EMC Centera Universal Access/FileArchiver · wrongly, and Richard Stallman agrees with me, when...

EMC® Centera® Universal Access/FileArchiverVersion 4.0.1

Monitor and Administration GuideP/N 300-004-806

REV A03

EMC CorporationCorporate Headquarters:

Hopkinton, MA 01748-9103

1-508-435-1000www.EMC.com

2

Copyright © 2003 - 2009 EMC Corporation. All rights reserved.

Published July, 2009

EMC believes the information in this publication is accurate as of its publication date. The information is subject to change without notice.

THE INFORMATION IN THIS PUBLICATION IS PROVIDED “AS IS.” EMC CORPORATION MAKES NO REPRESENTATIONS OR WARRANTIES OF ANY KIND WITH RESPECT TO THE INFORMATION IN THIS PUBLICATION, AND SPECIFICALLY DISCLAIMS IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

Use, copying, and distribution of any EMC software described in this publication requires an applicable software license.

For the most up-to-date listing of EMC product names, see EMC Corporation Trademarks on EMC.com.

All other trademarks used herein are the property of their respective owners.

Third-Party License Agreements

EMC Centera Software Development Kit

The EMC Software Development Kit (SDK) contains the intellectual property of EMC Corporation or is licensed to EMC Corporation from third parties. Use of this SDK and the intellectual property contained therein is expressly limited to the terms and conditions of the License Agreement.

Use of GPL

The EMC® version of Linux®, used as the operating system on the Centera server, is a derivative of Red Hat® and SuSE Linux. The operating system is copyrighted and licensed pursuant to the GNU General Public License (GPL), a copy of which can be found in the accompanying documentation. Please read the GPL carefully, because by using the Linux operating system on the Centera server, you agree to the terms and conditions listed therein.

SKINLF

This product includes software developed by L2FProd.com (http://www.L2FProd.com/).

Bouncy Castle

The Bouncy Castle Crypto package is Copyright © 2000 of The Legion Of The Bouncy Castle (http://www.bouncycastle.org).

RSA Data Security

Copyright © 1991-2, RSA Data Security, Inc. Created 1991. All rights reserved.

License to copy and use this software is granted provided that it is identified as the “RSA Data Security, Inc. MD5 Message-Digest Algorithm” in all material mentioning or referencing this software or this function. RSA Data Security, Inc. makes no representations concerning either the merchantability of this software or the suitability of this software for any particular purpose. It is provided “as is” without express or implied warranty of any kind.

These notices must be retained in any copies of any part of this documentation and/or software.

EMC Centera Universal Access/FileArchiver Monitor and Administration Guide

E

ICU License (IBM International Component on Unicode library)

Copyright (c) 1995-2002 International Business Machines Corporation and others. All rights reserved.

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, provided that the above copyright notice(s) and this permission notice appear in all copies of the Software and that both the above copyright notice(s) and this permission notice appear in supporting documentation.

THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT OF THIRD PARTY RIGHTS. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR HOLDERS INCLUDED IN THIS NOTICE BE LIABLE FOR ANY CLAIM, OR ANY SPECIAL INDIRECT OR CONSEQUENTIAL DAMAGES, OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE

ReiserFS

ReiserFS is hereby licensed under the GNU General Public License version 2.

Source code files that contain the phrase “licensing governed by reiserfs/README” are “governed files” throughout this file. Governed files are licensed under the GPL. The portions of them owned by Hans Reiser, or authorized to be licensed by him, have been in the past, and likely will be in the future, licensed to other parties under other licenses. If you add your code to governed files, and don't want it to be owned by Hans Reiser, put your copyright label on that code so the poor blight and his customers can keep things straight. All portions of governed files not labeled otherwise are owned by Hans Reiser, and by adding your code to it, widely distributing it to others or sending us a patch, and leaving the sentence in stating that licensing is governed by the statement in this file, you accept this. It will be a kindness if you identify whether Hans Reiser is allowed to license code labeled as owned by you on your behalf other than under the GPL, because he wants to know if it is okay to do so and put a check in the mail to you (for non-trivial improvements) when he makes his next sale. He makes no guarantees as to the amount if any, though he feels motivated to motivate contributors, and you can surely discuss this with him before or after contributing. You have the right to decline to allow him to license your code contribution other than under the GPL.

Further licensing options are available for commercial and/or other interests directly from Hans Reiser: [email protected]. If you interpret the GPL as not allowing those additional licensing options, you read it wrongly, and Richard Stallman agrees with me, when carefully read you can see that those restrictions on additional terms do not apply to the owner of the copyright, and my interpretation of this shall govern for this license.

Finally, nothing in this license shall be interpreted to allow you to fail to fairly credit me, or to remove my credits, without my permission, unless you are an end user not redistributing to others. If you have doubts about how to properly do that, or about what is fair, ask. (Last I spoke with him Richard was contemplating how best to address the fair crediting issue in the next GPL version.)

MIT XML Parser

MIT XML Parser software is included. This software includes Copyright (c) 2002,2003, Stefan Haustein, Oberhausen, Rhld., Germany

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction, including without limitation the rights included in all copies or substantial portions of the Software.

MC Centera Universal Access/FileArchiver Monitor and Administration Guide 3

4

THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.


E

Contents

Figures ............................................................................................................................. 11

Tables............................................................................................................................... 13

Preface............................................................................................................................ 15

Chapter 1 IntroductionProduct overview.............................................................................. 20

System operation ....................................................................... 20Disaster recovery ....................................................................... 22Enhanced CUA availability...................................................... 22Support for very high object count ......................................... 24Retention periods/classes ........................................................ 25Error alerts .................................................................................. 25

CUA/CFA limitations ...................................................................... 26

Chapter 2 Installation and ConfigurationHardware specifications .................................................................. 28

EMC Centera – CUA Software and Hardware Requirement Clarification ................................................................................ 28Dell PowerEdge 2950 ................................................................ 29Dell PowerEdge 2850 ................................................................ 29Dell PowerEdge 2650 ................................................................ 30EMC Centera Gen4 node .......................................................... 30EMC Centera Gen3 node .......................................................... 30

Installing the software...................................................................... 32Site requirements ....................................................................... 32


Required network information................................................ 32Accounts and passwords.......................................................... 33Installation on a Dell ................................................................. 34Installation on EMC Centera Gen3 and Gen4 nodes............ 37

Initial CUA/CFA configuration ..................................................... 38Phase 1: Starting the CUA/CFA configuration..................... 39Phase 2: Entering CUA/CFA configuration information.... 40Phase 3: Finishing the CUA/CFA configuration .................. 43

Verifying CUA/CFA functionality................................................. 46Assumptions .............................................................................. 46Verify email home functionality.............................................. 46Verifying communication from CUA/CFA to the EMC Centera ........................................................................................ 46Verifying communication from CUA/CFA to multiple nodes 47Verifying writeback functionality ........................................... 49Verifying invalidation functionality ....................................... 50Verifying read functionality ..................................................... 51

Completing the CUA/CFA configuration .................................... 52

Chapter 3 Advanced ConfigurationAccessing the CUA user interfaces ................................................ 54

CUA setup utility ...................................................................... 54CUA graphical user interface (GUI) ....................................... 57

Connect to CUA using CIFS............................................................ 59Adding access control to a CIFS share ................................... 60Map a drive in Microsoft Windows ........................................ 71Editing the smb.conf file........................................................... 71

Connect to CUA using NFS ............................................................ 75Add NFS mount points ............................................................ 75Adding access control to NFS mount points ......................... 76Manipulating files and directories using NFS....................... 77

Configuring FTP access.................................................................... 78Configuring HTTP access................................................................ 79

Accessing files using HTTP...................................................... 80Restoring a CUA configuration ...................................................... 81

To restore a CUA........................................................................ 83Restoring a CUA in enhanced availability ............................ 85

Altering file protection..................................................................... 86Using access profiles ........................................................................ 88

Configuring CUA access profiles ............................................ 89Mapping access profiles ........................................................... 90

EMC Centera Universal Access/FileArchiver Monitor and Administration Guide6

E

Backward compatibility using PEA files .................................92Working with metadata.................................................................... 95

Using custom metadata............................................................. 95Browsing metadata .................................................................... 96

Testing email alerts............................................................................ 97Entering permanent licensing information ................................... 98

License key verification............................................................. 98Entering a permanent license................................................... 98

Setting retention periods ................................................................ 100Retention notes ......................................................................... 101

Changing CUA passwords ............................................................ 103Setting system time ......................................................................... 104

Using the CUA GUI................................................................. 104Using the CUA Setup Utility.................................................. 105Setting system time zone ........................................................ 105

Setting up SNMP............................................................................. 107System information.................................................................. 108Agent access.............................................................................. 109SNMP trap destinations .......................................................... 110Email alerts ............................................................................... 111

Configuring UTF-8 file name support ......................................... 112

Chapter 4 EMC Centera FileArchiverIntroduction ..................................................................................... 124

Policies ....................................................................................... 124Retention ................................................................................... 125Simulation ................................................................................. 125

Enabling EMC Centera FileArchiver............................................ 126Phase 1 - Enabling CFA ........................................................... 127Phase 2 - Enabling Filename Translation.............................. 128Phase 3 - Enable HTTP access from CFA.............................. 130Phase 4 - Give CFA NFS permissions ................................... 130Phase 5 - Enable Celerra FileMover....................................... 131Phase 6 - Establish connection from the Celerra Network Server ......................................................................................... 131

Configuring EMC Centera FileArchiver...................................... 134Configuring Digest Access Authentication.......................... 134Configuring the DNS setting.................................................. 135Using CFA access profiles....................................................... 136

Defining migration policies ........................................................... 138Editing the EMC Centera FileArchiver configuration file . 138EMC Centera FileArchiver configuration options ...............139

7MC Centera Universal Access/FileArchiver Monitor and Administration Guide

Example EMC Centera FileArchiver configuration file...... 145Scheduling a policy run ................................................................. 147

Deleting a schedule ................................................................. 148Manually starting and stopping a policy run............................. 149Simulating a policy run.................................................................. 150

Stopping a policy simulation ................................................. 151Monitoring EMC Centera FileArchiver....................................... 152

Viewing EMC Centera FileArchiver statistics ..................... 152Viewing EMC Centera FileArchiver activity logs............... 153Viewing the EMC Centera FileArchiver migration log...... 153

Managing orphans.......................................................................... 155Finding orphans....................................................................... 155Excluding orphans from deletion ......................................... 157Deleting orphans ..................................................................... 157Stopping the orphan processes.............................................. 158

Chapter 5 CUA/CFA Enhanced AvailabilityIntroduction to CUA/CFA enhanced availability ..................... 160Overview of CUA/CFA enhanced availability.......................... 161Types of CUA/CFA EA configurations....................................... 163

Local standby CUA/CFA....................................................... 164Remote standby CUA/CFA with remote EMC Centera .... 166

Configuring enhanced availability .............................................. 169Configuring the Active CUA/CFA....................................... 170Configuring the Standby CUA/CFA.................................... 171

Resolving an Active CUA/CFA failure....................................... 173Testing CUA/CFA failover .................................................... 174Initiating failover in a production environment ................. 175

Reconfiguring the old Active system........................................... 178Recovering unwritten data from the old Active CUA ....... 179Reconfiguring the old system................................................ 183Implementing second failovers (failback)............................ 184

Using a remote Standby CUA/CFA and EMC Centera............ 185Failing over to a Standby CUA/CFA and a secondary EMC Centera ....................................................................................... 187Swapping your primary site with your remote site ............ 188

Using a remote Standby CFA, EMC Centera, and Celerra ....... 190Failover for full remote CFA configuration ......................... 192

Using DNS round robin with CFA............................................... 195Enabling DNS round robin .................................................... 196

Chapter 6 Monitoring CUA Status


E

Storage presentation ....................................................................... 200From a NFS mount point ........................................................ 200From a CIFS share .................................................................... 200Using the Monitor GUI ........................................................... 200

Accessing the GUI........................................................................... 201Monitoring system status............................................................... 203

Field descriptions..................................................................... 204Getting detailed node status.......................................................... 208

Field descriptions..................................................................... 209Monitoring services ........................................................................ 215

Field descriptions..................................................................... 215Monitoring a standby CUA ........................................................... 216Viewing writeback audit trails ...................................................... 218Viewing event messages ................................................................ 220

Field descriptions..................................................................... 221Disk monitoring .............................................................................. 223Changing the GUI password......................................................... 224

Chapter 7 Servicing the CUAUpgrading CUA on a Dell ............................................................. 228

Upgrading a Dell 2650 from v3.6 SP1 to v4.0.1.................... 228Upgrading a Dell 2850 from CUA v3.6 SP1 to 4.0.1............ 228

Upgrading CUA on an EMC Centera node ................................ 230Upgrade error messages................................................................. 231Changing CUA hardware .............................................................. 233Replacing disk drives and nodes .................................................. 235

Determining the drive to replace........................................... 235Disk replacement on a Dell..................................................... 238Disk replacement on an EMC Centera Gen3 or Gen4 node ..... 239

Connecting to CUA remotely ........................................................ 240Connecting to CUA with a modem....................................... 241

Chapter 8 Alerts and TroubleshootingEnabling email alert notification................................................... 244

Alert parameters ...................................................................... 244Responding to an email alert.................................................. 245

Index .............................................................................................................................. 257


E

Title Page

Figures

1 EMC Centera gateway ................................................................................... 202 EMC Centera Universal Access .................................................................... 213 Local EA topology .......................................................................................... 234 Dell 2950 back-panel indicators and features ............................................. 355 Dell 2850 back-panel indicators and features ............................................. 356 Dell 2650 back-panel indicators and features ............................................. 367 Login page ....................................................................................................... 578 Home page ...................................................................................................... 589 Set system time ............................................................................................. 10410 Configure SNMP .......................................................................................... 10711 EMC Centera FileArchiver .......................................................................... 12412 EA configuration with a local standby CUA ............................................ 16413 EA configuration with a local standby CFA ............................................ 16514 Remote standby CUA with remote EMC Centera .................................. 16715 CUA/CFA failover and failback procedure ............................................. 17316 Failover of remote Standby system with remote EMC Centera ............ 18617 EA configuration with remote standby CFA, remote EMC Centera, and re-

mote Celerra 19118 CFA Reads with DNS Round Robin .......................................................... 19519 Login page ..................................................................................................... 20120 Home page .................................................................................................... 20221 Node summary ............................................................................................. 20422 Node detail .................................................................................................... 20823 Monitor services ........................................................................................... 21524 Node detail page .......................................................................................... 21625 Event log ........................................................................................................ 22026 Monitoring disk repair ................................................................................ 22327 Changing an account password ................................................................. 22428 Disk failure on a Dell 2850 .......................................................................... 23629 Disk failure on a Dell 2650 .......................................................................... 236


Figures

30 Disk failure on an EMC Centera node ...................................................... 237


E

Title Page

Tables

1 Maximum number of objects .........................................................................242 CUA/CFA limitations ....................................................................................263 CUA/CFA access details ................................................................................344 CUA configuration prompts ..........................................................................415 CUA Setup Utility menu description ...........................................................556 smb.conf parameters for a new installation ................................................657 smb.conf values for the new share ................................................................688 krb5.conf values ...............................................................................................699 smb.conf values ...............................................................................................7210 CUA access details ........................................................................................10311 System Information panel elements ...........................................................10812 Agent Access panel elements ......................................................................10913 SNMP Trap Destination panel elements ....................................................11014 SNMP trap events .........................................................................................11015 Email Alert panel elements ..........................................................................11116 Locales supported by CUA ..........................................................................11217 FileArchiver Migration Log Information ...................................................15418 CUA/CFA failover overview ......................................................................16119 Stages of CUA EA configuration and failover ..........................................17420 Source column acronyms .............................................................................22121 Upgrade error messages ...............................................................................23122 CUA email alerts ...........................................................................................245


Tables


E

Preface

As part of an effort to improve and enhance the performance and capabilities of its product lines, EMC periodically releases revisions of its hardware and software. Therefore, some functions described in this document may not be supported by all versions of the software or hardware currently in use. For the most up-to-date information on product features, refer to your product release notes.

If a product does not function properly or does not function as described in this document, please contact your EMC representative.

Audience The audience for this Monitor and Administration Guide is a customer who is responsible for administering the EMC Centera® Universal Access (CUA) that links a fixed content application to an EMC Centera. This guide assumes that the CUA administrator is familiar with the Local Area Network (LAN) architecture and configuration at the site. It also assumes that the administrator has familiarity with the operation of a personal computer, workstation, or laptop running Microsoft Windows, Solaris, or Linux operating systems and the use of a Web browser (Netscape Navigator or Microsoft Internet Explorer).

Relateddocumentation

Other EMC Centera system publications include:

◆ EMC Centera Universal Access/EMC Centera FileArchiver Monitor and Administration Guide, P/N 069001199

◆ EMC Centera Programmer's Guide, P/N 069001127

◆ EMC Centera API Reference Guide, P/N 069001185

◆ Cabinet Setup Guide for the 40 U Cabinet, P/N 014003099


16

Preface

◆ Site Preparation and Unpacking Guide for the 40 U Cabinet, P/N 014003100

◆ EMC Centera Quick Start Guide, P/N 300-002-546

◆ EMC Centera Online Help, P/N 300-002-656

Conventions used inthis document

EMC uses the following conventions for special notices.

Note: A note presents information that is important, but not hazard-related.

CAUTION!A caution contains information essential to avoid data loss or damage to the system or equipment.

IMPORTANT!An important notice contains information essential to operation of the software.

WARNING

A warning contains information essential to avoid a hazard that can cause severe personal injury, death, or substantial property damage if you ignore the warning.

DANGER

A danger notice contains information essential to avoid a hazard that will cause severe personal injury, death, or substantial property damage if you ignore the message.

Typographical conventionsEMC uses the following type style conventions in this document:

Normal Used in running (nonprocedural) text for:• Names of interface elements (such as names of windows,

dialog boxes, buttons, fields, and menus)• Names of resources, attributes, pools, Boolean expressions,

buttons, DQL statements, keywords, clauses, environment variables, filenames, functions, utilities

• URLs, pathnames, filenames, directory names, computer names, links, groups, service keys, file systems, notifications


E

Preface

Where to get help EMC support, product, and licensing information can be obtained as follows.

Product information — For documentation, release notes, software updates, or for information about EMC products, licensing, and service, go to the EMC Powerlink website (registration required) at:

http://Powerlink.EMC.com

Technical support — For technical support, go to EMC Customer Service on Powerlink. To open a service request through Powerlink, you must have a valid support agreement. Please contact your EMC sales representative for details about obtaining a valid support agreement or to answer any questions about your account.

Bold: Used in running (nonprocedural) text for:• Names of commands, daemons, options, programs,

processes, services, applications, utilities, kernels, notifications, system call, man pages

Used in procedures for:• Names of interface elements (such as names of windows,

dialog boxes, buttons, fields, and menus)• What user specifically selects, clicks, presses, or types

Italic: Used in all text (including procedures) for:• Full titles of publications referenced in text• Emphasis (for example a new term)• Variables

Courier: Used for:• System output, such as an error message or script • URLs, complete paths, filenames, prompts, and syntax when

shown outside of running text

Courier bold: Used for:• Specific user input (such as commands)

Courier italic: Used in procedures for:• Variables on command line• User input variables

< > Angle brackets enclose parameter or variable values supplied by the user

[ ] Square brackets enclose optional values

| Vertical bar indicates alternate selections - the bar means “or”

{ } Braces indicate content that you must specify (that is, x or y or z)

... Ellipses indicate nonessential information omitted from the example


http://powerlink.emc.com

18

Preface

Your comments Your suggestions will help us continue to improve the accuracy, organization, and overall quality of the user publications. Please send your opinion of this document to:

[email protected]


1

This chapter introduces EMC Centera Universal Access (CUA), an application that provides access to an EMC Centera cluster through NFS, CIFS, FTP, and HTTP protocols.

This chapter provides the following:

◆ Product overview............................................................................... 20◆ CUA/CFA limitations ....................................................................... 26

Introduction

Introduction 19

20

Introduction

Product overviewEMC® Centera® Universal Access (CUA) enables applications using network file protocols (NFS, CIFS, FTP, and HTTP) to store and retrieve fixed content from EMC Centera. CUA acts as a high performance store and forward protocol translator and data cache. It communicates with application servers using network file protocols and with an EMC Centera cluster through the EMC Centera native API. As shown in Figure 1 on page 20, the CUA appears to application servers as a standard NFS or CIFS file server accessed over the LAN.

Figure 1 EMC Centera gateway

System operation Figure 2 on page 21 shows, in more detail, how CUA converts between network file protocols and the EMC Centera native API. Application servers connect to mount points on CUA using NFS or CIFS protocols.

When an application writes a file to the CUA, CUA stores the file locally and writes the new file data to EMC Centera at a time defined by an internal policy. Generally, data will start to be written back to the EMC Centera within 1 minute of the last update to a file.

When an application reads a file from the CUA, the data is returned from the local CUA storage if it is currently cached (a “hit”) or is fetched immediately from EMC Centera if it is not currently stored on the CUA (a “miss”). With the exception of access time, an access to


Introduction

the EMC Centera performed on behalf of an NFS or CIFS read is completely transparent to the application server.

A database on the CUA holds the content addresses of the data stored on the EMC Centera. On a daily basis, CUA automatically backs up sufficient information to permit a complete restore of the database in the event of a catastrophic CUA failure.

To manage its local disk cache, CUA automatically invalidates data for files which have current copies on the EMC Centera when the total CUA storage in use exceeds 70%. When CUA reaches very high levels of storage, it will automatically throttle its write performance to allow the invalidate process to free sufficient disk space. This algorithm is designed to minimize the chance that the CUA is forced to return an out of space error for new files.

Figure 2 EMC Centera Universal Access

Product overview 21

22

Introduction

Disaster recovery A standalone CUA has a built-in disaster recovery feature. Every night at 01:14 (AM local time) the CUA runs a backup utility that stores backup information on the EMC Centera. The system then sends an email to the administrator with the Backup Content Address (Backup CA). In the event of a system failure, the CUA can be reinstalled or the system can be replaced with a new CUA. The administrator or CE can then do a restore using the Backup CA to restore the file metadata to the CUA.

This backup and restore operation is intended for disaster recovery purposes, not for file restore operations. Only files that were written to an EMC Centera will be restored. Files that have not yet been written to an EMC Centera will not be recovered. Files are written to the EMC Centera one minute after the last modification to the file. Timing is dependent on file size and the number of new files. Files written to an EMC Centera after the backup will be restored because the restore operation performs a time-based query on the EMC Centera to recover those files.

Enhanced CUA availability

CUA supports enhanced availability through an optional dedicated Standby CUA. The Standby CUA is designed to continuously monitor the functions of the main CUA (Active CUA) and provides the ability to manually failover to the Standby CUA in the event of the failure of the Active CUA. The Active CUA logs the creation and deletion of objects stored on the EMC Centera and synchronizes this information with the Standby CUA so that service can be restored by transitioning the Standby CUA to an active state.

The Standby CUA is only aware of files that have been written to EMC Centera and is unaware of files written to the Active CUA but not yet written to the EMC Centera. A file is written to EMC Centera one minute after the last modification to the file. Timing is dependent on file size and the number of new files. Chapter 6, ”Monitoring CUA Status,” describes how to view the current status of the Standby CUA. Figure 3 on page 23 provides a graphical overview of a local EMC Centera EA topology.


Introduction

Figure 3 Local EA topology

The Standby CUA must have the same software version as the Active CUA. In addition, the Standby CUA must have a hardware configuration within the same hardware family and class (in other words, a Dell 2850 Active must use Dell 2850 Standby, and a Gen4 Active must use a Gen4 Standby). Active and Standby CUAs may be physically separated but must be within a corporate network. There should be no firewall between two CUAs. Each CUA must have a unique IP address and hostname. There is no requirement for a private network between the Active and Standby CUA.

The installer defines the initial operating mode of a CUA when the system is configured. You can install and configure a Standby CUA any time after the Active CUA is operational.

Note: The Active and Standby CUAs must be configured to point to the same EMC Centera.

Product overview 23

24

Introduction

Support for very high object count

CUA has the capability to map up to 200 million objects stored on an EMC Centera and referenced in CUA’s metadata. The actual maximum number of objects a CUA can store will vary by host and disk type.

Also, the number of objects is set during the initial install. If the CUA/CFA is upgraded to a release which supports a higher object count, this higher object count will not be realized unless the CUA/CFA is fresh installed and restored from a backup.

As an example in Table 1 on page 24, we have a Dell 2850 fresh installed at 3.5.x as supporting 200 million objects while the same server which is upgraded from 3.0.x or lower to 3.5.x as supporting only 100 million objects. The same issue exists for the Dell 2650 with 146GB disks. If this CUA was installed at 1.x, it can only support 10 million objects.

Note: CUA/CFA writes files of less than 74 KB as embedded blobs to EMC Centera. This is not a configurable option.

Table 1 Maximum number of objects

Platform Maximum number of objects

EMC Centera Gen4 Hardware 200 million

EMC Centera Gen3 Hardware 100 million

Dell 2950 200 million

Dell 2850 CUA 3.5 fresh installation 200 million

Dell 2850 CUA 3.5 upgrade 100 million

Dell 2650 with 146GB drives 100 million


Introduction

Retention periods/classes

The EMC Centera Governance Edition (GE) and Compliance Edition Plus (CE+) support retention periods. Retention periods assure storage of objects for a predetermined length of time.

CUA supports the configuration and setting of retention periods. CUA provides the ability to assign a retention period as an attribute to directories one level under the NFS and CIFS partitions. As the CUA writes the files to EMC Centera, the retention period is set. The files cannot be modified or deleted from the CUA or EMC Centera until the retention period has expired.

Note: CUA v4.0.1 also supports named retention classes; however, retention classes are not supported on CFA.

Error alerts The system automatically generates email alerts (if configured) and SNMP traps for the following error conditions:

◆ Disk failure

◆ File system full

◆ Number of directories or number of file per directory exceeds or is approaching limit

◆ No communication with the EMC Centera

◆ Network interface down

◆ Delivery application (NFS, CIFS) down

◆ CPU usage exceeds threshold

◆ Temporary license expired

Chapter 2, ”Installation and Configuration,” provides more information on SNMP trap setup and Chapter 8, ”Alerts and Troubleshooting,” contains the procedure to enable email alerts and for troubleshooting information.

Product overview 25

26

Introduction

CUA/CFA limitationsTable 2 on page 26 lists the limitations for CUA/CFA v4.0.1.

Table 2 CUA/CFA limitations

Element Limitation Notes

Number of application clients per CUA

1-6 clients With the restriction of a total of 24 threads

MaxThreads from all application clients

24 Maximum thread count for communication with client applications

Number of CUAs 2 Conditional on the EMC Centera configuration supporting 80 write threads.A CUA 4.0.1 and a CUA 4.0.1 running FileArchiver may point to the same EMC Centera.

Number of EMC Centera access nodes (Mn)

2 or 4 Gen3 and Gen4 hardware: Minimum of 2 access nodesGen1 and Gen2 hardware: Minimum of 4 access nodes

Maximum number of file systems (CFA)

65 per CFA 65 file systems are supported regardless of the number of Data Movers or the number of EMC Celerras.

MaxFiles (CUA) 200 million Maximum object count for CUA 4.0.1 - actual maximum varies by hardware platform.

Max File Size (CUA) 60 GB

Max Directories (CUA) 1 million

Max number of directories per folder 30 thousand

Max number of files per directory 100 thousand

Max thread count 40 Maximum SDK thread count for communication to the EMC Centera


2

This chapter explains how to install, set up, and perform the initial configuration of EMC Centera Universal Access/EMC Centera FileArchiver (CUA/CFA).

The main sections are:

◆ Hardware specifications.................................................................... 28◆ Installing the software ....................................................................... 32◆ Initial CUA/CFA configuration....................................................... 38◆ Verifying CUA/CFA functionality .................................................. 46◆ Completing the CUA/CFA configuration...................................... 52

Installation andConfiguration

Installation and Configuration 27

28

Installation and Configuration

Hardware specifications

EMC Centera – CUA Software and Hardware Requirement Clarification

To clear up any confusion in the field, EMC Centera Product Management has clarified their position on support of both the CUA software and the Dell hardware appliance sold via EMC Select. Below are the rules as provide by Product Management:

◆ All new CUA installations utilizing the Dell 2950 server must ship with both the CUA external software (CNRCUAESW) and EMC qualified Dell 2950 server available on EMC Select (CUA02-3).

◆ The Dell 2950 hardware must be purchased through EMC Select and is preconfigured to meet EMC specifications. This is the only external CUA hardware configuration that is EMC qualified.

◆ CUA external software (CNRCUAESW) implemented on the EMC Select Dell 2950 server (CUA02-3) is supported by EMC.

◆ CUA external software (CNRCUAESW) implemented on customer provided hardware is not supported by EMC.

◆ The customer is responsible for the installation of the EMC Select Dell 2950 server (CUA02-3). For hardware support issues, Dell directly supports the 2950 server hardware, not EMC. Dell servers purchased thru EMC Select include a 3-year 7x24 Dell service warranty.

◆ The CUA external software (CNRCUAESW) can be either customer- or EMC-implemented. PS-BAS-CUA (Implementation for Centera Universal Access) is recommended for new implementations if the customer requires an EMC implementation.

◆ For CUA implementations on an internal Centera Gen4 node (CNRCUA1NG4), the internal software (CNRCUAISW) can be sold separately and be installed by either the customer or EMC. CUA internal software (CNRCUAISW) implemented on internal Centera Gen4 nodes is supported by EMC.

(PS-BAS-CUA) Implementation for Centera Universal Access is a fixed-price service that installs and configures CUA for operating with an EMC Centera Storage System. The Service Brief as well as all accompanying documentation can be found on PowerLink.



Note: The rails that ship with the Dell systems do not fit the EMC Centera racks. EMC provides two different rail kits for the customer to purchase. Contact your EMC representative for more information on purchasing rails.

Note: The following specifications for Dell systems are provided for informational purposes only. Do not use them to purchase a Dell system from any other source besides EMC.

Dell PowerEdge 2950

◆ Single Dual Core Xeon Processor 5130 4 MB Cache, 2.00 GHz, 1333 MHz FSB

◆ 2 GB 667 MHz (4 x 512 MB), Single Ranked DIMMs

◆ PERC 5/i, x6 Backplane Integrated Controller Card

◆ Integrated SAS/SATA RAID 5 PERC 5/i Integrated

◆ 6 x 300 GB, SAS, 3.5-inch 10K RPM Hard Drive

◆ 1 x 24X IDE CD-RW/DVD ROM Drive

◆ Embedded Broadcom NetXtreme II5708 Gigabit Ethernet NIC

◆ Redundant power Supply

◆ Bezel for PE2950

◆ Rack Chassis w/Versarail RoundHole-Universal for third-party racks

Dell PowerEdge 2850

◆ 2 x 2.8 GHz or 2 x 3.2 GHz CPU with 1 MB Cache

◆ 2 GB DDR2 RAM, 2 x 1GB Single Ranked DIMMs

◆ Motherboard SCSI (341-1351)

◆ 1 x 6 Hard Drive Backplane PE2850 (311-4282)

◆ No hardware RAID

◆ 6 x 146 GB 10 K RPM 1" SCSI hot plug HDD

◆ 1 x 24 speed IDE internal CD-ROM drive

◆ 1 x dual On-Board NICS

◆ Redundant power Supply

◆ Bezel for PE2850 (313-2693)

◆ Versa Rails, third Party Rack

Hardware specifications 29

30


For other specifications, refer to the Dell PowerEdge 2850 hardware manuals delivered on CD with the hardware.

Dell PowerEdge 2650

◆ 2 x 2.0 GHz CPU with 512 K cache

◆ 2GB DDR RAM, 4 x 512 MB DIMMS

◆ On-board SCSI, 1-5 drives connected to on-board SCSI

Note: Dell 2650s with 73 GB drives are no longer supported. Dell 2650s with 146 GB drives are only supported for fresh installations.

◆ 5 bay (1x5) hot plug SCSI HD backplane

◆ No hardware RAID

◆ 5 x 146 GB 10 K RPM, 1” (ultra 160) SCSI hot plug HDD

◆ 1 x 24 Xspeed IDE internal CD-ROM drive

◆ 1 x dual on-board NICs (GigE copper)

◆ Redundant power supplies (2 x 500 Watt)

◆ Active bezel option for the Dell PowerEdge 2650

◆ “Versarack” for non-Dell racks or “Rapid Rails” for Dell rack

For other specifications, refer to the Dell PowerEdge 2650 hardware manuals delivered on CD with the hardware.

Note: During the CD-ROM load, the software checks the server BIOS to validate that the hardware platform is an approved Dell PowerEdge 2650 with the above specifications.

EMC Centera Gen4 node

Refer to the EMC Centera Online Help, P/N 300-002-547 for the hardware specifications of an EMC Centera Gen 4 node.

Note: An EMC Centera Gen4 node must be running CentraStar version 3.0 or higher.

EMC Centera Gen3 node

Refer to the EMC Centera Online Help, P/N 300-002-547 for the hardware specifications of an EMC Centera Gen 4 node.



Note: An EMC Centera Gen3 node must be running CentraStar version 2.3 or higher.

Hardware specifications 31

32


Installing the softwareAfter the hardware has been installed and powered up, carry out the following procedures to install the software, configure the device for the customer’s network, and verify that the server is working properly.

Note: Before beginning installation on a Dell, refer to the manufacturer’s documentation for an overview of the hardware. If there is a bad disk drive on the Dell, you should not attempt to install the CUA/CFA software.

CAUTION!Under no circumstances are you allowed to install third party applications on this machine, nor are you allowed to place data anywhere other than the mount/share points gateway_cifs and gateway_nfs using supported file sharing protocols.

Site requirements This version of CUA/CFA runs on either a Dell PowerEdge 2950, a Dell PowerEdge 2850, a Dell PowerEdge 2650 (146 GB drives only) or an EMC Centera Gen3 or Gen4 node.

The network requirements for all servers are a 100/1000BaseT (copper) Transmission Control Protocol/Internet Protocol (TCP/IP) based LAN. Gen4 nodes also support optical networking. During the installation, you will assign a host name, network IP address, subnet mask, and gateway/router address to the CUA/CFA. Contact the customer’s network administrator for this information.

Refer to the site CCRF for additional requirements and specifications.

Required network information

Gather the following information before continuing:

◆ The fully-qualified domain name (the nodename, myhost.company.com, for example)

◆ IP address for the default TCP/IP gateway

◆ DNS search domain

◆ DNS Server IP addresses

◆ ethn interface IP address, network mask, speed, duplex



◆ CUA/CFA username and password. Refer to “Accounts and passwords” on page 33 for more information.

◆ Mail server (SMTP) IP address

◆ Email addresses of the people to be notified if an event is triggered

◆ SMTP From field

◆ SMTP Replyto field

◆ EMC Site ID

◆ NTP server IP address

◆ Name or IP address of the EMC Centera cluster (you may enter multiple IP addresses)

◆ Metadata backup email address

◆ IP address of SMTP server for backup email

After the initial configuration, you can use the CUA Setup Utility to edit or change system settings. “Advanced Configuration” on page 53 contains instructions on changing system settings.

Accounts and passwords

To perform various installation and configuration procedures, you need to provide a CUA/CFA username and password.

Table 3 on page 34 lists the default CUA/CFA login and password details.

Note: “Changing CUA passwords” on page 103 provides instructions on changing the CUA/CFA password. If you changed the default passwords for any account and then upgraded to a later version of CUA, passwords are set to the default values for the new version.

Installing the software 33

34


Installation on a Dell Carry out the following procedure to install CUA/CFA on a Dell:

1. Do one of the following:

• Connect a PC or laptop to the serial port on the back of the chassis.

Note: If you install the Dell over the serial cable, you must also configure CUA/CFA over the serial cable. You may unplug the cable once the system has been installed and configured.

• Connect a monitor and keyboard to the video and keyboard connectors on the back of the chassis.

Refer to Figure 4 on page 35, Figure 5 on page 35, and Figure 6 on page 36 for the locations of these ports.

Note: Use a DB-9 null modem cable (not supplied) to connect a PC or laptop to the serial port (to serial port 1 on a Dell 2650). The DB-9 serial port runs at a speed of 9600 baud with specifications of 8n1. Start a HyperTerminal session with settings of VT100, 9600, and 8,n,1.

Note: A Dell 2950 requires bios version 2.2.6 or higher.

Table 3 CUA/CFA access details

MAP Login Password

CUA AdminGUI AdminFTP Admin

gwsetupsysmongwsetup

EMCCUAEMCCUAEMCCUA

FTP NFS gateway_nfs gateway_nfs

FTP CIFS gateway_cifs gateway_cifs



Figure 4 Dell 2950 back-panel indicators and features


EB143702112004ETKR

USB connectors (2)10/100/1000 NIC connectors (2)

Status LEDs

PS/2 Mouse connectorPS/2 Keyboard connector

Video connectorSerial port connector

Service connector

Primary power connectorSecondary power connector


36



2. Start a HyperTerminal session.

3. Insert the installation CD into the Dell’s drive.


• If this is a new pristine system, power cycle to reboot.

• If this is an existing system, type reboot from the prompt on the laptop.

After the system reboots, the following message appears:

At the "boot:" prompt below, you will need to type either serial or monitor. Entering "serial" will connect to the first serial console; Entering "monitor" will connect to a monitor and keyboard; Hitting enter at the boot prompt will select serial console by default.

The boot prompt will default to "serial" after 60 seconds.




• If using a laptop connected to the CUA/CFA, choose serial (default) by pressing Enter.

Note: After 60 seconds, the system will default to serial.

• If using a monitor and keyboard directly attached to the CUA/CFA, type monitor and press Enter.

The following message appears:

Loaded USB storage support driver <usb-storage>Installation from CD-ROM media detected.In order to prevent accidental installation, please type "yes" below.Type "yes" and press the Enter key to continue installation:

6. Enter yes and press Enter. The system reads the data off of the CD. After the data is read, the CD ejects and the installation begins. Installation will take approximately a half hour.

After the system reboots, the installation is complete and you are prompted to log in. Enter gwsetupand type theEMCCUA password. The system will now launch the EMC Centera Universal Access Setup Utility. You use this utility to configure and license the Gateway. “Initial CUA/CFA configuration” on page 38 provides instructions on configuring the CUA/CFA.

Installation on EMC Centera Gen3 and Gen4 nodes

For installation on an EMC Centera Gen 3 or Gen 4 node, contact your EMC representative.


38


Initial CUA/CFA configurationThe CUA/CFA configuration consists of the following phases:

Prerequisites Before configuring CUA/CFA you must do the following:

◆ Obtain valid information such as IP addresses and subnet masks from the network administrator before proceeding with the CUA/CFA configuration (refer to “Required network information” on page 32 for more information).

◆ The EMC Centera serial number must be available and the profile you use for authentication must have the Monitor capability enabled. Refer to the CLI commands set owner and profile update in the EMC Centera On-line Help.

Notes ◆ During configuration, you may have to delete the defaults (use backspace) to correctly type the configuration options.

◆ You can run the CUA/CFA Setup Utility at any time after the installation process by logging into the system. Refer to Chapter 3, ”Advanced Configuration,” for more information.

◆ On a Dell, if you performed the installation over a serial cable you must also perform the configuration over the serial cable. You may unplug the cable once the system has been installed and configured.

Phase Activity

“Phase 1: Starting the CUA/CFA configuration” on page 39

Connect to the CUA/CFA and begin the configuration

“Phase 2: Entering CUA/CFA configuration information” on page 40

Enter configuration information

“Phase 3: Finishing the CUA/CFA configuration” on page 43

Complete and test the configuration



Phase 1: Starting the CUA/CFA configuration

To start the configuration, do the following:

1. Once the system has rebooted the following menu appears:

Welcome to the EMC Centera Universal Access Setup Utility

--------------------------------------B) Browse MetadataC) Configuration optionsD) Configure Centera FileArchiverM) Send E-mail notificationO) Set CLI operational parameters (editor, locale)P) Alter File ProtectionR) Restore system configurationS) Shutdown systemT) Reboot systemU) Upgrade systemV) Initiate FailoverW) Backup systemX) Enter bash shell

Q or E) Exit

2. Enter the License key by carrying out the procedure “Entering permanent licensing information” on page 98.

3. Select C at the Setup Utility menu to open the CUA Configuration menu.

4. Select C from the CUA Configuration menu to begin the configuration:

• If you are using a Gen3 node or a Dell 2650 system, go to “Phase 2: Entering CUA/CFA configuration information” on page 40.

• If you are using a Gen4 node or a Dell 2850/2950 system, go to “Bonding CUA/CFA NICs” on page 39.

Bonding CUA/CFANICs

For Gen4 nodes and Dell PowerEdge 2850/2950 systems, CUA/CFA supports the bonding of the two available Ethernet ports to the same IP address. This feature provides port redundancy if one NIC fails.

In addition, CUA/CFA also supports the eth3 optical port on Gen4 nodes.

Bonding NICs on a Gen4 NodeAfter you type C at the CUA Configuration menu, CUA/CFA displays the following prompt:

Initial CUA/CFA configuration 39

40


Do you want to bond eth2 and eth0 as one adapter (Y/N)?

1. Type Y if you want to bond the NICs, or press Enter to accept the default N value.

2. If you entered N, the system displays the following prompt:

Do you want to use the optical port as your main adapter (Y/N)? :

Type Y to use the optical port, or press Enter to accept the default value.

3. Go to “Phase 2: Entering CUA/CFA configuration information” on page 40 to continue the configuration.

Bonding NICs on a Dell 2850/2950After you type C at the CUA Configuration menu, CUA/CFA displays the following prompt:

Do you want to bond eth0 and eth1 as one adapter (Y/N)?

1. Type Y if you want to bond the NICs, or press Enter to accept the default N value.

2. Go to “Phase 2: Entering CUA/CFA configuration information” on page 40 to continue the configuration.

Phase 2: Entering CUA/CFA configuration information

Table 4 on page 41 lists the prompts that appear during the CUA/CFA configuration. You can either type in new information at the ? prompts or repeatedly press Enter to move through the prompts to accept the default values without making changes.

After you have entered all the configuration information, go to “Phase 3: Finishing the CUA/CFA configuration” on page 43 to complete the installation.

Note: The CUA/CFA does not validate IP addresses or host names on your network before saving the configuration. Settings made during the initial configuration are saved as default values. During this initial configuration, you must enter all of the network parameters. The only optional fields are related to Email Notification and NTP.



Following is an explanation of the required information at each prompt:

Table 4 CUA configuration prompts (1 of 2)

Prompt Input

Hostname in FQDN format: Type the hostname for the system in Fully-Qualified Domain Name (FQDN) format.

Note: FQDN name consists of a text string of up to 24 characters from the alphabet (A-Z), digits (0-9), minus sign (-), and period (.). No blank or space characters, or any other special characters are permitted as part of a domain name.

DNS search domain: Type the domain name for this node.

DNS IP addresses (comma separated, MAX 3):

Type the IP addresses of up to 3 DNS servers used by this node. Separate each IP address with a comma.

Default TCP/IP gateway address:

Type the default gateway IP address.

ethn interface IP Address: Type the IP address for the network interface in dotted-decimal format; for example, 190.118.40.1.

Note: To remove or clear an address, enter the value 0.0.0.0.

ethn Network Mask: Type the network mask for this interface; for example, 255.255.255.0.

ethn Speed (0=auto, 10/100/1000):

Type the Ethernet speed for this interface; for example, 100 or 1000.

ethn Duplex (0=auto, 1=half, 2=full):

Type the Ethernet duplex for this interface.

Email Alert Settings

IP Address of SMTP server: Type the address of any valid SMTP server; for example the EMC Centera Alert workstation address.

Note: If an alert is issued, CUA first tries to reach the default email (MX) server for the domain. If the MX server cannot be reached, CUA then uses the SMTP configured in this step.

E-mail alert recipients (comma separated):

Type a list of email recipients separated by commas. This list must include the following addresses:• [email protected]• [email protected]

SMTP From field: Type the email address of the sending host (where the message originated).


42


SMTP ReplyTo field: Type the email address to whom reply messages will be sent (where your reply will be sent).

EMC Site Id: Type the EMC site identification as assigned by the EMC service group for tracking purposes.

NTP Settings

Enable NTP (Y/N): Input Yes to enable Network Time Protocol (NTP). NTP is used to synchronize the time of a server to another server or reference time source. Enabling NTP will keep all nodes in the network time synchronized so that email and file time stamps are coordinated. The default value is No (disabled).

NTP Servers (comma separated, MAX 2):

If you entered Yes in step 1, enter up to two NTP servers separated by commas.

EMC Centera Specific Settings

Note: If the customer site supports dual CUAs the setup and configuration of the parameters in this section must be exactly the same on each CUA.

Centera Names or IP Addresses (comma separated):

Type the names or IP addresses of all the access nodes on the EMC Centera cluster. If you do not configure all access nodes during the CUA/CFA installation you may see a fault detector error: Application Centera is not running.

Backup e-mail address: Type the backup email address (the default entry for this field are the entries that were made for the SMTP server above).

Note:If EMC addresses appear in this field, you must delete them.

Without a valid address in the Backup e-mail address field, there is no way to obtain the Content Address (CA) from the daily backup to the EMC Centera. If the CUA/CFA crashes you will need the most recent CA when you run the Setup Utility to restore the CUA/CFA configuration. Refer to “Restoring a CUA configuration” on page 81.

You cannot use # in front of an email address when setting backup email addresses.

Back-up E-mail SMTP server address:

Type an SMTP server to be used for delivering mail to the end user. Typically, this is not the EMC Centera Alert SMTP address (the default entry for this field is the entry that was made for the SMTP server above).

Back-up time: Enter the time that the daily backup should run. The default format is mm:hh AM.

Table 4 CUA configuration prompts (2 of 2)

Prompt Input



Phase 3: Finishing the CUA/CFA configuration

To complete the configuration:

1. When prompted, press any key. The system displays the first of three summary pages similar to those shown in the examples below. View the different pages by typing 1, 2, or 3 on your keyboard.

Example of page 1 Welcome to the EMC Centera Universal Access Module. Hostname : domain.company.com*Search Domain : company.com Nameserver : xxx.xxx.xxx.xx Nameserver : Nameserver : Default TCP/IP gateway : xxx.xxx.xxx.xx eth2 Address : xxx.xxx.xxx.xx/255.255.255.0 eth2 Speed : Auto eth2 Duplex : Auto eth0 Address : 0.0.0.0 eth0 Speed : Auto eth0 Duplex : Auto eth1 Address : 0.0.0.0 eth1 Speed : Auto eth1 Duplex : Auto eth3 Address : 0.0.0.0 eth3 Speed : Unknown eth3 Duplex : Unknown

* = modified setting page 1 of 3----------------------------------------------1..3) Page numberR) Retry setupS) Save changes & Exit

Q or E) Exit without saving changes

Example of page 2 CURRENT Configuration----------------------------------------------------

SMTP Server address : mail.legato.com E-mail recipients : [email protected] SMTP From field : SMTP ReplyTo field : EMC Site Id : EMC Cluster Serial Number : NTP enabled : Yes NTP Server : xx.xxx.xxx.xx

* = modified setting page 2 of 3----------------------------------------------------


44


1..3) Page numberR) Retry setupS) Save changes & Exit


Example of page 3 CURRENT Configuration---------------------------------------------------

CenteraId :cube3a,cube3b,cube3c,cube3dMetaBackupEmail : [email protected] : mail.legato.comMetaBackupTime : 01:14 AM

* = modified setting page 3 of 3---------------------------------------------------

1..3) Page numberR) Retry setupS) Save changes & Exit


2. Select R to restart the Setup Utility and make further changes or select S to save all changes.

If you save, the Setup Utility reboots and displays messages similar to the example shown below:

Changes successfully saved.Press any key to reboot.

3. Press any key. The following appears:

The system is going down for reboot NOW.

Note: If you receive an error when removing the node because of multiple non-online instances, re-execute the command.

4. After reboot completes, go to a remote system on the network, open a command shell, and test the interface with the ping command using the address of the CUA/CFA.

Example ping IP Address

Pinging node name.domain.com [102.118.100.13] with 32 bytes of data:Reply from 102.118.100.13: bytes=32 time<10ms TTL=254Reply from 102.118.100.13: bytes=32 time<10ms TTL=254Reply from 102.118.100.13: bytes=32 time<10ms TTL=254Reply from 102.118.100.13: bytes=32 time<10ms TTL=254



Ping statistics for 102.118.100.13: Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),Approximate round trip times in milliseconds: Minimum = 0ms, Maximum = 0ms, Average = 0ms

C:\>


46


Verifying CUA/CFA functionalityThis section details the steps required to verify proper CUA/CFA functionality. The steps below will verify that the correct writeback, invalidation, and retrieval functions are occurring between the CUA/CFA and the EMC Centera.

Assumptions It is assumed that the proper network parameters have been configured and cabling is in place to allow the CUA/CFA to communicate over the LAN to the EMC Centera.

Verify email home functionality

After installation, it is very important that you verify that the CUA/CFA can send email alerts to the EMC Customer Service Center. Follow the procedure in “Testing email alerts” on page 97 to test email home functionality.

Verifying communication from CUA/CFA to the EMC Centera

To verify communication from the CUA/CFA to the EMC Centera, do the following:

1. Log in to the CUA/CFA using an SSH client and log in as gwsetup using the password EMCCUA.

2. Select X at the Setup Utility menu to enter the Bash Shell.

3. Verify that the CUA/CFA is capable of communicating to the EMC Centera by typing the following command:

sudo sfsproxyd_centera_info

Output similar to the following should be returned:

Centera status:version: 1.2.0.114-620cluster id: ddb2a5cc-1dd1-11b2-b4f1-c74d29d43aafcapacity: 2434130806755free: 1718987034275

If this returns successfully, you have made an API call from the CUA/CFA to the EMC Centera. If it does not, it is likely that the IP addresses/DNS names for the EMC Centera have not been set properly or here is a network issue that requires resolution.



Verifying communication from CUA/CFA to multiple nodes

To verify communication to multiple nodes in an EMC Centera cluster, do the following:


2. Verify that the CUA/CFA can communicate with the EMC Centera, by typing the following command:

/usr/local/centera/bin/CenteraPing -address <IP_address1>,<IP_address2>,<IP_address3>

Note: Replace <IP_address1>, <IP_address2>, <IP_address3> with the correct IP addresses of the EMC Centera nodes with the access role.

Example command /usr/local/centera/bin/CenteraPing -address 190.160.110.31,190.160.110.32,190.160.110.33

Expected result 190.160.110.31 is accessible190.160.110.32 is accessible190.160.110.33 is accessible

3. Verify that the CUA/CFA can write to the EMC Centera, by typing the following command:

/usr/local/centera/bin/CenteraVerify -address <IP_address1>,<IP_address2>,<IP_address3> [-authentication {name=<name>,secret=<secret>} | {<pea file>}]

Notes:

• Replace <IP_address1>, <IP_address2>, <IP_address3> with the correct IP addresses of the nodes with the access role.

• If using authentication, indicate the name/secret pair or PEA file with the -authentication switch.

• The verify command can only be run from the directory in which you entered the Bash Shell (/usr/local/storigen/gwsetup).

Verifying CUA/CFA functionality 47

48


Example command /usr/local/centera/bin/CenteraVerify -address 190.160.110.31 -authentication name=c2c,secret=c2c

/usr/local/centera/bin/CenteraVerify -address 190.160.110.31 -authentication peafile.pea

Expected result Open the centeraverify.log file to view the results:

Test options-------------Num Files: 3File Size (Kb): 10File Size (Kb): 100File Size (Kb): 1000Restore Folder: /tmp/Log File Name: ./centeraverify.logGenerate file folder: .-------------

Step 1

Open 190.160.110.31 Success

Step 2

Version: 2.3.2-215-396-5792Capacity (GB): 36,279FreeCapacity (GB): 25,795ClusterID: f8e9bbc8-1dd1-11b2-8d3e-d38dfa138xy8Write allowed: trueRead allowed: trueDelete allowed: truePurge allowed: false

Step 3

Write successWrite time (ms): 3270.69Write throughput (MBps): 0.0031Read successRead time (ms): 211.851Read throughput (MBps): 0.0483

Write successWrite time (ms): 3280.57Write throughput (MBps): 0.0312Read successRead time (ms): 300.98Read throughput (MBps): 0.3402

Write successWrite time (ms): 3915.14



Write throughput (MBps): 0.2615Read successRead time (ms): 138.399Read throughput (MBps): 7.3989

Average Write throughput (MBps): 0.0986312Average Read throughput (MBps): 2.59582End of test-----------

Verifying writeback functionality

Writeback is the process where data is written from the CUA to the EMC Centera. If writeback is not functioning properly, the CUA will be unable to store data on the EMC Centera. Following is an overview of the steps necessary to verify Writeback:

◆ Copy a file from outside the /sfs directory structure into the /sfs/gateway_xxx directory structure.

Note: In the path above, replace gateway_xxx with either gateway_cifs or gateway_nfs depending on which directory you are performing the writeback on.

◆ Verify that the EMC Centera informed the CUA/CFA of the C-Clip Descriptor File (CDF) value.

To verify Writeback, do the following:


2. Copy the file /usr/local/storigen/centera_checkfile to /sfs/gateway_nfs/centera_test by typing the following command:

cp /usr/local/storigen/centera_checkfile /sfs/gateway_nfs/centera_test

3. Wait one minute before continuing.


50


4. Read the returned CDF to verify success by typing the following command:

sudo sfsproxyd_read_rfi -CAH /sfs/gateway_nfs/centera_test

Output similar to the following will be returned:

Remote File Id: 7JIE12QTI7733e1B4E4CI7N9254G60VVF6TDPE000000000000000 Resident: YesDirty: NoFile Name: /sfs/gateway_nfs/centera_test

Note: It will take one minute for the file to be written to the EMC Centera. If you do not see the remote file ID in the output, repeat the command until the file ID appears.

Note: The sudo sfsproxyd_read_rfi command provides usage information (a list of valid switches) if the command is run without arguments.

Verifying invalidation functionality

Invalidation is the process where the CUA/CFA frees up disk space in the CUA/CFA file system by deleting data blocks for files that have already been written to the EMC Centera.

To execute the command to invalidate the blocks used by /sfs/gateway_nfs/centera_test, do the following:


2. Type the following command:

sudo sfsproxyd_invalidate_blocks -F -s /sfs/gateway_nfs/centera_test

3. To verify that the file's data blocks are no longer in the CUA/CFA cache, type the following command:

sudo sfsproxyd_read_rfi -CAH /sfs/gateway_nfs/centera_test



4. Output similar to the following will appear:

Remote File Id: 7JIE12QTI7733e1B4E4CI7N9254G60VVF6TDPE000000000000000 Resident: NoDirty: NoFile Name: /sfs/gateway_nfs/centera_test

The Resident flag shows false, verifying that the file's data blocks are no longer in the CUA/CFA cache.

Verifying read functionality

After configuration, it is important to verify that the file can be read back from the EMC Centera.

To verify read functionality, do the following:



diff –s /usr/local/storigen/centera_checkfile /sfs/gateway_nfs/centera_test

Read functionality is successful if centera_checkfile and centera_test are identical.


52


Completing the CUA/CFA configurationAfter the initial configuration of the CUA/CFA, there are advanced configuration steps that you should carry out to enable the full functioning of your CUA/CFA system.

For CUA If using the system as CUA:

◆ Set up access protocols:

• “Connect to CUA using CIFS” on page 59

• “Connect to CUA using NFS” on page 75

• “Configuring HTTP access” on page 79

◆ Set retention periods on your data—“Setting retention periods” on page 100.

In addition to these tasks, there are other advanced configuration tasks listed in Chapter 3, ”Advanced Configuration.”

For CFA If using the system as CFA:

◆ Enable CFA on the CUA/CFA system —“Enabling EMC Centera FileArchiver” on page 126.

◆ Configuring CFA features such as:

• “Configuring Digest Access Authentication” on page 134

• “Configuring the DNS setting” on page 135

• “Using CFA access profiles” on page 136

◆ Define your data migration policies—“Defining migration policies” on page 138.

In addition to these tasks, there are other configuration tasks listed in Chapter 4, ”EMC Centera FileArchiver.”


3

This reference chapter explains how to change the various configuration options and gives an explanation of the two CUA user interfaces.


◆ Accessing the CUA user interfaces.................................................. 54◆ Connect to CUA using CIFS............................................................. 59◆ Connect to CUA using NFS.............................................................. 75◆ Configuring FTP access ..................................................................... 78◆ Configuring HTTP access ................................................................. 79◆ Restoring a CUA configuration ....................................................... 81◆ Altering file protection ...................................................................... 86◆ Using access profiles.......................................................................... 88◆ Working with metadata..................................................................... 95◆ Testing email alerts ............................................................................ 97◆ Entering permanent licensing information .................................... 98◆ Setting retention periods................................................................. 100◆ Changing CUA passwords ............................................................. 103◆ Setting system time.......................................................................... 104◆ Setting up SNMP.............................................................................. 107◆ Configuring UTF-8 file name support .......................................... 112

AdvancedConfiguration

Advanced Configuration 53

54

Advanced Configuration

Accessing the CUA user interfacesThis section provides information on how to access the two CUA user interfaces used to configure the CUA:

◆ CUA Setup Utility

◆ CUA Graphical User Interface (GUI)

CUA setup utility Connect to the CUA using an SSH client and log in as gwsetup using the password EMCCUA.

The following menu appears:

Welcome to the Centera Universal Access Setup Utility-----------------------------------------------------B) Browse MetadataC) Configuration optionsD) Configure Centera FileArchiverM) Send E-mail notificationO) Set CLI operational parameters (editor, locale)P) Alter File ProtectionR) Restore system configurationS) Shutdown systemT) Reboot systemU) Upgrade systemV) Initiate FailoverW) Backup systemX) Enter bash shell

Q or E) Exit

Note: Changing the Hostname, DNS info, Gateway, IP address, Subnet mask, Speed, Duplex, NIC bonding, or EMC Centera IP addresses will cause the system to reboot.

Table 5 on page 55 gives an overview of the options in this menu and where you can find instructions for the menu item in this manual.



Table 5 CUA Setup Utility menu description (1 of 2)

Action Location

B) Browse Metadata Refer to “Browsing metadata” on page 96.

C) Configuration options Opens the CUA configuration menu that contains the following options:

C) Configure Centera Universal Access

Refer to “Initial CUA/CFA configuration” on page 38.

F) Configure FTP access Refer to “Configuring FTP access” on page 78.

G) Configure Access Profiles Refer to “Using access profiles” on page 88.

L) Enter licenses Refer to “Entering permanent licensing information” on page 98.

M) Configure Custom Metadata Refer to “Using custom metadata” on page 95.

N) Modify NFS configuration Refer to “Connect to CUA using NFS” on page 75.

R) Configure Retention Refer to “Setting retention periods” on page 100.

P) Change passwords Refer to “Changing CUA passwords” on page 103.

S) Modify CIFS configuration Refer to “Connect to CUA using CIFS” on page 59.

V) Configure Enhanced Availability Refer to “Configuring enhanced availability” on page 169.

D) Configure Centera FileArchiver Refer to “EMC Centera FileArchiver” on page 123.

M) Send E-mail notification Refer to “Testing email alerts” on page 97.

O) Set CLI operational parameters (editor, locale)

Refer to “Configuring UTF-8 file name support” on page 112.

P) Alter File Protection Refer to “Altering file protection” on page 86.

R) Restore system configuration Refer to “Restoring a CUA configuration” on page 81.

Accessing the CUA user interfaces 55

56


S) Shutdown system Shuts down CUA.

Note: Do not shut down CUA unless you have physical access to the machine. To restart CUA, push the power button.

After you choose to shutdown the system, you will see following prompt:Please note that shutting down the system will save some information in the system logs, but not the complete current system state. If you are shutting down the system because you are experiencing problems with the current operation of the system, you may wish to intentionally crash the system in order to capture the complete system state.Do you want to crash the system to capture the complete state?:

T) Reboot system Reboots CUA.After you choose to reboot the system, you will see following prompt:Please note that rebooting the system will save some information in the system logs, but not the complete current system state. If you are rebooting the system because you are experiencing problems with the current operation of the system, you may wish to intentionally crash the system in order to capture the complete system state.Do you want to crash the system to capture the complete state?:

U) Upgrade system Refer to “Servicing the CUA” on page 227.

V) Initiate Failover Refer to “Resolving an Active CUA/CFA failure” on page 173.

W) Backup system Refer to “Restoring a CUA configuration” on page 81.

X)Enter bash shell Exits the menu to the Bash Shell.

Table 5 CUA Setup Utility menu description (2 of 2)

Action Location



CUA graphical user interface (GUI)

Access CUA’s GUI through a standard web browser. Use the following URL (this URL uses the HTTPS (HTTP Secure) protocol):

https://<machine_name>:7227/

Replace machine with the host name or IP address of the CUA.

Example The following example connects to a CUA named demo1:

https://demo1:7227/

Once connected to the CUA GUI, the browser displays the Login page shown in Figure 7 on page 57.

Figure 7 Login page

Enter a valid user name and password for the system monitor account, then click Login. The default account is named gwsetup and the default password, EMCCUA.

Note: Username and password entries are case-sensitive. Type them exactly as shown.

Accessing the CUA user interfaces 57

58


After logging in, the page shown in Figure 8 on page 58 appears. This is the Home Page of the CUA GUI

Figure 8 Home page



Connect to CUA using CIFSThis section describes how to connect to CUA through the CIFS protocol. This section also provides information on the following topics:

◆ “Adding access control to a CIFS share” on page 60

◆ “Map a drive in Microsoft Windows” on page 71

◆ “Editing the smb.conf file” on page 71

The following examples show how to connect to a CUA using a CIFS/SMB connection to access or add content to a CUA.

CUA ships with the share \\<machine-name>\gateway_cifs as the default CIFS share. If you want more than one CIFS share, do the following:


2. Change directory to /sfs/gateway_cifs.

3. Type the mkdir command to make directories

Example cd /sfs/gateway_cifs

mkdir shareA

mkdir shareB

Note: You cannot export directories that contain UTF-8 characters; however, directories beneath the mount point can contain UTF-8 characters.

4. The new directories are owned by the user gwsetup.

5. Type exit to return to the CUA Setup Utility menu.

6. To change ownership of the subdirectories, select P at the Setup Utility menu to display the following:

File :User :Group :Permissions :Recurse down directory tree (Y/N) :

7. Input gateway_cifs at the User prompt.

8. Input gateway_cifs at the Group prompt.

Connect to CUA using CIFS 59

60


9. The new sub-directories are now owned by gateway_cifs.

10. To modify /etc/samba/smb.conf to export the multiple mount points, select C at the Setup Utility menu to enter the CUA Configuration menu.

11. Select S at the CUA Configuration menu to enter the editor.

12. In the editor, remove the section for [gateway_cifs].

Add the following to the end of the file:

[shareA]path = /sfs/gateway_cifs/shareAguest account = gateway_cifsforce user = gateway_cifswriteable = yesguest ok = yes[shareB]path = /sfs/gateway_cifs/shareBguest account = gateway_cifsforce user = gateway_cifswriteable = yesguest ok = yes

Samba will automatically reread the configuration file within two minutes. You should then be able to mount the shares \\machine-name\shareA and \\machine-name\shareB.

Adding access control to a CIFS share

CUA ships without access control in place for the \\machine-name\gateway_cifs share; however, you can add access control through one of the following methods:

◆ “Add IP access control to a CIFS share” on page 60.

◆ “Adding Windows NT authentication to CIFS” on page 61.

◆ “Adding Active Directory Authentication to CIFS” on page 65.

Add IP access controlto a CIFS share

To add an IP Access Control List to a CIFS share, do the following:

1. Select C at the Setup Utility menu to enter the CUA Configuration menu.

2. Select S at the CUA Configuration menu to enter the editor.

3. Go to the section with your share (default is [gateway_cifs]) and add a new line to this section in one of the following formats:



Example In this example, only clients from 190.160.0.23 will be able to access the share:

hosts allow = 190.160.0.23

Example In this example, only clients in the 190.160.0.0 network will be able to access the share:

hosts allow = 190.160.0.0/255.255.255.0

Always run the command /usr/bin/testparm after modifying the smb.conf file to verify that the configuration options are correct. Output similar to the following appears:

bash-2.04$ /usr/bin/testparmLoad smb config files from

/etc/samba/smb.confProcessing section "[serviceinfo]"Processing section "[gatewayd]"Processing section "[gateway_cifs]"Processing section "[.alogro$]"Processing section "[.alogrw$]"Loaded services file OK.

Samba will automatically reread the configuration file within 2 minutes.

Adding Windows NTauthentication to CIFS

To add Windows NT authentication to CIFS, you must do one of the following:

1. Configure the /etc/samba/smb.conf file with one of the following procedures:

• “Add Windows NT authentication to CIFS for new installations” on page 62.

• “Add Windows NT authentication to CIFS for upgrades” on page 63.

2. Configure Kerberos and finish the Active Directory configuration. Refer to “Finish Windows NT authentication configuration” on page 64 for instructions.


62


Add Windows NT authentication to CIFS for new installationsTo add Windows NT authentication to CIFS, do the following:

1. Log in to the CUA using an SSH client and log in as gwsetup using the password EMCCUA.


3. Select S at the CUA Configuration menu to modify the CIFS configuration.

4. Modify the CIFS configuration file:

a. Add or modify the following variables in the global section:

workgroup = <DOMAIN>security = domainpassword server = <PDC>idmap uid = 10000-20000idmap gid = 10000-20000encrypt passwords = yeswinbind enum users = nowinbind enum groups = no

Note: Replace <PDC> with the hostname of primary domain controller and <DOMAIN> with the NT domain name.

b. Add or modify the following variables in the gateway_cifs share to look like the following:

path = /sfs/gateway_cifsbrowseable = noread only = yeswriteable = noguest ok = novalid users = <DOMAIN\USER1> <DOMAIN\USER2>write list = <DOMAIN\USER1> <DOMAIN\USER2># force user = gateway_cifs# create mask = 0774# directory mask= 0775

Note: Replace <DOMAIN\USERx> with the NT domain and user name.

c. Save the CIFS configuration file.

d. Select Q or E to return to the Setup Utility main menu.



5. Proceed to “Finish Windows NT authentication configuration” on page 64 to finish the configuration.

Add Windows NT authentication to CIFS for upgradesTo store new files under Windows NT Authentication, you continue to use the share name gateway_cifs. To enable Windows NT Authentication for new files, you must complete the steps in “Add Windows NT authentication to CIFS for new installations” on page 62. This procedure restricts /sfs/gateway_cifs access to the defined Windows NT users.

To allow user gateway_cifs to continue accessing existing files under gateway_cifs, you must create a new share (for example: gateway_xxxx) to /sfs/gateway_cifs, and force to the gateway_cifs user.

To add the gateway_xxxx share, do the following:

1. Log in to the CUA as gwsetup using the password EMCCUA.


3. Select S at the CUA Configuration menu to modify the CIFS configuration.

4. In the /etc/samba/smb.conf file:

5. Modify the CIFS configuration file:

a. Add another share called gateway_xxxx with the following parameters:

[gateway_xxxx]path = /sfs/gateway_cifs/browseable = noguest ok = novalid users = <DOMAIN\USER1> <DOMAIN\USER2>write list = <DOMAIN\USER1> <DOMAIN\USER2>force user = gateway_cifswriteable = yescreate mask = 0774directory mask = 0775

Note: Replace xxxx with a logical name for the share.

Replace <DOMAIN\USERx> with the NT domain and user name.

b. Save the CIFS configuration file.


64


c. Select Q or E to return to the Setup Utility main menu.

6. Proceed to “Finish Windows NT authentication configuration” on page 64 to finish the configuration.

Finish Windows NT authentication configuration1. Select P at the Setup Utility to alter the file protection. Enter the

following at the prompts:

Alter File Protection---------------------File : /sfs/gateway_cifs/User : gateway_cifsGroup : gateway_cifsPermission : 777Recurse down directory tree (Y/N) : N

2. Select X at the Setup Utility to enter the Bash Shell.


sudo /usr/bin/net rpc join -S <PDC> -U <ADMIN>

Note: Replace <PDC> with the hostname of primary domain controller and <ADMIN> with the privileged Windows NT Domain user.

4. Enter the password for the Windows NT Domain user.

5. Restart the CIFS Daemon, nmb, and winbind by typing the following commands:

sudo /etc/init.d/smb restartsudo /etc/init.d/nmb restartsudo /etc/init.d/winbind restart

6. Verify that the CUA can ping, by typing the following:

/usr/bin/wbinfo -p

7. Verify that the CUA can view users and groups within the Windows Domain, by typing the following command:

/usr/bin/wbinfo -ug

8. Type exit at the Bash Prompt to return to the Setup Utility.



Adding Active Directory Authentication to CIFSTo add Active Directory authentication to CIFS, you must do one of the following:

1. Configure the /etc/samba/smb.conf file with one of the following procedures:

• “Add Active Directory authentication to CIFS for fresh installations” on page 65.

• “Add Active Directory authentication to CIFS for upgrades” on page 67.

2. Configure Kerberos and finish the Active Directory configuration. Refer to “Finish Active Directory authentication configuration” on page 68 for instructions.

Add Active Directory authentication to CIFS for fresh installationsTo add Windows Active Directory authentication to CIFS, do the following:



3. Select S at the CUA Configuration menu to modify the CIFS configuration file.

4. Edit parameters in the /etc/samba/smb.conf file and save the file. The bolded parameters in Table 6 on page 65 are the ones that you must change and/or add.

Table 6 smb.conf parameters for a new installation (1 of 3)

smb.conf parameters Notes

[global]log level = 0os level = 1

No change.

security = ADS Set value to ADS.

server string = EMC CIFS Server No change.

workgroup = <domain name> Replace <domain name> with the hostname of the CUA.

realm = <DOMAIN.TEST.COM> Replace <DOMAIN.TEST.COM> with the name of the CUA’s workgroup. Value must be uppercase.


66


password server = <DC1> Replace <DC1> with the hostname of the DC. Value must be uppercase.

idmap uid = 10000-20000 Set idmap uid range to 10000-20000.

idmap gid = 10000-20000 Set idmap gid range to 10000-20000.

log file = /var/log/samba/%m.log No change.

encrypt passwords = yes Make sure value is yes.

syslog = 1 No change.

winbind enum users = no Set value to no.

winbind enum groups = no Set value to no.

# Performance Enhancementssocket options = SO_RCVBUF=65536

SO_SNDBUF=65536strict sync = yesoplocks = yeskernel oplocks = no

# write cache size = 1048576wide links = yeschange notify timeout = 120deadtime = 1case sensitive = nomap to guest = bad user

# use sendfilev = no# read size has been deprecated

[serviceinfo]path = /sfs/serviceinfo/browseable = noforce user = serviceinfowriteable = noguest ok = yes

[gateway_cifs]path = /sfs/gateway_cifs/

No change.

browseable = no Set value to no.

read only = yes Set value to yes.

writeable = no Set value to no.

guest ok = no Set value to no.

valid users = <DOMAIN\USER1> <DOMAIN\USER2>

Add the list of users that will be granted rights to the share.





5. Proceed to “Finish Active Directory authentication configuration” on page 68 to finish the configuration.

Add Active Directory authentication to CIFS for upgradesTo store new files under Active Directory Authentication, you continue to use the share name gateway_cifs. To enable Active Directory Authentication for new files, you must complete the steps in “Add Active Directory authentication to CIFS for fresh installations” on page 65. This procedure restricts /sfs/gateway_cifs access to the defined Active Directory users.

To allow user gateway_cifs to continue accessing existing files under gateway_cifs, you must create a new share (for example: gateway_xxxx) to /sfs/gateway_cifs, and force to the gateway_cifs user.

To add the gateway_xxxx share, do the following:



3. Select S at the CUA Configuration menu to modify the CIFS configuration file.

4. In the /etc/samba/smb.conf file:

a. Add another share called gateway_xxxx after the existing [gateway_cifs] share.

b. Add the variables shown in Table 7 on page 68 to the new share.

write list = <DOMAIN\USER1> <DOMAIN\USER2>

Grant the above users write access to the share.

# force user = gateway_cifs Disable this setting by commenting out this line with an #.

# create mask = 0774 Disable this setting by commenting out this line with an #.

# directory mask= 0775 Disable this setting by commenting out this line with an #.




68


c. Save the CIFS configuration file.

d. Select Q or E to return to the Setup Utility main menu.

5. Proceed to “Finish Active Directory authentication configuration” on page 68 to finish the configuration.

Finish Active Directory authentication configuration1. Modify the Kerberos configuration file:

a. Select X at the Setup Utility to enter the Bash Shell.

b. Edit values /etc/krb5.conf file and save the file. The bolded parameters in Table 8 on page 69 are the ones that you must change and/or add.

Table 7 smb.conf values for the new share

smb.conf parameter Instructions

[gateway_xxxx] Replace xxxx with the logical name of the new share.

path = /sfs/gateway_cifs/ Set path to /sfs/gateway_cifs/

browseable = no Set value to no.

force user = gateway_cifs Set force user to gateway_cifs

writeable = yes Set value to yes.

create mask = 0774 Set value to 0774.

directory mask= 0775 Set value to 0775.

guest ok = no Set value to no.



2. Return to the Setup Utility and alter the file protections on /sfs/gateway_cifs:

a. Select P at the Setup Utility to alter the file protection.

b. Enter the following at the prompts:

Alter File Protection---------------------File : /sfs/gateway_cifs/User : gateway_cifsGroup : gateway_cifsPermission : 777Recurse down directory tree (Y/N) : N


Table 8 krb5.conf values

krb5.conf parameters Notes

[logging]default = FILE:/var/log/krb5libs.logkdc = FILE:/var/log/krb5kdc.logadmin_server =

FILE:/var/log/kadmind.log

[libdefaults]ticket_lifetime = 24000

No change.

default_realm = <domain.test.com> Replace <domain.test.com> the name of the CUA’s domain. Value must be lowercase.

dns_lookup_realm = true Set value to true.

dns_lookup_kdc = true Set value to true.

[realms]

<DOMAIN.TEST.COM> = { Replace <DOMAIN.TEST.COM> with the name of the CUA’s domain. Value must be uppercase.

kdc = <DC1>}

Replace <DC1> with the hostname of the Domain Controller (not the FQDN). Value must be uppercase.

[domain_realm]

.<domain.test.com> = <DOMAIN.TEST.COM> Replace <domain.test.com> with the name of the CUA’s domain. The first value must be lowercase and the second value uppercase. A dot must precede the first value.

dns_lookup_kdc = true Set value to true.


70


4. Restart the CIFS Daemon and the NetBIOS Daemon by typing the following commands:

sudo /etc/init.d/smb restartsudo /etc/init.d/nmb restart

5. Adjust the clock skew by typing the following command:

sudo /usr/bin/net time set -S <server>

Note: Replace <server> with the name of the Kerberos server.

6. Verify the Active Directory information by typing the following command:

/usr/bin/net ads info

7. Join the Active Directory Domain:

a. Type the following command:

sudo /usr/bin/net ads join -S <server> -U <ADMIN>

Note: Replace <server> with the name of the Kerberos server and <ADMIN> with the name of the Active Directory administrator.

b. Enter the password for the Windows Active Directory administrator.

8. Start the winbind Daemon by typing the following command:

sudo /etc/init.d/winbind restart


/usr/bin/wbinfo -p




Map a drive in Microsoft Windows

To create a mapped drive to a CUA named node_name.domain_name.com from a PC running Windows, do the following:

1. Open Windows Explorer.

2. Choose Tools > Map Network Drive...

3. In the Folder field type the following:

\\<node_name.domain_name.com>\gateway_cifs

You can now open the gateway_cifs or gateway_xxxx mapped drive in Windows Explorer and copy or drag-and-drop files to the CUA. However, the following security conditions may occur:

◆ If this is a new install, with no previous data on the system, and you map to the gateway_cifs directory you may be prompted for the domain username and password if you are not already logged into that domain.

◆ If you map to the gateway_cifs on either a new installation or an upgrade you may be prompted for the domain username and password if you are not already logged into that domain.

◆ If this is an upgrade and there was data on the system, and you map to the gateway_xxxx directory you will need to provide the username gateway_cifs and the password gateway_cifs.

Editing the smb.conf file

CUA uses Samba software to provide and regulate access by CIFS clients. Samba is configured through the /etc/samba/smb.conf file. The following example shows the raw headers for the smb.conf file:

Example [global] log level = 0 os level = 1 security = share server string = EMC CUA Samba Server workgroup = EMC_CUA log file = /var/log/samba/%m.log encrypt passwords = yes syslog = 1

# Performance Enhancements socket options = SO_RCVBUF=65536 SO_SNDBUF=65536 strict sync = yes oplocks = yes kernel oplocks = no# write cache size = 1048576


72


wide links = yes change notify timeout = 120 deadtime = 1 case sensitive = no map to guest = bad user

You can change some of the parameters in the smb.conf file; however, there are certain parameters that you absolutely must not change. Table 9 on page 72 specifies which parameters you can change and which you cannot change.

Table 9 smb.conf values (1 of 3)

Parameter

Can be edited by customer Notes

log level yes Default: 0Allowable values: numberSets the logging level used. Values of 3 or more slow Samba noticeably. A synonym is debug level. Recommended value: 1.

os level yes Default: 1Allowable values: numberSets the candidacy of the server when electing a browse master. Used with the domain master or local master options. You can set a higher value than a competing operating system if you want Samba to win. Windows for Workgroups and Windows 95 use 1, Windows NT client uses 17, and Windows NT Server uses 33.

security NO Important: Do not change this value.

server string yes Default: EMC CUA Samba ServerAllowable values: stringSets the name that appears beside a server in browse lists. Honors the %v (Samba version number) and %h (hostname) variables.

workgroup yes Default: EMC_CUAAllowable values: workgroup nameSets the workgroup to which things will be served. Overrides compiled-in value. Choosing a name other than WORKGROUP is strongly recommended.

log file NO Important: Do not change this value.

encrypt passwords

NO Important: Do not change this value. The default value is yes.

syslog NO Important: Do not change this value.



socket options

yes Default: SO_RCVBUF=65536 SO_SNDBUF=65536Allowable values: listSets OS-specific socket options. The main options are:TCP_NODELAY Have the server send as many packets as necessary to keep delay low. This is used on telnet connections to give good response time, and is used - somewhat counter-intuitively - to get good speed even when doing small requests or when acknowledgments are delayed (as seems to occur with Microsoft TCP/IP). This is worth a 30-50 percent speedup by itself. Incidentally, in Samba 2.0.4, socket options = TCP_NODELAY became the default value for that option.IPTOS_LOWDELAY This is another option that trades off throughput for lower delay, but which affects routers and other systems, not the server. All the IPTOS options are new; they're not supported by all operating systems and routers. If they are supported, set IPTOS_LOWDELAY whenever you set TCP_NODELAY.SO_SNDBUF and SO_RCVBUF The send and receive buffers can often be the reset to a value higher than that of the operating system. This yields a marginal increase of speed (until it reaches a point of diminishing returns). SO_KEEPALIVE This initiates a periodic (four-hour) check to see if the client has disappeared. Expired connections are addressed somewhat better with Samba's keepalive and dead time options. All three eventually arrange to close dead connections, returning unused memory and process-table entries to the operating system.

strict sync NO Important: Do not change this value.

oplocks yes Default: yesAllowable values: YES, NOIf YES, support local caching of opportunistic locked files on client. This option is recommended because it improves performance by about 30%. For more information, see also fake oplocks and veto oplock files.

kernel oplocks

NO Value must be NO. Important: Do not change this value.

wide links yes Default: yesAllowable values: YES, NOIf set to YES, Samba will follow symlinks out of the current disk share(s). See also the root dir and follow symlinks options.

change notify timeout

yes Default: 120Allowable values: positive numberSets the number of seconds between checks when a client asks for notification of changes in a directory. Used to limit the performance cost of the checks. Avoid lowering.


Parameter



74


deadtime yes Default: 1 Allowable values: minutesThe time in minutes before an unused connection will be terminated. Zero means forever. Used to keep clients from tying up server resources forever. If used, clients will have to auto-reconnect after minutes of inactivity. For more information see also keepalive.

case sensitive

NO Important: Do not change this value.

map to guest NO Important: Do not change this value.


Parameter




Connect to CUA using NFSThe following example shows how to connect to the CUA using an NFS connection.

Note: Do not use the async option when mounting an NFS file system. The async option can result in serious instability of the CUA system.

The mount command is case-sensitive. If you use the wrong case, the CUA generates a Permission Denied error message.

To create a mount point on a Linux or UNIX system to gain access to a CUA, do the following:

1. Open a shell window and become root.

2. At the prompt type:

mount -t nfs <node_name.company.com>:/sfs/gateway_nfs /mnt/<directory_name>

Add NFS mount points

CUA ships with NFS exporting the mount point /sfs/gateway_nfs. Additional mount points must be under /sfs/gateway_nfs. To create additional mount points, do the following:


2. Change directory to /sfs/gateway_nfs.

3. Type the mkdir command to make directories.

Example mkdir mountpointA

mkdir mountpointB


4. The new directories are owned by the user gwsetup.

5. Type exit to return to the CUA Setup Utility menu.

Connect to CUA using NFS 75

76


6. To change ownership of the subdirectories, select P at the Setup Utility menu to display the following:


7. Input gateway_nfs at the User prompt.

8. Input gateway_nfs at the Group prompt.

9. The new sub-directories are now owned by gateway_nfs.

10. To modify /etc/exports to export the multiple mount points, select C at the Setup Utility menu to enter the CUA Configuration menu.

11. Select N at the CUA Configuration menu to enter the editor.

12. In the editor, change the line for /sfs/gateway_nfs to:

/sfs/gateway_nfs/mountpointA

13. Add the following line:

/sfs/gateway_nfs/mountpointB

Adding access control to NFS mount points

CUA ships without access restrictions for /sfs/gateway_nfs mount point. To add an IP Access Control List to an NFS mount point:


2. Select N at the CUA Configuration menu to enter the editor.

3. Go to the line with your mount point (default is /sfs/gateway_nfs) and change the * to one of the following formats:

Example An IP address — In this example, only clients from 192.168.0.23 will be able to access the mount point:

/sfs/gateway_nfs 192.168.0.23(rw,no_root_squash,insecure,sync)

Example IP networks — In this example, only clients in the 192.168.0.0 network will be able to access the mount point:

/sfs/gateway_nfs 192.168.0.0/255.255.255.0(rw,no_root_squash,insecure,sync)



Example This is an example of multiple IP addresses — if the list of IP addresses exceeds a single line, split the list into two parts as follows:

/sfs/gateway_nfs/dir1 10.2.4.5(rw) 10.2.3.4(rw)/sfs/gateway_nfs/dir1 12.1.1.1(rw) 10.1.1.2(rw)

Manipulating files and directories using NFS

This section discusses some issues when manipulating files on an NFS share.

Moving files underretention

Moving a file under retention may give unexpected results if you are using NFS.

The issue is with NFS using the /bin/mv command. If you attempt to move a file containing an unexpired retention period, the move will copy the file to the new name and then give an error while trying to delete the file with the old name. The /bin/mv command successfully copies the file then fails while trying to remove the old filename because the retention period has not expired. For example, moving a file named test with an unexpired retention period produces the following:

#mv test test.renamedmv: cannot unlink ’test’: Permission deniedmv: cannot remove ’test’: Permission denied#ls -ltotal 1-rw-r--r-- 1 root root 2 May 19 11:27 test-rw-r--r-- 1 root root 2 May 19 11:27 test.renamed

Using rmdir Doing a rmdir (example: rm -rf directory) over NFS may require the client to issue the rmdir multiple times. Depending on the size of the directory and how the NFS client mounted, only a portion of a directory may be removed with each rmdir call.

Connect to CUA using NFS 77

78


Configuring FTP accessFTP access to a CUA is turned off by default. You can turn on FTP access to both the /sfs/gateway_nfs and /sfs/gateway_cifs directories. Access to the service account can also be turned on.

Note: The home directory of the service account is /var/tmp/ftp.

To turn on FTP access, do the following:


2. Select F at the CUA Configuration menu to display the following:

FTP access----------Enable ftp access to account gateway_cifs? (Y/N):Enable ftp access to account gateway_nfs? (Y/N):Enable ftp access to account gwsetup? (Y/N):

Enter Y to enable access to the CIFS, and/or NFS, and /or gwsetup directories.

Note: Refer to “CUA access details” on page 103 for login and password details for FTP. To change FTP passwords, refer to “Changing CUA passwords” on page 103.



Configuring HTTP accessHTTP access to CUA is turned on by default for the directory /sfs/logs. Turn on HTTP access to both the /sfs/gateway_nfs and /sfs/gateway_cifs directories, by doing the following:


2. Change directory, by typing the following:

cd /etc/apache2/vhosts.d/

3. Edit the configuration file, by typing the following command:

pico cua_httpd.conf

4. Use the search functionality of the editor to find the following lines under the section Directory "/sfs":

Order deny, allow Allow from 127.0.0.1 Deny from all

5. Change the lines above to the required specification. To allow everyone access, change the lines to:

Order allow, deny Allow from all

6. Exit the editor.

7. Restart HTTP, by typing the following command at the Bash Shell prompt:

sudo /etc/init.d/apache2 restart

8. Once HTTP has restarted, HTTP access is enabled.

Configuring HTTP access 79

80


Accessing files using HTTP

To access files on the CUA using HTTP, use the following URLs:

◆ To access the file /sfs/gateway_nfs/dirA/fileB use:

http://<server.company.com>/gateway_nfs/dirA/fileB

◆ To access the file /sfs/gateway_cifs/dirA/fileB use:

http://<server.company.com>/gateway_cifs/dirA/fileB

◆ To access the log files that contain the list of files that were written to the EMC Centera use:

http://<server.company.com>/logs/writeback

Note: Replace server.company.com above with the Fully Qualified Domain Name of the CUA.



Restoring a CUA configurationCUA backup and restore features provide a way to recover the file system if there is a catastrophic failure. CUA backup is a backup of metadata and not of the actual files.

The restore feature is used in case of a catastrophic failure of the CUA. It cannot be used for retrieving an older version of a file. Restore is only supported on a standalone CUA.

CUA maintains a daily backup of:

◆ The CUA configuration information

◆ File system directory information

◆ All EMC Centera metadata for files stored since the last backup

The backup is done at 01:14 (AM local time) to the EMC Centera. Additionally, an administrator may force a backup by selecting W at the Setup Utility menu.

You cannot restore a CUA configuration to a system that already contains content. If there is a catastrophic hardware failure, install the new hardware, reinstall the CUA, then use this procedure to restore the configuration (to reinstall the CUA, refer to “Installation and Configuration” on page 27).

To restore the configuration of a CUA that contains content do one of the following:

◆ On a Dell, you must use the CD-ROM to reinstall the software before doing the restore.

◆ On an EMC Centera Gen 3 or Gen 4 Node, contact your EMC representative to reinstall the software.

Reinstalling the software will prepare the disks for the CUA restore process.

Note: This process restores all file metadata for files written to the EMC Centera before the CUA failed, thereby restoring access to these files via the CUA. Files that were on the CUA that had not yet been written to the EMC Centera will not be restored. In a CUA EA environment, the query portion of the restore is based on the metadata from the original CUA and CA.

Restoring a CUA configuration 81

82


Before starting this procedure, note the following:

◆ Hostname — you may use the same hostname or choose a new one.

◆ Network information

◆ DNS search domain string

◆ DNS IP address

◆ Default TCP/IP gateway address

◆ ethn interface IP address

◆ ethn network mask

◆ ethn speed

◆ ethn duplex

◆ All Email Alert settings

◆ Name or IP address of the EMC Centera cluster(s) — use the name or IP of the CUA that you are replacing.

Restores and accessprofiles

You must configure an access profile for restore operations—this profile is used to read the backup files from the EMC Centera. The access profile used for a restore operation must be the same access profile used to create the backup file. For information on creating and mapping an access profile for backups/restores, refer to “Using access profiles” on page 88.

Restores andconfiguration

changes

If you restore CUA/CFA from a backup and you have made configuration changes after the date of the backup file, you will need to update configuration information (including access profiles) that has changed since the date of the backup used for the restore.

Metadata backupemail address

Before restoring a CUA configuration, a Metadata Backup Email Address must be assigned while configuring CUA. An email message with the Backup Content Address (CA) is mailed to this email address each time the system is backed up.



The email looks similar to the following:

CUA cua6.cua.emc.com (xx.xxx.xxx.xxx) v2.1 Full Backup started at Tue Apr 13 01:14:00 2004

Configuration files backup completed at Tue Apr 13 01:14:02 2004

XMD backup completed at Tue Apr 13 01:14:07 2004

Aggregate backup completed at Tue Apr 13 01:14:08 2004

CUA cua6.cua.emc.com (xx.xxx.xxx.xxx) v2.1 Full Backup completed at Tue Apr 13 01:14:08 2004

CA is5TPRHQKAG1POEe9B5PV9R7DB81HG60VFGBT5LM000000000000000

Restore times The length of the restore depends on the size of your CUA database and how many files were written since the backup file you use to do the restoration. A restoration on a Dell 2950 with 200 million records and using the most recent backup takes approximately 4 hours. To minimize restore time, make sure you use the most recent backup file to restore the system.

To restore a CUA Note: If you need to restore a CUA in an Enhanced Availability configuration, refer to “Restoring a CUA in enhanced availability” on page 85.

To restore a lost CUA configuration, do the following:

1. Type R at the Setup Utility menu.

The Setup Utility prompts you for the CA of the system backup as follows:

Restoring System

CA of the system backup to restore :2. Type in the CA from the email message and press Enter. You will

see output similar to the following:

CA of the system backup to restore :5TPRHQKAG1POEe9B5PV9R7DB81HG60VFGBT5LM000000000000000

Restoring system...

Preparing CUA 'xxx.xxx.emc.com' for restoration. Please wait...Preparing CUA for setup ...


84


At the end of the restore process you will see a confirmation message similar to the following:

[02/03/2007 22:39:50] Starting Recovery using CA C6NEELJ4QCK6Ke95GOM4EF3PP26G41222JQE35080Q7H29SBO9F7M [02/03/2007 22:39:50] Recovering a version 0.3 backup [02/03/2007 22:39:50] Hostname for backup is xxx.xxx.emc.com [02/03/2007 22:39:51] Recovering XMD database [02/04/2007 02:21:37] Using starting timestamp of 2007.02.02 23:23:32 GMT from database [02/04/2007 02:21:38] Time based query initiated from cluster time 2007.02.02 23:23:32 GMT [02/04/2007 02:21:38] About to perform SFS proxy query on host 'xxx.xxx.emc.com'... [02/04/2007 02:21:39] SFS proxy query on host 'xxx.xxx.emc.com' completed. status 0 [02/04/2007 02:21:39] Time based query complete, updating file information [02/04/2007 02:21:39] Recovery complete, updating local status information post restore activity.

The CUA restore process reinstates the NFS configuration to /etc/exports and the CIFS configuration to /etc/samba/smb.conf. There is no need to modify these files before restoration.

The CUA also sends events that can be viewed in the GUI Event Log. Refer to “Viewing event messages” on page 220 for more information on the CUA GUI Event Log.



Restoring a CUA in enhanced availability

If you attempt to restore an Active CUA in Enhanced Availability, you see the following error:

Restore cannot be performed while the CUA is configured in an Enhanced Availability environment

To perform the restore for an Active CUA, follow these steps:

1. Remove the Active CUA from the Enhanced Availability configuration:

a. Enter the C) Configuration Options section of the Utility menu and then choose V) Configure Enhanced Availability.

b. Type Y to change the configuration and then type Y to make it the Active CUA. Type N when prompted if there is a Standby CUA.

2. Perform the restore on the system that will be the Active CUA.

Note: To minimize restore time, make sure you use the most recent backup file to restore the system.

3. When the restore is done, reconfigure both the Active and Standby CUAs for Enhanced Availability by following the steps in the Enhanced Availability section of the Monitor and Administration Guide.

Note: Even though the Standby CUA may not have been touched, you must run through the entire EA setup to re-sync the two CUAs.


86


Altering file protectionAlter File Protection allows you to modify the ownership and permissions of a file or directory tree.

Note: The CUA does not know the user or group name ID mappings of the clients using the file system. Therefore, the numeric user or group IDs will normally be used to specify the user and group information other than CUA’s gateway_nfs or gateway_cifs.

To alter file protection, select P at the Setup Utility menu to display the following:


The following is an explanation of each of the above fields:

◆ File — Enter the file or directory name you want to change.

◆ User — Enter the UID for the owner of the file or directory.

◆ Group — Enter the GID for the group in which the file or directory resides.

◆ Permissions — Enter the permissions for the file or directory. The following is an explanation of the Permissions:

• The Hundreds place represents the permission for the owner of the file.

• The Tens place is the permission for group members of the file.

• The Units place is the permission for everyone.

The following figures can be used:

• The number 7 gives read, write, and execute permissions.

• The number 6 gives read and write permissions.

• The number 5 gives read and execute permissions.

• The number 4 gives read permissions.

• The number 0 gives no access.



Examples The following are examples of permissions:

• 775 Owner and group members have read, write, and execute permissions. Everyone else has read and execute permissions.

• 777 Everyone has read, write, and execute permissions.

• 770 Owner and group members have read, write, and execute permissions.

◆ Recurse down directory tree (Y/N) — Use this option to make changes to only the current file or directory or to all files and sub-directories in the current directory.

Altering file protection 87

88


Using access profilesCUA supports the EMC Centera access security model that prevents unauthorized applications from storing data on or retrieving data from an EMC Centera. This security model operates at the application level—not at the end-user level.

Profiles are the means by which authentication and authorization are enforced on the EMC Centera. An EMC Centera authenticates CUA before giving it access with profile-specific permissions. The EMC Centera access security model is based on the concept of pools and profiles.

Note: For the purposes of backward compatibility, CUA still supports the use of PEA files for authorization; however, it is strongly recommended that you use access profiles for CUA authentication. If you do not configure a default access profile, CUA first looks for a PEA file. If there is no PEA file on the CUA, CUA attempts to use the Anonymous Profile to authorize the EMC Centera. The use of the Anonymous Profile is discouraged.

Refer to “Backward compatibility using PEA files” on page 92 for more information about PEA files.

Newly created profiles need to have the following capabilities:

◆ For EMC Centera 2.4 and 3.0:

• Read (r)• Write (w)• Query (q)• Delete (d)• Monitoring (m)

◆ For EMC Centera 3.1

• Read (r)• Write (w)• Delete (d)• Exist (e)• Query (q)• Clip-Copy (c)

In addition, you must enable the Monitoring role when creating the access profile.

CUA lets you configure multiple access profiles and then assign (map) them for use in different circumstances. You can assign a



default access profile for CUA. In addition, you can map other access profiles for backup and restore operations, and for access to directories under gateway_cifs and gateway_nfs. The default access profile is overridden by specific profiles mapped for backup/restore and gateway access.

Note: Access profiles and mappings are stored in two files on CUA. The passwords in these files are encrypted. These files are included in backups and Enhanced Availability configurations.

This section discusses the following topics:

◆ “Configuring CUA access profiles” on page 89

◆ “Mapping access profiles” on page 90

◆ “Backward compatibility using PEA files” on page 92

Configuring CUA access profiles

Access profiles must exist on the EMC Centera before you can configure them in CUA. Refer to EMC Centera Online Help for information on creating access profiles on the EMC Centera.

To configure access profiles in CUA, do the following:


2. Select G at the CUA Configuration menu. CUA displays the following Access Profiles Configuration menu.

Access Profile Configuration----------------------------A) Add/modify access profilesR) Remove access profilesM) Map access profilesS) Show access profiles

Q) Quit WITHOUT saving changesE) Save changes and exit

3. Select A. CUA displays the following prompts:

Access User:Access Password:Enter password again:

Using access profiles 89

90


4. Enter the Access Username and Access Password and then confirm the password. CUA displays the following prompt:

Is this a cluster profile? (Y/N)

Note: It is strongly recommended that you create a cluster profile for restores. Do not map cluster profile to any specific use by CUA. It will be used automatically by restore.

5. Enter Y or N depending on the kind of access profile you added.

CUA verifies the validity of this access profile with EMC Centera. If this is not a valid profile, you receive the following error:

Error validating the access profile for user 'xxxxx' (and/or the password).

6. Save changes and exit.

Viewing accessprofiles

To view access profiles:

At the Access Profile menu, select S. CUA displays a list of all configured access profiles.

Mapping access profiles

Mapping access profiles assigns them for use in specific situations. You can map access profiles for the following uses:

◆ Default access profile — This profile is used for all file system access unless overridden by a specific backup/restore or gateway access profile.

Note: It is strongly recommended that you map a default access profile. If you do not configure a default access profile, CUA first looks for a PEA file. If there is no PEA file on the CUA, CUA attempts to use the Anonymous Profile to authorize with the EMC Centera. On fresh installs of CentraStar v3.1, the Anonymous Profile is disabled by default. Use of the Anonymous Profile is strongly discouraged.



◆ Backup/restore access profile — This profile is used to place nightly backup files and EA activity logs on the EMC Centera and to read backups off the EMC Centera. If you do not configure a backup/restore access profile, CUA tries to use the anonymous profile for backup and restore. Use of the Anonymous Profile is strongly discouraged.

Note: You must configure this access profile to perform backups and restores. In addition to the mapped profile for backup/restores, it is also strongly recommended that you create a cluster profile for restore purposes. Performance on restores will be significantly better if you have added a cluster profile to the CUA.

◆ Access profiles for gateways—these profiles are used for access to /sfs/gateway_cifs and /sfs/gateway_nfs. You can also set distinct access profiles for subdirectories one level under gateway_cifs and gateway_nfs. If you do not set a specific access profile for a subdirectory, it defaults to the parent directory’s access profile. For example, you could set the following access profiles:

/sfs/gateway_cifs User: cifs_main/sfs/gateway_cifs/test User: cifs_sub

Any subdirectory under gateway_cifs, other than /sfs/gateway_cifs/test, would use cifs_main for its access profile.

To map an accessprofile

To map an access profile, do the following:

1. Select G at the CUA Configuration menu. CUA displays the following Access Profiles Configuration menu.

2. Select M from the Access Profiles Configuration menu. The following menu appears.

Map Access Profiles Configuration---------------------------------A) Set default access profileB) Set access profiles for backups/restoresC) Set access profiles for gateway_cifsN) Set access profiles for gateway_nfsD) Remove access profiles from gateway_cifsR) Remove access profiles from gateway_nfsS) Show Access Profiles Mappings



92


3. From this menu you can map the following access profiles:

• To set the default access profile, enter the Access Username.

• To set the access profile for backups/restores, enter the Access Username.

• To set access profiles for gateway_cifs, enter the path for the subdirectory and then the Access Username.

• To set access profiles for gateway_nfs, enter the path for the subdirectory and then the Access Username.


Note: If this installation is an Enhanced Availability setup, this procedure must be carried out on both the Active and Standby CUAs.

Search order foraccess profiles

When authorizing access, the CUA looks for these elements in the following order:

1. Access profile for gateways or backup

2. Default access profile

3. PEA file

4. Anonymous profile

Backward compatibility using PEA filesCUA 3.6 and earlier used a Pool Entry Authorization (PEA) files to provide a authentication on the EMC Centera. The PEA file is passed to the SDK during a pool open and allows the CUA to have a profile associated with it.

For the purposes of backward compatibility, CUA still supports the use of PEA files for authorization. However, if you map a default access profile, this access profile is used in place of the PEA file. Refer to “Using access profiles” on page 88 for more information about access profiles.

About PEA files All EMC Centera pool open calls on the CUA will look for the PEA file named: /usr/local/PAI/PAI_info.

Note: The name of the file is xxx.pea on EMC Centera but must be named PAI_info on the CUA.



If this file exists it will be passed to the FPPool_Open SDK call. Both the users service and gwsetup have rights to create the /usr/local/PAI/PAI_info file.

Refer to the CLI command profile update in the EMC Centera System Operators Guide, P/N 069001126 for information on creating the Pool Entry Authorization file (xxx.PEA file). The PEA file must be on your local system before performing the following procedure. The following procedure copies the file on to the CUA and renames the file.

Note: If this installation is an Enhanced Availability setup, this procedure must be carried out on both the Active and Standby CUAs.

Copying PEA files To copy the PEA file from your local system onto the CUA, do the following:

1. Enable FTP access on the CUA by following the procedure in “Configuring FTP access” on page 78.

Note: Ensure that you select Y at the prompt Enable ftp access to account gwsetup? (Y/N).

2. FTP to the CUA by typing the following command at a command prompt:

ftp xxx.xx.x.xx

Note: Replace xxx.xx.x.xx with the IP address of the CUA.

3. Log in using gwsetup and the password EMCCUA.

4. Type the following command to set the transfer type to binary mode:

bin

5. Type the following command to transfer the PEA file to the CUA

put xxx.PEA file

Note: Replace xxx.PEA file with the actual name of the .pea file.

6. Exit FTP by typing the following command:

bye


94



Note: Ensure that you are logged in to the CUA as gwsetup.

8. Move the PEA file to the correct directory, by typing the following command:

mv /var/tmp/ftp/xxx.PEA /usr/local/PAI/PAI_info

Note: In the step above, the file name on the CUA must be PAI_info and the file must be copied to the correct directory. No other file name or directory will work.

Note: The Unix command mv is equivalent to the Windows command rename.

9. Select T at the Setup Utility menu to reboot the CUA.



Working with metadataThis section describes how to add custom metadata to files written to the EMC Centera and how to browse metadata.

Using custom metadata

CUA allows you to add custom metadata to the CDFs of files written back to the EMC Centera. The custom metadata, specified by a special file called custom_metadata.txt, is added on a per directory basis. The use of custom metadata must be enabled on CUA.

You can control the custom metadata for each directory by having multiple custom_metadata.txt files. During the writeback process for a file, CUA searches the directory tree of the file from bottom to top. The custom metadata contained in the first occurrence of a custom_metadata.txt file is written to the file’s CDF.

Example You have the following directory tree:

./dirA

./dirA/file1

./dirA/custom_metadata.txt

./dirA/dirB

./dirA/dirB/custom_metadata.txt

./dirA/dirB/dirC

./dirA/dirB/dirC/dirD

./dirA/dirB/dirC/dirD/file1

When writing file ./dirA/file1 to EMC Centera, CUA uses the custom metadata found in ./dirA/custom_metadata.txt.

Using custommetadata

To enable custom metadata, do the following:

Note: If you do not use custom metadata, do not enable this option— doing so could impact CUA’s performance.


2. Select M at the CUA Configuration menu. CUA displays the following prompt:

Do you want to allow the end user to provide custom metadata in custom_metadata.txt files (Y/N): N

3. Replace the N with Y and press Enter.

4. Press any key to continue.

Working with metadata 95

96


5. Create custom_metadata.txt files using the format described in “Format of the Custom Metadata File” on page 96.

6. Place the custom_metadata.txt files in the appropriate directories on the CUA server.

Format of the Custom Metadata FileThe format of the content of custom_metadata.txt file is:

Tag = value

Example The contents of the custom_metadata.txt file look like this:

Patient = Joe SmithID = 1234

Note: Custom metadata tag names are case insensitive and must be XML (Extensible Markup Language) compliant.

The maximum number of 'name = value' pairs allowed in any custom_metadat.txt file is 100.

When written to the CDF, the custom metadata looks like this:

<CUA_Custom_Metadata Patient="Joe Smith" ID="1234"/>

During the writeback process, if the content of the custom metadata is not in the correct format, the writeback continues excluding invalid custom metadata content and a message is logged.

Browsing metadata To see a list of files on a Standby CUA you can browse the metadata (you cannot use the ls or dir commands on a Standby CUA).

To Browse metadata select B at the CUA Setup Utility menu.



Testing email alertsTo test the email alert notification system:

1. Select M at the Setup Utility menu. The following prompt appears:

Send E-mail Alert

E-mail recipient (NULL to cancel) :

2. Enter a valid email address for the intended recipient of the message (if sending to EMC use the following address: [email protected]). Separate multiple entries with a space, comma, or semi-colon (to cancel the operation, press Enter without entering an email address).

Note: You cannot use # in front of an email address when setting alert email addresses through the configuration menu.

3. You will then be prompted for the subject and content of the message as shown below.

E-mail subject:Message text (NULL to cancel) :

4. Type a subject and the content of the message and press Enter (to cancel the operation, press Enter without entering a subject). The following output appears:

Message sent.Press any key to continue.

5. Verify that the email alert was processed correctly by:

• Calling the EMC Customer Service - Help Desk to verify that the email alert was processed by the EMC CenteraAlert workstation or

• Calling the customer’s administrator if the CUA was configured with the local SMTP server and administrator’s email address.

Testing email alerts 97

98


Entering permanent licensing informationThe CUA tracks the expiration date of a temporary license key and generates an event message or sends an email alert (if configured) 21 days before the temporary license expires and every day thereafter.

Additionally, an event (information) is logged daily in the event viewer specifying how much time remains on the temporary key.

Note: Upgrading from CUA v3.6 to v4.0 or higher requires a new license.

Note: The license key is the 15-character Serial Number found on the part number label on the outside of the CUA kit.

License key verification

To determine whether the CUA license key is a temporary key or a permanent key, log into the CUA as gwsetup using the password EMCCUA and select X at the CUA Setup Utility menu to enter the Bash Shell. Type the following command:

cat /proc/sys/storigen/sfs/sfs_enabled

The system will return a numeric value. The values are:

0 = invalid key1 = valid permanent key2 = valid temporary key

Entering a permanent license

To enter a permanent license, do the following:


2. Select L at the CUA Configuration menu to enter the license details. The following information is displayed:

Enter License-------------Your system currently has a temporary CUA license with 79 day(s) remaining.

Do you want to enter the license key? (Y/N): YEnter CUA license (15 characters) : xxxxxxxxxxxxxxx

You've entered the key: xxxxxxxxxxxxxxxDo you want to continue? (Y/N) : Y



A valid license key has been entered.

Press any key to continue

3. Type Y, input the license number and press Enter on your keyboard.

4. Type Y again to continue and then press any key to finish the procedure.

Note: If you already have a permanent license and you select this option, the following message is displayed:

A valid SFS key already exists and may not be changed.Press any key to continue

Note: If the temporary license expires before you install the permanent license, you will have to reboot the CUA after entering the license key.

Entering permanent licensing information 99

100


Setting retention periodsRetention periods are configured on directories one level under /sfs/gateway_nfs or /sfs/gateway_cifs. All files and directories under the specified directory inherit the configured retention period. You can set the retention period to a valid EMC Centera retention class or a specific number of days. You can also specify an infinite retention where the file can never be deleted.

Note: Retention classes can be created through the EMC Centera CLI using the create retention command. Retention classes used on CUA must start with an alphabetic character.

Retention example The following example shows how to set a retention period on a NFS directory. Retention periods can only be set on subdirectories one level under /sfs/gateway_nfs and /sfs/gateway_cifs.

This example sets a retention period of one year for the directory /sfs/gateway_nfs/directoryA.


2. Select R at the CUA Configuration menu to display the following Data Retention menu.

Data Retention Configuration----------------------------C) Insert/modify retention into gateway_cifsN) Insert/modify retention into gateway_nfsD) Remove retention from gateway_cifsR) Remove retention from gateway_nfsS) Show retention configuration



• If you are setting the retention period for NFS, select N at the Data Retention menu.

• If you are setting the retention period for CIFS, select C at the Data Retention menu.

The system prompts you to enter the directory name for which you are setting the retention period.



4. Enter the name of the directory and press Enter. The system prompts for the retention period.


• If you want to set the period to a retention class, type the class name and press Enter.

Example: Retention time (in days), or retention class: class123

Note: The CUA only supports retention classes that begin with an alphabetic character.

• If you want to set the period for a regular interval, type the number of days in the format #_daysd and press Enter.

Example: Retention time (in days), or retention class: 55

Note: There is no need to reboot or restart any program after modifying retention periods.

6. Select S at the Data Retention Configuration menu to verify that the retention period has been set correctly. The output will look similar to the following:

Current Retention Configuration-------------------------------System default retention period :0

/sfs/gateway_nfs/directoryA: 365d


All new files created under the hierarchy of /sfs/gateway_nfs/directoryA will have a retention period of 1 year.

Retention notes ◆ When configuring the retention settings, you can only set different retention settings on directories one level below /sfs/gateway_cifs and /sfs/gateway_nfs. For example /sfs/gateway_nfs/directoryA can have a retention setting of 1 year and /sfs/gateway_nfs/directoryB can have a retention setting of 2 years. The settings are inherited for any of their subdirectories.

Setting retention periods 101

102


◆ Valid retention periods are set in days or by a retention class. A value of infinite represents an infinite retention period.

If /sfs/gatway_nfs/directoryA has a retention of one year, all files under this directory inherit the one year retetention.

Example /sfs/gatway_nfs/directoryA/accounting/file1.dat will inherit the one year retention period.

◆ EMC does not recommend setting the retention period after files are already written to the directory. Files created prior to setting the retention period have a retention period of 0. Only new files will receive the new retention period. A better practice is to set the retention period and then start writing files to the directory.

◆ EMC does not recommend modifying a retention period for a directory. If you do modify a retention period on a directory, all files that have already been written out to EMC Centera will not get the new retention period (you cannot change the retention period on a file). Any new files written to the directory will get the new retention period. A better practice is to create a new directory with the new retention period.

◆ Once a file has a retention period there is no way to deactivate or modify the retention period.

◆ Retention periods are not set on files until 30 minutes after the last file modification.



Changing CUA passwordsThe table below lists the default CUA login and password details.

To change account passwords:

1. Select C at the CUA Setup Utility menu.

2. Select P at the CUA Configuration menu.

3. Either fill in the new passwords at the prompts or repeatedly press Enter to move through the prompts without making changes.

Note: If you changed the default passwords for any account and then upgraded to a later version of CUA, passwords are set to the default values for the new version.

Table 10 CUA access details

MAP Login Password

CUA AdminGUI AdminFTP Admin

gwsetupsysmongwsetup

EMCCUAEMCCUAEMCCUA

FTP NFS gateway_nfs gateway_nfs

FTP CIFS gateway_cifs gateway_cifs

Changing CUA passwords 103

104


Setting system timeThe System Time page is used to configure the server system time.

System Time is changed through the CUA GUI or CUA Setup Utility.

Using the CUA GUI Change the System Time using the CUA GUI, by doing the following:

1. Navigate to the following the path:

Configure ⇒ System Time

Figure 9 Set system time

2. On the System Time page, enter the new system date and time and click on the Apply button.

The system time is changed.



Using the CUA Setup Utility

To change the System Time using the CUA Setup Utility, do the following:



sudo date [MMDDhhmm]

Note: Replace MMDDhhmm with:MM — Month (example: 01 for January)DD — Day (example: 25)hh — Hour (example: 22)mm — Minute (example: 33)

The system will be changed.

To see more System Time options type the following command:

date --help

Setting system time zone

The default time zone on the CUA is set to US Eastern Standard Time.

To change the time zone, do the following:



sudo /usr/sbin/timeconfig

3. The following menu appears:

Please identify a location so that time zone rules can be set correctly.Please select a continent or ocean. 1) Africa 2) Americas 3) Antarctica 4) Arctic Ocean 5) Asia 6) Atlantic Ocean 7) Australia 8) Europe 9) Indian Ocean10) Pacific Ocean11) none - I want to specify the time zone using the Posix TZ format.

Setting system time 105

106


4. Enter the number for the region in which your time zone is located. You are prompted to select a country.

5. Select a country. You are now prompted to select a time zone.

6. Select a time zone. You are asked to confirm the time zone setting.

7. Select Yes to confirm the time zone setting.



Setting up SNMPCUA supports SNMPv2 data collection and specifically MIB-2. T.

SNMP is set through the CUA GUI. To reach the CUA GUI, do the following:

1. Open a standard web browser.

2. Navigate to the following URL, which uses the HTTPS (HTTP Secure) protocol:

https://machine:7227/

Replace machine with the host name or IP address of CUA.

Example The following connects to a CUA named demo1:

https://demo1:7227/

3. Log in as gwsetup using the password EMCCUA.

Once you are connected to the CUA GUI, navigate to the Configure SNMP Parameters page through the following path:

Configure ⇒ SNMP

Figure 10 Configure SNMP

Setting up SNMP 107

108


Once at this page, do the following:

1. Fill in the required MIB-2 system-description objects.

2. Enable access to the SNMP agent on this node.

3. Set destination(s) for the node’s SNMP traps.

Click Apply to change the settings.

Following sections provide a description of each panel on the SNMP page:

System information Use the System Information panel to set three system-description objects from the MIB-2 system table, and to optionally set a trap for authentication failure. Table 11 on page 108 describes the elements on the System Information Panel.

Table 11 System Information panel elements

Screen Element Description

System Description Maps to the MIB-2 object, system.sysDescr (mib-2 1 1 in the MIB tree). This is a read-only description of CUA.

System Location Maps to the MIB-2 object, system.sysLocation (mib-2 1 6 in the MIB tree). Enter up to 255 characters to describe the location of the CUA.

System Contact Maps to the MIB-2 object, system.sysContact (mib-2 1 4 in the MIB tree). Enter the name of someone to contact regarding the CUA, along with a method for contact. Use up to 255 characters.

System Name Maps to the MIB-2 object, system.sysName (mib-2 1 5 in the MIB tree). This is a name for the CAG.By convention, use the fully-qualified domain name (FQDN); for example, myhost.mycompany.com.

Enable Auth Trap Option to enable an authentication trap, which is a trap message indicating that some user tried to access the SNMP agent with an incorrect community string.Set the community string with the Community Name field in the Agent Access panel.



Agent access Use the Agent Access panel to allow access to the SNMP agent on the CUA. Table 12 on page 109 describes the elements on the Agent Access panel.

Table 12 Agent Access panel elements


Community Name Community string to access this SNMP agent. Enter up to 255 characters.

Restricted Host1 Address IP address of a host with access to the SNMP agent. Enter the IP address in dotted-decimal format; for example, 192.168.50.11.

Note: You cannot use host names in this field.

Netmask Network mask for the above IP host. Choose a mask from the pull-down menu.

SNMP Version Version of SNMP to use with the host. Choose v1 or v2c from the pull-down menu.

Restricted Host2 Address Optional second IP host to which you can grant access. This host has its own Netmask and SNMP Version settings.


Setting up SNMP 109

110


SNMP trap destinations

Use the SNMP Trap Destinations panel to identify one or two hosts to receive SNMP trap messages from this node. Table 13 on page 110 describes the elements on the SNMP Trap Destinations panel.

Table 14 on page 110 lists the trap events generated by SNMP.

Table 13 SNMP Trap Destination panel elements


Community Name Community string to access this SNMP agent. Enter up to 255 characters.

Host1 Address IP address of a host with access to the SNMP agent. Enter the IP address in dotted-decimal format; for example, 192.168.50.11.


Netmask Network mask for the above IP host. Choose a mask from the pull-down menu.

SNMP Version Version of SNMP to use with the host. Choose v1 or v2c from the pull-down menu.

Host2 Address Optional second IP host to which you can grant access. This host has its own Netmask and SNMP version settings.


Table 14 SNMP trap events (1 of 2)

Object ID Description

1.3.6.1.4.1.7979.3.1.1.0.2 CPU usage %1 exceeds threshold %2

1.3.6.1.4.1.7979.3.1.1.0.4 File system %1 usage %2 exceeds threshold %3

1.3.6.1.4.1.7979.3.1.2.0.2 Monitored Network Interface %1 is down

1.3.6.1.4.1.7979.3.1.2.0.2 %1 out of %2 raid devices are inactive

1.3.6.1.4.1.7979.3.1.2.0.2 Service %1 is not running

1.3.6.1.4.1.7979.3.1.2.0.2 EMC CUA temporary key %1 expired %2 days ago on %3

1.3.6.1.4.1.7979.3.1.2.0.1 EMC CUA temporary key %1 expires in %2 days on %3



Email alerts You can have the system send email alerts to one or more destinations if an SNMP trap message is received. To enable this service you must check the Enable Traps check box on the Configure Fault Detector page. Table 15 on page 111 describes the elements on the Agent Access panel.

Note: To have email alerts sent to EMC as well as to a system administrator, you must configure an SMTP server that can send email externally. One of the addresses should be [email protected].

1.3.6.1.4.1.7979.3.1.2.0.2 Gateway Backup failed at %2 for %1

1.3.6.1.4.1.7979.3.1.2.0.2 Active CUA (%1) is in an [%2] state; pid %3

1.3.6.1.4.1.7979.3.1.2.0.2 Active CUA (%1) is in an [%2] state; pid %3

1.3.6.1.4.1.7979.3.1.2.0.2 Standby CUA (%1) is in an [%2] state

1.3.6.1.4.1.7979.3.1.2.0.2 Standby CUA (%1) failover for Active CUA (%2)

1.3.6.1.4.1.7979.3.1.2.0.1 Disk %1 (mount point %3) has been removed

1.3.6.1.4.1.7979.3.1.2.0.1 Array md%1 disk failure on %2; removing %2 from array

1.3.6.1.4.1.7979.3.1.2.0.2 Error processing activity log (%1) on CUA (%2); pid %3

Table 14 SNMP trap events (2 of 2)

Table 15 Email Alert panel elements


SMTP Server IP address or the hostname of the mail server.

Email Destinations Email addresses to which error alerts and SNMP trap messages will be sent. Separate multiple addresses with a space character. One email message per error alert or SNMP trap will be sent.

Setting up SNMP 111

112


Configuring UTF-8 file name supportThe CUA Setup Utility supports UTF-8 file names. To display UTF-8 file names correctly, you must configure your locale in the Setup Utility and log into the CUA from a system that has the corresponding fonts.


To set a locale, do the following:

1. Select O at the Setup Utility menu. The following prompt appears:

Select the editor to use for future cli commandsvi - standard unix editorpico - easy to use editor

Editor :pico

2. Press Enter. The following prompt appears:

To view the locale choices, enter "locale -a" at the bash promptSelect locale :

3. Type the appropriate locale name from Table 16 on page 112 and press Enter.

Table 16 Locales supported by CUA (1 of 10)

Locale name Description

aa_DJ.utf8 Afar language locale for Djibouti (CaduLaaqo Dialects).

aa_ER Afar language locale for Eritrea (CaduLaaqo Dialects).

aa_ER@saaho Afar language locale for Eritrea (Saaho Dialect).

aa_ER.utf8 Afar language locale for Eritrea (CaduLaaqo Dialects).

aa_ET Afar language locale for Ethiopia (CaduCarra Dialects).

aa_ET.utf8 Afar language locale for Ethiopia (CaduCarra Dialects).

af_ZA.utf8 Afrikaans locale for South Africa

am_ET Amharic language locale for Ethiopia.



am_ET.utf8 Amharic language locale for Ethiopia.

an_ES.utf8 Aragonese locale for Spain

ar_AE.utf8 Arabic language locale for United Arab Emirates

ar_BH.utf8 Arabic language locale for Bahrain

ar_DZ.utf8 Arabic language locale for Algeria

ar_EG.utf8 Arabic language locale for Egypt

ar_IN Arabic language locale for India

ar_IN.utf8 Arabic language locale for India

ar_IQ.utf8 Arabic language locale for Iraq

ar_JO.utf8 Arabic language locale for Jordan

ar_KW.utf8 Arabic language locale for Kuwait

ar_LB.utf8 Arabic language locale for Lebanon

ar_LY.utf8 Arabic language locale for Libyan Arab Jamahiriya

ar_MA.utf8 Arabic language locale for Morocco

ar_OM.utf8 Arabic language locale for Oman

ar_QA.utf8 Arabic language locale for Qatar

ar_SA.utf8 Arabic locale for Saudi Arabia

ar_SD.utf8 Arabic language locale for Sudan

ar_SY.utf8 Arabic language locale for Syrian Arab Republic

ar_TN.utf8 Arabic language locale for Tunisia

ar_YE.utf8 Arabic language locale for Yemen

az_AZ.utf8 Azeri language locale for Azerbaijan (latin)

be_BY@latin Belarusian Latin-Script locale for Belarus

be_BY.utf8 Belarusian locale for Belarus

bg_BG.utf8 Bulgarian locale for Bulgaria



Configuring UTF-8 file name support 113

114


bn_BD BengaliBangla language locale for Bangladesh

bn_BD.utf8 BengaliBangla language locale for Bangladesh

bn_IN Bengali language locale for India

bn_IN.utf8 Bengali language locale for India

br_FR.utf8 Breton language locale for France

bs_BA.utf8 Bosnian language locale for Bosnia and Herzegowina

byn_ER ‘Blin language locale for Eritrea

byn_ER.utf8 Blin language locale for Eritrea

ca_ES.utf8 Catalan locale for Spain

cs_CZ.utf8 Czech locale for the Czech Republic

cy_GB.utf8 Welsh language locale for Great Britain

da_DK.utf8 Danish locale for Denmark

de_AT.utf8 German locale for Austria

de_BE.utf8 German locale for Belgium

de_CH.utf8 German locale for Switzerland

de_DE.utf8 German locale for Germany

de_LU.utf8 German locale for Luxemburg

dz_BT Dzongkha language locale for Bhutan

el_GR.utf8 Greek locale for Greece

en_AU.utf8 English locale for Australia

en_BE.utf8 IBM LOCALE SOURCE FOR GLIBC 2.2

en_BW.utf8 English locale for Botswana

en_CA.utf8 English locale for Canada

en_DK.utf8 English locale for Denmark

en_GB.utf8 English locale for Britain





en_HK.utf8 English locale for Hong Kong

en_IE.utf8 English locale for Ireland

en_IN English language locale for India

en_IN.utf8 English language locale for India

en_NZ.utf8 English locale for New Zealand

en_PH.utf8 English language locale for Philippines

en_SG.utf8 English language locale for Singapore

en_US.utf8 English locale for the USA

en_ZA.utf8 English locale for South Africa

en_ZW.utf8 English locale for Zimbabwe

es_AR.utf8 Spanish locale for Argentina

es_BO.utf8 Spanish locale for Bolivia

es_CL.utf8 Spanish locale for Chile

es_CO.utf8 Spanish locale for Colombia

es_CR.utf8 Spanish locale for Costa Rica

es_DO.utf8 Spanish locale for Dominican Republic

es_EC.utf8 Spanish locale for Ecuador

es_ES.utf8 Spanish locale for Spain

es_GT.utf8 Spanish locale for Guatemala

es_HN.utf8 Spanish locale for Honduras

es_MX.utf8 Spanish locale for Mexico

es_NI.utf8 Spanish locale for Nicaragua

es_PA.utf8 Spanish locale for Panama

es_PE.utf8 Spanish locale for Peru

es_PR.utf8 Spanish locale for Puerto Rico




116


es_PY.utf8 Spanish locale for Paraguay

es_SV.utf8 Spanish locale for El Salvador

es_US.utf8 Spanish locale for the USA

es_UY.utf8 Spanish locale for Uruguay

es_VE.utf8 Spanish locale for Venezuela

et_EE.utf8 Estonian locale for Estonia

eu_ES.utf8 Basque locale for Spain

fa_IR Persian locale for Iran

fa_IR.utf8 Persian locale for Iran

fi_FI.utf8 Finnish locale for Finland

fo_FO.utf8 Faroese locale for Faroe Islands

fr_BE.utf8 French locale for Belgium

fr_CA.utf8 French locale for Canada

fr_CH.utf8 French locale for Switzerland

fr_FR.utf8 French locale for France

fr_LU.utf8 French locale for Luxemburg

ga_IE.utf8 Irish locale for Ireland

gd_GB.utf8 Scots Gaelic language locale for Great Britain

gez_ER Ge'ez language locale for Eritrea.

gez_ER@abegede Ge'ez language locale for Eritrea With Abegede Collation.

gez_ET Ge'ez language locale for Ethiopia

gez_ET@abegede Ge'ez language locale for Ethiopia With Abegede Collation

gl_ES.utf8 Galician locale for Spain

gu_IN Gujarati Language Locale For India

gv_GB.utf8 Manx Gaelic locale for Britain





he_IL.utf8 Hebrew locale for Israel

hi_IN Hindi language locale for India

hi_IN.utf8 Hindi language locale for India

hr_HR.utf8 Croatian locale for Croatia

hsb_DE.utf8 Upper Sorbian locale for Germany

hu_HU.utf8 Hungarian locale for Hungary

id_ID.utf8 Indonesian locale for Indonesia

is_IS.utf8 Icelandic locale for Iceland

it_CH.utf8 Italian locale for Switzerland

it_IT.utf8 Italian locale for Italy

iw_IL.utf8 Hebrew locale for Israel

ja_JP.utf8 Japanese language locale for Japan

ka_GE.utf8 Georgian language locale for Georgia

kk_KZ.utf8 Kazakh locale for Kazakhstan

kl_GL.utf8 Greenlandic locale for Greenland

km_KH Khmer locale for Cambodia

kn_IN Kannada language locale for India

ko_KR.utf8 Korean locale for Republic of Korea

ku_TR.utf8 Kurdish (latin) locale for Turkey

kw_GB.utf8 Cornish locale for Britain

ky_KG Kyrgyz Language Locale for Kyrgyzstan

lg_UG.utf8 Luganda locale for Uganda

lo_LA Lao locale for Laos

lt_LT.utf8 Lithuanian locale for Lithuania

lv_LV.utf8 Latvian locale for Latvia




118


mg_MG.utf8 Malagasy locale for Madagascar

mi_NZ.utf8 Maori language locale for New Zealand

mk_MK.utf8 Macedonian locale for Macedonia

ml_IN Malayalam language locale for India

ml_IN.utf8 Malayalam language locale for India

mn_MN Mongolian locale for Mongolia

mn_MN.utf8 Mongolian locale for Mongolia

mr_IN Marathi language locale for India

mr_IN.utf8 Marathi language locale for India

ms_MY.utf8 Malay language locale for Malaysia

mt_MT.utf8 Maltese language locale for Malta

nb_NO.utf8 Norwegian (Bokmal) locale for Norway

ne_NP Nepali language locale for Nepal

ne_NP.utf8 Nepali language locale for Nepal

nl_BE.utf8 Dutch locale for Belgium

nl_NL.utf8 Dutch locale for the Netherlands

nn_NO.utf8 Nynorsk language locale for Norway

no_NO.utf8 Norwegian (Bokmal) locale for Norway

nso_ZA Northern Sotho locale for South Africa

oc_FR.utf8 Occitan Language Locale for France

om_ET Oromo language locale for Ethiopia.

om_ET.utf8 Oromo language locale for Ethiopia.

om_KE.utf8 Oromo language locale for Kenya.

pa_IN Punjabi language locale for Indian Punjabi(Gurmukhi)

pa_IN.utf8 Punjabi language locale for Indian Punjabi(Gurmukhi)





pl_PL.utf8 Polish locale for Poland

pt_BR.utf8 Portuguese locale for Brasil

pt_PT.utf8 Portuguese locale for Portugal

ro_RO.utf8 Romanian locale for Romania

ru_RU.utf8 Russian locale for Russia

ru_UA.utf8 Russian locale for Ukraine

rw_RW Kinyarwanda language locale for Rwanda

se_NO Northern Saami language locale for Norway

se_NO.utf8 Northern Saami language locale for Norway

sh_YU.utf8 IBM LOCALE SOURCE FOR GLIBC 2.2

sid_ET Sidama language locale for Ethiopia.

sid_ET.utf8 Sidama language locale for Ethiopia.

si_LK Sinhala language locale for Sri Lanka

sk_SK.utf8 Slovak locale for Slovak

sl_SI.utf8 Slovenian locale for Slovenia

so_DJ.utf8 Somali language locale for Djibouti.

so_ET Somali language locale for Ethiopia

so_ET.utf8 Somali language locale for Ethiopia

so_KE.utf8 Somali language locale for Kenya

so_SO.utf8 Somali language locale for Somalia

sq_AL.utf8 Albanian language locale for Albania

sr_CS.utf8 Serbian locale for Serbia and Montenegro

ss_ZA Swati locale for South Africa

st_ZA.utf8 Sotho locale for South Africa

sv_FI.utf8 Swedish locale for Finland




120


sv_SE.utf8 Swedish locale for Sweden

ta_IN Tamil language locale for India

ta_IN.utf8 Tamil language locale for India

te_IN Telugu language locale for India

te_IN.utf8 Telugu language locale for India

tg_TJ.utf8 Tajik language locale for Tajikistan

th_TH.utf8 Thai locale for Thailand

ti_ER Tigrigna language locale for Eritrea.

ti_ER.utf8 Tigrigna language locale for Eritrea.

ti_ET Tigrigna language locale for Ethiopia.

ti_ET.utf8 Tigrigna language locale for Ethiopia.

tig_ER Tigre language locale for Eritrea

tig_ER.utf8 Tigre language locale for Eritrea

tl_PH.utf8 Tagalog language locale for Philippines

tn_ZA Tswana locale for South Africa

tr_TR.utf8 Turkish locale for Turkey

ts_ZA Tsonga locale for South Africa

tt_RU.utf8 Tatar language locale for Tatarstan

uk_UA.utf8 Ukrainian locale for Ukraine

ur_PK Urdu Language Locale for Pakistan

ur_PK.utf8 Urdu Language Locale for Pakistan

uz_UZ@cyrillic Uzbek (cyrillic) locale for Uzbekistan

ve_ZA Venda locale for South Africa

vi_VN Vietnamese language locale for Vietnam

wa_BE.utf8 Walloon Language Locale for Belgium





xh_ZA.utf8 Xhosa locale for South Africa

yi_US.utf8 Yiddish Language locale for the USA

zh_CN.utf8 Chinese locale for Peoples Republic of China

zh_HK.utf8 Chinese language locale for Hong Kong

zh_SG.utf8 Chinese language locale for Singapore

zh_TW.utf8 Chinese locale for Taiwan

zu_ZA.utf8 Zulu locale for South Africa




122



4

This chapter explains how to enable, configure, and use the EMC Centera FileArchiver.


◆ Introduction ...................................................................................... 124◆ Enabling EMC Centera FileArchiver............................................. 126◆ Configuring EMC Centera FileArchiver....................................... 134◆ Defining migration policies ............................................................ 138◆ Scheduling a policy run .................................................................. 147◆ Manually starting and stopping a policy run .............................. 149◆ Simulating a policy run................................................................... 150◆ Monitoring EMC Centera FileArchiver ........................................ 152◆ Managing orphans........................................................................... 155

EMC CenteraFileArchiver

EMC Centera FileArchiver 123

124

EMC Centera FileArchiver

IntroductionEMC Centera FileArchiver (CFA) is a policy engine appliance that moves data between the Celerra and EMC Centera platforms. CFA uses the Celerra FileMover API in order to provide complete transparency to end-users mounting both CIFS and NFS Celerra file systems.

Archived files on Celerra file systems are converted to stub files which contain an abundance of metadata including but not limited to the remote location of the actual file on the secondary storage platform.

Figure 11 EMC Centera FileArchiver

Policies To use CFA, users define file systems on the Celerra, develop policies, and schedule data movement. CFA will scan the appropriate file systems and move data from the Celerra directly to an EMC Centera. Data is never cached on the CFA system during migrates; however, data is cached on CFA when it is read back from the EMC Centera. Only when the data has successfully been written to the EMC Centera will the file be stubbed on the Celerra. This eliminates any concern over data loss during migration.

FileArchiver Node

NFS, HTTP, TCP/IP FileMover



Users can develop policies based on accessed or modified time, file size, extension, file names, or directory location. They can also use logical operators such as and, or, and not to create more complex policies. For instance, the user can configure CFA to migrate all PDF files on a Celerra that have not been modified in one month and are at least 10 MB in size.

Refer to “Defining migration policies” on page 138 for more information on configuring policies.

Retention Administrators can attach EMC Centera retention periods to specific policies ensuring that files may not be deleted from EMC Centera until the retention period expires.

Note: Retention classes are not supported on CFA.

Refer to “Defining migration policies” on page 138 for more information on configuring retention.

Simulation Another feature of CFA is the simulation mode which allows administrators to develop polices and run them without actually moving data. Logs show how much data would have been moved if actual migration had occurred.

Refer to “Simulating a policy run” on page 150 for more information on simulating a policy run.

Introduction 125

126


Enabling EMC Centera FileArchiverEMC Centera FileArchiver (CFA) is built on top of core CUA technology and runs on all supported CUA hardware platforms. CFA and CUA cannot run simultaneously. Once you enable CFA the general purpose CIFS and NFS exports that CUA provides will be disabled.

For further information on Celerra FileMover (called DHSM in previous versions) refer to the following Celerra Network Server documentation:

◆ Celerra Network Server v5.3 — Configuring DHSM on Celerra

◆ Celerra Network Server v5.4 and later — Using Celerra FileMover

Pre-Requisites◆ The Celerra and EMC Centera must be located within the same

Local Area Network.

◆ The minimum Celerra version for CFA is v5.4.18.3 for Celerra v5.4 and v5.5.20.1 for Celerra v5.5.

IMPORTANT!If you are using HTTP stubbing with CFA, your Celerra must be patched to at least version 5.5.27-5. For more information on HTTP stubbing, refer to “Phase 6 - Establish connection from the Celerra Network Server” on page 131.

◆ You must have a valid Celerra NFS license to use CFA.



CFA configurationchecklist

This checklist provides a brief list of the tasks needed to configure CFA and the Celerra Network Server. Specific steps are described in more detail in the following sections.

❑ Configure the CFA by following the procedure in “Phase 1 - Enabling CFA” on page 127.

❑ Log onto the Celerra Control Station as user nasadmin.

❑ Setup UTF-8 filename translation by following this procedure: “Phase 2 - Enabling Filename Translation” on page 128.

❑ Enable HTTP access from the CFA by following this procedure: “Phase 3 - Enable HTTP access from CFA” on page 130.

❑ Give CFA NFS permissions to read the Celerra file system by following this procedure: “Phase 4 - Give CFA NFS permissions” on page 130.

❑ Enable Celerra FileMover on the Celerra file system by following this procedure: “Phase 5 - Enable Celerra FileMover” on page 131

❑ Establish the connection to the CFA as a secondary server by following this procedure: “Phase 6 - Establish connection from the Celerra Network Server” on page 131.

Phase 1 - Enabling CFA

Where performed: On the CUA.

To enable CFA, do the following:

1. Carry out the CUA installation as described in “Installation and Configuration” on page 27.


3. Select D at the Setup Utility menu.

4. Read the warning and type Y at the prompt if you want to proceed:

WARNING: Configuring this system as a EMC Centera FileArchiver will automatically disable general NFS, CIFS, and FTP access. This system will only be used as a Celerra FileMover.Do you want to continue? (N/Y) : N

Enabling EMC Centera FileArchiver 127

128


5. Once enabled, the EMC Centera FileArchiver Configuration menu will appear:

A) Configure policiesB) Schedule policiesC) Simulate policiesD) Kill policy simulationF) Show current statisticsG) Generate orphaned listH) Clean orphansI) Kill orphaned list threadsJ) Run policies nowK) Kill policy runL) DNS settingM) Configure Digest AuthenticationN) Configure Access ProfilesO) Convert Celerra stubs to HTTP formatP) Kill HTTP conversion

6. Select A at the EMC Centera FileArchiver Configuration menu to open the editor and enter the EMC Centera FileArchiver Configuration File.

7. In the EMC Centera FileArchiver Configuration File, under the Define File Systems section, define the Celerra file systems to be migrated to the EMC Centera. Refer to “Editing the EMC Centera FileArchiver configuration file” on page 138 for instructions on editing this file.

Note: You must add file systems to the EMC Centera FileArchiver Configuration File before you connect to the Celerra Network Server in “Phase 6 - Establish connection from the Celerra Network Server” on page 131.

8. Save the file and exit the editor.

Phase 2 - Enabling Filename Translation

Where performed: On the Celerra Control Station.

CFA expects all filenames to come from the Celerra Network Server in UTF-8 format. To preserve filenames correctly, you must perform the following procedure:

1. Log onto the Celerra Control Station as user nasadmin.

2. Open the /nas/site/locale/xlt.cfg file in a text editor.



3. Locate the last line of the file. Typically, the last line of this file is:

::::8859-1.txt: Any thing that didn't match above will be assumed to be latin-1

Add the following line immediately prior to the last line:

::CFA_IP_ADDR::: CFA requires no translation (UTF-8)

where:

CFA_IP_ADDR = the IP address of your CFA system.

Example If your CFA IP address is 10.10.10.1, you add the following line:

::10.10.10.1::: CFA requires no translation (UTF-8)

Note: The order of these lines is very important.

4. From the Celerra Control Station run the command:

/nas/sbin/uc_config -update xlt.cfg

5. Run the following command to verify the configuration change:

/nas/sbin/uc_config -verify CFA_IP_ADDR -mover ALL

Where: CFA_IP_ADDR = the IP address of your CFA system.

Example /nas/sbin/uc_config -verify 10.10.10.1 -mover ALLOutput server_2 : 10.10.10.1 is UTF-8

server_3 : 10.10.10.1 is UTF-8


130


Phase 3 - Enable HTTP access from CFA


While still logged into the Celerra Control Station issue the following command from the Celerra CLI:

server_http { movername | ALL } -append dhsm -hosts CFA_IP_address

where:movername = name of the Celerra Data Mover.CFA_IP_address = IP address of the CFA.

Note: If you have multiple CFA addresses to which you need to enable access, the IP addresses can be wildcarded. For example, 10.10.10 will match everything in 10.10.10.*. Similarly, 1.2 will match 1.2.*.*.

If configuring for EA If configuring CFA in an enhanced availability configuration, issue the following command:

server_http { movername | ALL } -append dhsm -hosts Active_CFA_IP_address,Standby_CFA_IP_Address

Phase 4 - Give CFA NFS permissions


Give CFA NFS permissions to read the Celerra file system by entering the following command in the Celerra CLI:

server_export { movername | ALL } -Protocol nfs -option rw=client,root=client,access=client pathname

where:movername = name of the Celerra Data Mover.client = DNS name or IP address of the CFA server.

Note: If you have multiple CFA addresses to which you need to give permission, the IP addresses can be wildcarded. For example, 10.10.10.0/24 will match everything in 10.10.10 range.

pathname = NFS pathname on the Celerra.

Note: This can also be done from the Celerra Manager.



Phase 5 - Enable Celerra FileMover


Enable Celerra FileMover on the Celerra file system by typing the following command in the Celerra CLI:

fs_dhsm -modify fileystem_name -state enabled

where:filesystem_name = name of the Celerra file system.

Phase 6 - Establish connection from the Celerra Network Server


You can specify an NFS connection and/or an HTTP connection for stubbing files. With HTTP stubbing, the stubs reference the EMC Centera Content Address rather than the filename stored in the CFA database. If no connection is specified, the CFA will not function properly.

You can have both NFS and HTTP connections defined at the same time. If both types of connections exist, CFA uses the HTTP format when stubbing a file. If there are stubs on the system using NFS for access, you should not remove the NFS connection. If you want to use HTTP stubbing only, you need to convert all existing NFS stubs to HTTP. Refer to “Converting NFS stubs to HTTP” on page 133 for information on converting stubs.

Note: HTTP stubbing may be advantageous if you are setting up CFA in an enhanced availability configuration and wish to use DNS round robin for failover. Refer to “Using DNS round robin with CFA” on page 195 for information on DNS round robin.

Before establishing aconnection

You establish a connection from the Celerra with the fs_dhsm command. Before establishing a connection, you must define at least one file system in the FileArchiver configuration file. Refer to “Defining file systems” on page 139 for information.


132


Establishing an NFSconnection

To have CFA use NFS for stubbing files, enter the following command:

fs_dhsm -connection file_system_name -create -proto TCP -type nfsv3 -secondary DNS_hostname:/sfs/fileArchiver -read_policy_override passthrough -useRootCred true

where:filesystemname = the filesystem on the Celerra with FileMover enabled.DNS_hostname = the fully qualified hostname or IP address of the CFA. If using CFA in an EA configuration, it is strongly recommended that you use a DNS hostname.

Note: You must use the -read_policy_override passthrough option and the -useRootCred true option. CFA may not function properly without these options.

CFA does not support UDP connections from the Celerra to CFA.

Establishing an HTTPconnection

HTTP stubbing is only supported on Celerra Network Server version 5.5.27-5 and greater.

To have CFA use HTTP for stubbing files, enter the following command:

fs_dhsm -connection file_system_name -create -type http -read_policy_overrride passthrough -secondary http://DNS_hostname -cgi n

where:filesystemname = the filesystem on the Celerra with FileMover enabled.DNS_hostname = the FQDN DNS hostname, IP address of the CFA, or DNS Round Robin name. If using CFA in an EA configuration, it is strongly recommended that you use a DNS hostname.

Note: You must use the -read_policy_override passthrough option. CFA may not function properly without this option.



Converting NFS stubs to HTTPTo convert all NFS stubbed files to HTTP stubs:

IMPORTANT!This is an irreversible action. Once you convert the stubs to HTTP, they cannot be reconverted to NFS.

1. Select D at the Setup Utility menu to enter the EMC Centera FileArchiver Configuration menu.

2. Select O to convert the Celerra stubs to HTTP format. The following message appears:

WARNING: Once converted to use HTTP for secondary file access the stubs can not be converted back to NFS.

Do you want to continue? (Y/N)

3. Enter Y and press Enter.

Note: If you change the DNS name or a node name, you must rerun the HTTP conversion.


134


Configuring EMC Centera FileArchiverThis section describes how to configure the following EMC Centera FileArchiver features:

◆ “Configuring Digest Access Authentication” on page 134

◆ “Configuring the DNS setting” on page 135

◆ “Creating CFA access profiles” on page 136

Configuring Digest Access Authentication

When communicating with CFA, Celerra authenticates FileMover requests by IP address ACLs or by digest authentication. When using IP address ACLs, the Celerra accepts any incoming requests from the CFA’s IP address. The use of digest authentication restricts access to specific username/password pairs and so provides a higher level of access control.

If the Celerra uses digest authentication as the authentication method for the FileMover protocol, you must configure CFA to use digest authentication. If digest authentication is enabled, CFA uses a username and password to issue Celerra FileMover API requests to the Celerra.

Note: To configure digest access authentication on the Celerra, refer to the Celerra FileMover documentation.

To configure digest access authentication on CFA:


2. Select M at the EMC Centera FileArchiver Configuration menu. The following menu appears:

Centera FileArchiver Configuration digest authenticationA) Add/Modify userD) Delete userS) Show all users

Q) Quit without saving changesE) Save changes and exit



3. To configure a username/password pair, do the following:

a. Select A. CFA prompts you for the name of the Data Mover.

b. Enter the DNS name or IP address of the Celerra Data Mover and press Enter. CFA prompts you for the username.

c. Enter the username and press Enter. CFA prompts you for the password.

d. Enter the password for the account and press Enter.

e. Confirm the password for the account.

Example Data Mover name: datamover.company.comUser name: cfa_userPassword for server_3//cfa_user account: Enter password again:

4. Press E to save the changes and exit.

Configuring the DNS setting

When EMC Centera FileArchiver migrates a file from the Celerra to the EMC Centera, it places a stub file on the Celerra that points to the filename in the CFA’s database. If a connection is defined with the CFA’s DNS name, the file is stubbed using the DNS name. You can configure CFA to use the CFA’s IP address instead of its DNS name in stub files.

In addition, you can configure DNS round robin processing to support enhanced availability configurations. For information on configuring DNS round robin, refer to “Using DNS round robin with CFA” on page 195.

Note: It is strongly recommended that you use the CFA’s DNS name as this assists in migrations and disaster recovery scenarios.

To configure CFA to use the IP address in stub addresses:


2. Select L at the EMC Centera FileArchiver Configuration menu. The following prompt appears:

Is the Centera FileArchiver in a DNS domain? Y

3. Press the Backspace key and type N.

4. Press Enter.

Configuring EMC Centera FileArchiver 135

136


Using CFA access profiles

This version of CFA supports assigning multiple access profiles for migrating Celerra file systems to an EMC Centera. Access profiles are assigned to individual file systems through the CFA Configuration file. If a file system is not assigned a specific access profile, CFA uses the default CUA access profile to migrate the file system. If CFA cannot find a default CUA access profile, it then looks for a PEA file.

Creating CFA accessprofiles

Access profiles must exist on the EMC Centera before you can configure them in CFA. Refer to EMC Centera Online Help for information on creating access profiles on the EMC Centera.

Newly created profiles need to have the following capabilities:

◆ For EMC Centera 2.4 and 3.0:

• Read (r)• Write (w)• Query (q)• Delete (d)• Monitoring (m)

◆ For EMC Centera 3.1

• Read (r)• Write (w)• Delete (d)• Exist (e)• Query (q)• Clip-Copy (c)

In addition, you must enable the Monitoring role when creating the access profile.

To configure access profiles in CFA, do the following:


2. Select N at the EMC Centera FileArchiver Configuration menu. The following menu appears:

Access Profile Configuration----------------------------A) Add/modify access profilesR) Remove access profilesM) Map access profilesS) Show access profiles




Note: Option M is not used in configuring access profiles on CFA.

3. Select A. CFA displays the following prompts:

Access User:Access Password:Enter password again:

4. Enter the Access Username and Access Password and then confirm the password. CUA displays the following prompt:

Is this a cluster profile? (Y/N)

5. Enter Y or N depending on the kind of access profile you added.

CFA verifies the validity of this access profile with EMC Centera. If this is not a valid profile, you receive the following error:

Error validating the access profile for user 'xxxxx' (and/or the password).


7. Define the access profiles in the CFA configuration file. Refer to “Defining access profile groups” on page 140 for instructions.

Viewing accessprofiles

To view access profiles:

At the Access Profile menu, select S. CUA displays a list of all configured access profiles.

Configuring EMC Centera FileArchiver 137

138


Defining migration policiesEMC Centera FileArchiver (CFA) must be configured to define which files should be migrated from Celerra to EMC Centera.

Editing the EMC Centera FileArchiver configuration file

To edit the CFA configuration file, do the following:


2. Select A at the EMC Centera FileArchiver Configuration menu to open the editor and enter the EMC Centera FileArchiver Configuration File.

3. In the EMC Centera FileArchiver Configuration file, add the necessary information. Refer to “EMC Centera FileArchiver configuration options” on page 139, for information on the EMC Centera FileArchiver Configuration file.

Note: Comments in the configuration file are denoted by a # in the first column. All keywords are case insensitive.

4. Save the file and exit the editor.

5. If the format of the file is incorrect, the a messages such as the following will be displayed:

Parse error: Unknown keyword in line '[incorrectline] = 'Invalid config file format detectedDo you want to re-edit? (Y/N) (default = Y): y

6. Type Y to re-edit the file and correct the problem.

Note: If you choose not to re-edit the file, all changes will be lost and the file will revert to the previously saved version.



EMC Centera FileArchiver configuration optionsThis section explains the EMC Centera FileArchiver Configuration file in detail. These are the sections that can be configured:

◆ Define File Systems (required)

◆ Define File System Groups (optional)

◆ Define Access Profile Groups (optional)

◆ Define Policies (required)

◆ Define Migration (required)

Defining file systems The location of the files on the Celerra to be migrated to the EMC Centera are defined in the Define File Systems section. CFA supports up to 65 file systems per CFA system.

Note: You should add file systems to the CFA configuration file before adding connections to those file systems through the Celerra.

The format for Define File Systems is:

<logical_fs_name> = filesystem, <Celerra filesystem>, <Celerra DNS name or IP> [,<username>]

where:

logical_fs_name = a label for the file systemCelerra_filesystem = the export path for the file system being migrated tousername = the username used to make a CIFS-connection

(the user must have administrator privileges to access

the migrated files on the secondary storage system)

Note: CFA supports a maximum path length of 1024 characters.

Note: If configuring CFA in an enhanced availability configuration, entering the DNS name instead of the IP address is the preferred method.

You must create a line for every file system for which you want to make a policy.

Example: MarketingFS = filesystem, /fs0/directory1, 10.10.1.3SalesFS = filesystem, /fs0, celerra1.company.comEngineeringFS = filesystem, /fs0, 10.10.1.4

Defining migration policies 139

140


Defining file systemgroups

File systems may be grouped together in the Define File System Groups section.

The format for Define File System Groups is:

logical_group_name = policygroup, logical_fs_name[,logical_fs_name2]

Example SalesandMarketing = policygroup, MarketingFS, SalesFSAllFilesystems = policygroup, MarketingFS, SalesFS, EngineeringFS

Defining accessprofile groups

CFA supports the assignment of access profiles for migrating specific Celerra file systems to an EMC Centera. Access profiles are assigned to individual file systems through the CFA Configuration file.

Before assigning access profiles, the access profiles must be created on the CFA. For instructions on creating access profiles, refer to “Creating CFA access profiles” on page 136.

You assign access profiles to different file systems in the Define Access Profile Groups section of the EMC Centera FileArchiver Configuration. The format for assigning access profiles is:

access_user = accessprofile, logical_fs_name [,logical_fs_name]

where:

access_user = name of the access profile used for migrationaccessprofile = literal valuelogical_fs_name = a label for the file system

Example CenteraEngUser = accessprofile, EngineeringFS, QualityAssuranceFSCenteraMarketUser = accessprofile, MarketingFS

Note: If you do not define an access profiles for a specific file system, the default profile set on the underlying CUA is used. If there is no default access profile, a PEA file is used.



Defining policies The type of files to be migrated are defined in the Define Policies section.

The format for Define Policies is:

logical_policy_name = keyword, arguments

The valid keywords and arguments are described in the following sections:

ModificationThe argument for the Modification keyword is a value specifying the amount of time since the file was last modified. The time may be expressed in hours (h,H), days (d,D), months (m,M), or years (y,Y).

The format for Modification is:

logical_policy_name = modification, [hours,days,months,or years]

Example policy-mod-1m = modification, 1mpolicy-mod-3d = modification, 3d

Note: A month is valued at 30 days and a year at 365 days.

Note: As there is no time synchronization between the Celerra and the CUA, the administrator must ensure that both systems are in the same time zone and are set to the same time.

AccessThe argument for the Access keyword is a value specifying the amount of time since the file was last accessed. The time may be expressed in days (d,D), months (m,M), or years (y,Y).

The format for Access is:

logical_policy_name = access, [hours,days,months,or years]

Example policy-access-1m = access, 1mpolicy-access-3d = access, 3d

Note: A month is valued at 30 days and a year at 365 days.


142


Note: As there is no time synchronization between the Celerra and the CUA, the administrator must ensure that both systems are in the same time zone and are set to the same time.

ExtensionThe argument for the Extension keyword is a comma separated list of case insensitive file extensions.

The format for Extension is:

logical_policy_name = extension, file_extension[,file_extension]

Example policy-pdf = extension, pdfpolicy-video = extension, mpeg, mpg, qt, mp2, mp4

SizeThe argument for the Size keyword is a value specifying the minimum size of a file. The size may be expressed as kilobyte (k,K), megabyte (m,M), or gigabyte (g,G).

The format for Size is:

logical_policy_name = size, size_in_bytes

Example policy-size-1g = size, 1gpolicy-size 10m = size, 10m

DirectoryThe argument for the Directory keyword is a directory name. All files and directories under this directory will be included in the policy. The name should not include the file system name specified in the Define File Systems section.

The format for Directory is:

logical_policy_name = directory, directory_name_under_a_filesystem

Example policy-dir-directory1 = directory, directory1policy-dir-dirA = directory, directory0/dirA

Note: A directory may be excluded from the policy by defining the directory in this section and then using the not operator in the Define Migration section.



File NameThe argument for the File Name keyword is an alphanumeric string. Any files with a file name that match the string are included in the policy. The string matching is case sensitive.

Valid characters are any valid UTF-8 character that Linux accepts for a file name. The asterisk (*) or question mark (?) characters can be used as wildcards—an asterisk selects any string of characters and a question mark selects for a single character. You cannot use slash (/) character.

The format for File Name is:

logical_policy_name = filename, string

Example policy-filename-name = filename, *dataMatches any files that end with the string "data."

policy-filename-name = filename, data?Matches any files that contain the string "data" plus one other character.

File Path The argument for the File Path keyword is an alphanumeric string. Any files with a file path that includes the string are included in the policy. The string matching is case sensitive.

Valid characters are any valid UTF-8 character that Linux accepts for a file path. The asterisk (*) or question mark (?) characters can be used as wildcards—an asterisk selects any string of characters and a question mark selects a single character.

The format for File Path is:

logical_policy_name = filepath, string

Note: All paths are relative to the mount point. For example, /*/file is the same as */file.

Example policy-filepath-name = filepath, /*/*/*/??Matches any files that have 2 character names and are 3 directory levels down.


144


Defining migration The Define Migration section is used to define:

◆ Which Policies are applied to which File Systems or File System Groups.

◆ The retention period for the files.

◆ The stub delay for the files.

The time may be expressed as hours, (h,H), days (d,D), months (m,M), or years (y,Y).

The operators and, or, and not can also be used.

Note: If you have multiple migrate policies that effect the same file, the first migrate command that matches the file is the one applied. Any subsequent matching migrate policies are ignored.

The format for Define Migration is:

migrate logical_policy_name on logical_group_name [retention num_days] [stub num_days]

Options for migrate:

retentionSets the retention period for the migrated files. Indicate the number of days the file should be retained. If you want to set the retention to infinite, enter the word infinite.

Note: The retention value is set on EMC Centera. The files can still be deleted from Celerra. The file will not automatically be deleted from EMC Centera if it is deleted from Celerra—these files become orphan files and must be manually deleted by the administrator. Refer to “Managing orphans” on page 155 for more information.

stubBy default, files migrated to the EMC Centera are immediately stubbed on the Celerra. By specifying the stub option, you can delay the stubbing. The stub delay for a file is calculated from the time that the file was migrated. Depending on how often you run the policy engine, the actual stub delay may exceed that set by the policy. For example, if you run the policy engine once a week, and the stub delay on a file is set to 30, the file may not actually be stubbed until 37 days after the file is migrated.



Example 1 migrate policy-mod-5d on EngineeringFS

migrate policy-mod-3d on MarketingFS

Note: Adding a retention period is not obligatory.

Example 2 migrate policy-video on AllFilesystems retention 365d

In this example, all files that have the extension mpeg, mpg, qt, mp2,or mp4 in all file systems would be migrated with a retention 365 days.

Example 3 migrate (policy-mod-3d or policy-size-10m) and policy-extension-pdf on AllFS

In this example, parentheses have been used to group policies with the or operator. In this case, all files in all file systems that have the extension pdf and that have not be modified for 3 days or have a size of 10 megabytes will be migrated. Parentheses can be nested.

Example 4 migrate policy-mod-3d on MarketingFS stub 30d

In this example, all files that have not been modified for the last 3 days in the file system MarketingFS would be migrated but they would not be stubbed until 30 days from the migration date.

Note: Adding a stub delay is not obligatory.

Example EMC Centera FileArchiver configuration file# Define File Systems#

MarketingFS = filesystem, 10.10.1.3:/fs0SalesFS = filesystem, 10.10.2.4:/fs0EngineeringFS = filesystem, /fs0, 10.10.1.4

# Define File System Groups#

AllFilesystems = policygroup, MarketingFS, SalesFS, EngineeringFS

# Define Policies#

policy-mod-3d = modification, 3dpolicy-size-1m = size, 1m


146


# Define Migration#

migrate policy-mod-3d and policy-size-1m on AllFilesystems retention 6m stub 30d

Note: Individual migration policies cannot contain page breaks.



Scheduling a policy runTo schedule a policy run, do the following:

1. Select B at the EMC Centera FileArchiver Configuration menu. The following menu appears:

Centera FileArchiver Schedule policies--------------------------------------A) AddD) DeleteS) Show schedule

Q) Quit without saving changesE) Save changes and Exit

2. Select A to schedule a policy. The following prompt appears:

Day(s) of the month, example: 1,10,20(comma separated, * => every day):

Note: The valid range of values is 1 to 31. Use * to denote all days of the week.

3. Enter the days of the month on which you want the policy to run and press Enter. The following prompt appears:

Day(s) of the week, example: Sunday,Monday(comma separated, * => every day):

Note: Use * to denote all days of the month.

4. Enter the days of the week on which you want the policy to run and press Enter. The following prompt appears:

Time :

5. Enter the time of day at which you want the policy to run and press Enter.

6. Select E to save the schedule.

Notes◆ Policy schedules specifying specific days override schedules

using an asterisk. If you use an * to denote all days of the month and then specify specific days of the week, the policy runs only

Scheduling a policy run 147

148


on the days of the week specified. Similarly, if you specify specific days of the month and use an * for days of the week, the policy runs only on the days of the month specified.

◆ If the policy includes a large number of files, the policy run may take more than 24 hours. If your schedule runs every 24 hours and a second one begins before the previous run ends, the new run will be cancelled. All errors are logged in the CUA Event Viewer (see “Viewing event messages” on page 220 for more information).

◆ If a policy is running and the system reboots, the policy must be manually re-started by selecting J at the EMC Centera FileArchiver Configuration menu. If the policy is not restarted, the policy will run again at the next scheduled time.

Deleting a schedule To delete a policy schedule:

1. Select B at the EMC Centera FileArchiver Configuration menu.

2. Select D to delete a schedule. The following prompt appears:

On the <list of scheduled days and time>Delete this schedule item? (Y/N): N

3. Replace the N with a Y and press Enter.

4. Select E to save the change.



Manually starting and stopping a policy runThis section describes how to manually start and stop policies.

Note: Only one migration policy may run at a time. If you try to run a migration policy while one is already in progress you will receive an error message asking if you want to restart the migration.

You can run a migration concurrent with a simulation.

To manually start a policy, do the following:

1. Select J at the EMC Centera FileArchiver Configuration menu.

2. The following will be displayed:

Centera FileArchiver Run policies---------------------------------A policy engine has been started in the background


3. Press any key to continue and return to the EMC Centera FileArchiver Configuration.

To stop a running policy, do the following:

1. Select K at the EMC Centera FileArchiver Configuration menu.

2. The following will be displayed:

Are you sure you want to stop process pid xxxx? (Y/N): yStopped background process xxxxxPress any key to continue

3. Press any key to continue and return to the EMC Centera FileArchiver Configuration.

Note: A stopped policy run may have already migrated some files. You can view the actual results of a policy run in the log files through HTTP at: http://system_name/logs/fileArchiver/migrated/1/ or /2/, etc.

Note: If a policy is running and the system reboots, the policy must be manually re-started by selecting J at the EMC Centera FileArchiver Configuration menu. If the policy is not restarted, the policy will run again at the next scheduled time.

Manually starting and stopping a policy run 149

150


Simulating a policy runPolicy simulation runs a policy in the background but does not carry out the actual migration to EMC Centera. At the end of the policy simulation, EMC Centera FileArchiver saves a log file of the files that would have been migrated. The file is saved in the following directory tree:

/sfs/logs/fileArchiver/simulation/

Note: The removal of files in the /sfs/logs directories will impact behavior of CUA writeback logging. If one or more files in this subdirectory is accidentally removed, the CUA must be rebooted to re-enable writeback logging.

The name of the log file conforms to the following format:

fileArchiver_simulation+date+time.log

Every 200 logs, a new subdirectory is created starting with the name of 1 and incrementing by 1 for each new subdirectory.

Example /sfs/logs/fileArchiver/simulation/1/fileArchiver_simulation-2004-12-20-10-37.log

To run a Policy Simulation, do the following:

1. Select C at the EMC Centera FileArchiver Configuration menu and the following output will be displayed:

Centera FileArchiver Simulate policies--------------------------------------A policy engine simulation has been started in the background


Note: Only one policy simulation may run at a time. If you try to run a simulation while one is already in progress you will receive an error message asking if you want to restart the simulation.

You can run a migration concurrent with a simulation.

A policy simulation is currently in progress, would you like to restart the simulation? (y/n)



2. View the log file through HTTP at:

http://system_name/logs/fileArchiver/simulation/1 /fileArchiver_simulation-2004-12-20-10-37.log

Stopping a policy simulation

To stop a Policy Simulation, do the following:

1. Select K at the EMC Centera FileArchiver Configuration menu.

2. Type Y at the prompt and press Enter:

Are you sure you want to kill the Policy Simulation? n (y/n)

3. If no simulation is running, the following error will be displayed:

No Policy Simulation is currently running

Simulating a policy run 151

152


Monitoring EMC Centera FileArchiverThe following sections provide information on monitoring the EMC Centera FileArchiver.

Viewing EMC Centera FileArchiver statistics

EMC Centera FileArchiver provides statistics about the currently running policy (or the last run policy) and a simulation of a policy.

To view Statistics, do the following:

1. Select F at the EMC Centera FileArchiver Configuration menu.

2. Output similar to the following will be displayed:

Running Policy:Start time: Thu Dec 15 12:08:38 2005End time: Thu Dec 15 12:08:43 2005Total: 20 filesMatching policy: 20 filesAlready stubbed: 0 filesMigrated: 20 files, 4000 bytesStubbed: 20 files, 4000 bytesStub pending: 0 files, 0 bytesPending due to err: 0 files, 0 bytesMoved/renamed: 0 files

Simulation:Start time: Fri Mar 10 16:10:11 2006End time: Fri Mar 10 16:10:16 200Total: 10 filesMatching policy: 10 filesAlready stubbed: 0 filesSimulated: 10 files, 10240 bytesStub pending: 0 files, 0 bytesMoved/renamed: 0 files



Viewing EMC Centera FileArchiver activity logs

The EMC Centera FileArchiver activity log provides a consolidated view of CFA activity. This log file is appended to each time the policy is run or orphans are deleted. When the log file size exceeds 5 MB, CFA creates a new version of the log file.

To view the CFA activity log:

1. From a web browser, access the following URL:

http://CFA_name/logs/fileArchiver/activity/1/

where: CFA_name = the IP address or DNS name of the CUA.

Note: You can view the actual results of a policy run in the log files through HTTP at: http://CFA_name/logs/fileArchiver/migrated

2. Select the specific log you want to view. Log files names are in the following format:

fileArchiver_activity-<date_time_stamp>.log

The format of information in the log file is as follows:

date, action, logical_file_system, file_count, total_bytes

ExampleWed Oct 12 09:21:35 2005, Action, File system, File count, Total bytesWed Oct 12 09:21:35 2005, SIMULATED, logicalFs2, 10, 10240Wed Oct 12 09:48:07 2005, MIGRATED, logicalFs2, 2, 2048Wed Oct 12 10:49:17 2005, RENAMED, logicalFs1, 17, naWed Oct 12 11:43:17 2005, DELETED, logicalFs2, 55, 56320

Viewing the EMC Centera FileArchiver migration log

The EMC Centera FileArchiver migration log provides detailed information about all migrated files. A new migration log file is created each time the policy is run. When the log file size exceeds 5 MB, CFA creates a new version of the log file.

To view the CFA migration log:

1. From a web browser, access the following URL:

http://CFA_name/logs/fileArchiver/migrate/1

where: CFA_name = the IP address or DNS name of the CUA.

Monitoring EMC Centera FileArchiver 153

154


2. Select the specific log you want to view. Log files names are in the following format:

fileArchiver_migrated-<date_time_stamp>.log

Table 17 on page 154 shows the format of information in the log file.

Table 17 FileArchiver Migration Log Information

Column Name Description

Date Date the file was migrated.

File System Name of the Celerra file system.

File Name Name of the file migrated.

Local File Name Name of the file on CFA.

File Size Size of the file.

Retention (days) Number of days that the file will be retained. If no retention policy was set, this column shows a 0.

Stub Delay Stub delay set on the file. If there is no stub delay set, this column shows a 0.

Data Copied Indicates if the data has been copied to the EMC Centera.

File Stubbed Indicates if the file has been stubbed.



Managing orphansFiles that are stored on the EMC Centera, but do not have a related file on the Celerra are called orphans. Orphans are created in two ways:

◆ When a file is deleted on Celerra — Deleted files remain archived on EMC Centera even though the file no longer exists on the EMC Celerra.

◆ When a user modifies a file on the Celerra — When a file is modified, CFA migrates the new version of the file while still retaining the old version on the EMC Centera. These older versions have no equivalent files on the Celerra and so are technically orphans.

Note: CFA does not automatically delete orphans. Orphan deletion is a manual process where the CFA administrator decides which, if any, orphans to delete.

This section describes how to manage orphans. Managing orphans includes:

◆ “Finding orphans” on page 155

◆ “Excluding orphans from deletion” on page 157

◆ “Deleting orphans” on page 157

◆ “Stopping the orphan processes” on page 158

Finding orphans In order to manage and delete orphans, you must first find orphans. FileArchiver searches for orphans by both their inode Celerra handles and their filenames and generates orphan lists.

To find orphans, do the following:

1. Select G at the EMC Centera FileArchiver Configuration menu. The following prompt appears:

Do you consider old versions of files as orphans? (Y/N): N

Managing orphans 155

156



• To exclude older versions of files from the orphan list, press Enter.

• To include older file versions in the orphan list, type Y and press Enter.

3. Output similar to the following will be displayed:

Centera FileArchiver Generate orphaned list-------------------------------------------Background tasks are generating orphaned lists:

/sfs/logs/fileArchiver/orphaned/orphaned.list/sfs/logs/fileArchiver/orphaned/orphaned_with_retention.listPress any key to continue

The two files that are created contain the following:

◆ orphaned.list — this file contains a list of the orphaned files. The file format is:

[CUA_filename], [CelerraFilename], [Date]

Example "/sfs/fileArchiver/00048/00015/FS3~~~~4294967327-78739-1113409940", "/mnt/gateway/10.10.10.10/FS3/localdirs/26210/data/disk0.dat", "Wed Apr 13 02:55:45 2005

The CUA_filename is the full path name of the file on the CUA. This filename includes the Celerra inode HANDLE. The CelerraFilename is the last known filename for this file.

◆ orphaned_with_retention.log — this file contains the list of files that are orphaned on EMC Centera but will not be deleted because their retention period has not expired. The format of this file is:

[CUA_filename], [CelerraFilename], [timestamp of, when retention expires]

Example "/sfs/fileArchiver/00036/00161/FS3~~~~4294967327-62979-1133634455", "/mnt/gateway/10.10.10.10/FS3/retentiontest.ppp", "Sat Apr 23 19:22:29 2005"

The CUA_filename is the full path name of the file on the CUA. This filename includes the Celerra inode HANDLE. The CelerraFilename is the last known filename for this file. The timestamp is the date when retention will expire.



Note: The orphaned_with_retention.list is for informational purposes only. In order to actually delete a file that currently has a retention period, a new orphan list must be generated when the retention period has expired.

Both of these files are replaced each time an orphaned list is generated. Errors are logged to the CUA Event Viewer (see “Viewing event messages” on page 220 for more information).

Excluding orphans from deletion

Before deleting the files from the orphaned list, you may first want to exclude certain files that you do not want deleted.

To exclude files that you do not want deleted, do the following:

1. Select X at the main Setup Utility menu to enter the Bash shell.

2. Type the following command to edit the file:

pico /sfs/logs/fileArchiver/orphaned/orphaned.list

3. Remove the files that you do not want deleted.

4. Exit the editor and save the file.

Deleting orphans Once you have created the orphaned list and excluded any files that you do not want deleted, you can run the process to delete all orphans.

Note: Before deleting orphans, you should consider whether older file versions are part of a snapshot or backup file that may need to be restored. If files referenced in a backup file or snapshot are deleted, the snapshot will be unable to find the referenced file on the EMC Centera.

To delete orphans, do the following:

1. Select H at the EMC Centera FileArchiver Configuration menu.

2. Type Y at the prompt:

Are you sure you want to delete orphaned files? (Y/N): y

3. Deleting orphans will begin and the following output will be returned when delete is complete:


Managing orphans 157

158


4. Press any key to return to the EMC Centera FileArchiver Configuration menu.

All deleted files are logged in the following file:

/sfs/logs/fileArchiver/deleteOrphans/1/fileArchiver_deleteOrphansTIMESTAMP.log


Errors are logged in the CUA Event Viewer (refer to “Viewing event messages” on page 220).

Stopping the orphan processes

To stop the creation of the orphan list or to stop the deleting of orphans, do the following:

1. Select I at the EMC Centera FileArchiver Configuration menu.

2. Type Y at the prompt:

Are you sure you want to stop process pid xxxx? (Y/N): Y

3. All processes will be stopped.


5

This chapter explains how to setup CUA/CFA for Enhanced Availability and includes information on configuring the Standby system and resolving a CUA/CFA failure.


◆ Introduction to CUA/CFA enhanced availability....................... 160◆ Overview of CUA/CFA enhanced availability ........................... 161◆ Types of CUA/CFA EA configurations ........................................ 163◆ Configuring enhanced availability................................................ 169◆ Resolving an Active CUA/CFA failure ........................................ 173◆ Reconfiguring the old Active system............................................ 178◆ Using a remote Standby CUA/CFA and EMC Centera ............. 185◆ Using a remote Standby CFA, EMC Centera, and Celerra ........ 190◆ Using DNS round robin with CFA ................................................ 195

CUA/CFA EnhancedAvailability

CUA/CFA Enhanced Availability 159

160

CUA/CFA Enhanced Availability

Introduction to CUA/CFA enhanced availabilityTo retrieve files stored on the EMC Centera (or in the case of CFA, to migrate files from the Celerra to the EMC Centera) requires all components in a CUA/CFA configuration (CUA/CFA, EMC Centera, and Celerra) to be functioning properly. A single failed component can result in business discontinuity or data unavailability.

In the simplest CUA/CFA configuration, there is a single component for each device type. This configuration provides no high availability protection. To provide higher availability, you should set up replication on each of the individual devices. It is important to understand the various protection mechanisms for individual components as well as their inter-relationships.

Note: Although Celerra FileMover provides various disaster recovery protection schemes—synchronous, point-in-time, and asynchronous—CFA and EMC Centera only support asynchronous disaster recovery protection. This forces the entire solution to be asynchronous regardless of which disaster recovery protection is used by Celerra FileMover.



Overview of CUA/CFA enhanced availabilityTable 18 on page 161 contains an in depth overview of this chapter and provides a workflow for configuring CUA/CFA failover and failing over a CUA/CFA system.

Table 18 CUA/CFA failover overview (1 of 2)

Overview of enhanced availability configurations

• Local standby CUA/CFA Configuration where the Active and Standby systems are local to the EMC Centera

“Local standby CUA/CFA” on page 164

• Remote standby CUA/CFA with remote EMC Centera

Configuration where there is a local EMC Centera and an Active CUA/CFA and a remotely sited standby EMC Centera with Standby CUA/CFA

“Remote standby CUA/CFA with remote EMC Centera” on page 166

• Remote Standby CFA, EMC Centera, and Celerra

Configuration with full site redundancy where this is a local EMC Centera, Celerra, and Active CUA/CFA and a remote EMC Centera, Celerra, and Standby CUA/CFA

“Using a remote Standby CFA, EMC Centera, and Celerra” on page 190

Configuring enhanced availability

• Configuring the Active CUA/CFA

Instructions on configuring the Active system “Configuring the Active CUA/CFA” on page 170

• Configuring the Standby CUA/CFA

Instructions on configuring the Standby system “Configuring the Standby CUA/CFA” on page 171

• Using DNS round robin for CFA EA

Instructions on configuring DNS round robin.

Note: The use of DNS round robin greatly simplifies failover in a CFA environment.

“Using DNS round robin with CFA” on page 195

Failing over an Active CUA/CFA system

• Failover of a local standby system

Failover instructions for a simple local standby configuration (one EMC Centera with a local Active and Standby CUA/CFA)

“Resolving an Active CUA/CFA failure” on page 173

• Failover of a remote standby CUA/CFA with a remote EMC Centera

Failover instructions for a configuration with a local EMC Centera with an Active CUA/CFA and a remote site with a secondary EMC Centera and a Standby CUA/CFA

“Failing over to a Standby CUA/CFA and a secondary EMC Centera” on page 187

• Failover of to a remote Standby CFA, EMC Centera, and Celerra

Failover instructions for a CFA configuration with full site (EMC Centera, Celerra, and CFA) redundancy

“Failover for full remote CFA configuration” on page 192

Overview of CUA/CFA enhanced availability 161

162


Recovering an old Active system

• Recovering unwritten data from a failed CUA

When failing over the Active CUA, some data may not have been written to the EMC Centera at the time of failover. This section provides instructions on recovering that data.

“Recovering unwritten data from the old Active CUA” on page 179

• Reconfiguring the failed system

Instructions for turning the failed system into the new Standby system in a CUA/CFA EA configuration

“Reconfiguring the old system” on page 183

Table 18 CUA/CFA failover overview (2 of 2)



Types of CUA/CFA EA configurationsYou enable enhanced availability for CUA/CFA by configuring either a local or remote Standby CUA/CFA. Each Active system can only use one Standby system. An EMC Centera cluster can accommodate two pair of Active and Standby CUA/CFAs (quantity 4 nodes).

The Active system constantly updates its activity logs to the Standby—by default this occurs every 5 minutes or after 500 operations. In case a disaster occurs on the Active system, you can manually failover to the Standby system so that the Standby system takes on the responsibilities of the active participant.

Note: CFA can use HTTP instead of NFS for file stubbing on the Celerra. In an EA configuration, HTTP stubbing along with DNS round robin, allows the Celerra to retrieve data from either the Active or Standby system. This read redundancy eliminates the need for failover if the Active CFA is unavailable. It is strongly recommended that you use HTTP stubbing and DNS round robin for your CFA EA configurations. Refer to “Using DNS round robin with CFA” on page 195 for a description of this type of configuration.

CUA/CFA can be set up in three types of enhanced availability configurations:

◆ Local standby CUA/CFA—This configuration provides redundancy for the CUA/CFA but not for EMC Centera.

◆ Remote standby CUA/CFA and remote standby EMC Centera—This configuration provides redundancy for both the CUA/CFA and EMC Centera. There is also enhanced protection due to site redundancy.

◆ Remote standby CFA with remote EMC Centera and Celerra — This configuration provides full redundancy for all components in the CFA configuration. There is also enhanced protection due to site redundancy. Refer to “Using a remote Standby CFA, EMC Centera, and Celerra” on page 190 for information about this configuration.

Types of CUA/CFA EA configurations 163

164


Local standby CUA/CFA

This configuration provides disaster recovery protection for the CUA/CFA but not for the EMC Centera. This is the simplest EA configuration to setup; however, with this configuration, the EMC Centera becomes a single point of failure. Because all components in this configuration are located at the same site, this configuration is more prone to downtime in the event of things like network and power failures. Figure 12 on page 164 shows the this configuration and the failover for CUA. Figure 13 on page 165 shows the this configuration and the failover for CFA.

Figure 12 EA configuration with a local standby CUA

`

Client Client

Centera

ActiveCUA

StandbyCUA

Logs

Local standby

During

failo

ver

During failover

`

Client Client

Centera

Failed Active CUA

New ActiveCUA

(original Standby

CUA)

Failed over local



Figure 13 EA configuration with a local standby CFA

System requirementsfor local standby

This configuration requires the following:

◆ One EMC Centera.

◆ Two CUA/CFA systems. Both the Active and Standby system must be running on the same hardware family and class (in other words, a Dell 2850 Active must use Dell 2850 Standby, and a Gen4 Active must use a Gen4 Standby). The CUA/CFA systems must be running v 3.6 SP1 or greater.

◆ For CFA: One Celerra running DART OS v5.4 or greater.

During

failove

r

Centera

Failed ActiveCFA

Celerra

New ActiveCFA

(original Standby

CFA)

Client

`

Client

During failover

Centera

ActiveCFA

Celerra

StandbyCFA

Client

Logs

`

Client

Local standby Failed over local standby


166


Remote standby CUA/CFA with remote EMC CenteraTo provide a higher level of disaster recovery, you can replicate both a Standby CUA/CFA and a replica EMC Centera to a remote site. In this configuration, the EMC Centera is replicating data uni-directionally.

For this configuration, communication must be established to all devices across sites. Both the Active and Standby systems should point to the primary EMC Centera as the target.

Note: There are special failover tasks that must be done with this type of configuration. Refer to “Using a remote Standby CUA/CFA and EMC Centera” on page 185 for failover instructions.

Figure 14 on page 167 shows the setup of this configuration. In this diagram, the dashed arrows indicate the direction of replication:

◆ For the CUA/CFAs, the activity logs are replicated.

◆ For the EMC Centera, data is replicated.



Figure 14 Remote standby CUA with remote EMC Centera

Centera

ActiveCFA

Celerra

`

Client Client

Centera

Standby CFA

Logs

Data

Centera

ActiveCUA

`

Client Client

Centera

Standby CUA

Logs

Data

Remote standby CFA with remoteEMC Centera

Remote standby CUA with remote EMC Centera


168


System requirementsfor remote standby

This configuration requires the following:

◆ Two EMC Centeras.

◆ Two CUA/CFA systems. Both the Active and Standby system must be running on the same hardware family and class (in other words, a Dell 2850 Active must use Dell 2850 Standby, and a Gen4 Active must use a Gen4 Standby). The CUA/CFA systems must be running v 3.6 SP1 or greater.

◆ Centera Replication License.

◆ A high speed network with reliable cross site network communication.

Note: ICMP Ping packets must be routable between the two sites.

◆ For CFA: One Celerra running DART OS v5.4 or greater.



Configuring enhanced availability Installing a Standby CUA/CFA ensures that service interruptions due to the failure of the main CUA/CFA (Active system) are minimized. A Standby system can be installed any time after the Active system becomes operational.

Before proceeding, ensure that you have the following information available:

◆ IP Addresses — different on Active and Standby systems

◆ Hostnames — different on Active and Standby systems

◆ License key — different on Active and Standby systems

Note: The license key is the 15-character Serial Number found on the part number label on the outside of the CUA/CFA kit.

◆ EMC Centera information — same on Active and Standby system

◆ Email configuration — same on Active and Standby systems

Notes◆ If you use access profiles (or a PEA file), the same profiles (or PEA

file) must be in place on both the Active and Standby CUA/CFAs. Before you set up the EA relationship, you must configure the Standby CUA/CFA with the same profile (or PEA file) used to do the last backup on the Active CUA/CFA.

For CFA systems with a secondary EMC Centera, the same profiles (or PEA files) must be on both the primary and secondary EMC Centera.

◆ It is highly recommended that the system times on the Active and Standby systems are synchronized in order to avoid unexpected results in the event of a failover.

◆ The Active and Standby systems must have unique IP addresses and hostnames while they are both up and running on the network.

◆ For CUA: In an Enhanced Availability configuration, the administrator on the Standby CUA should not configure FTP or modify the NFS or CIFS configurations. The Standby CUA will automatically receive these configuration settings from the Active CUA. These configuration settings should not be explicitly set on the Standby CUA.

Configuring enhanced availability 169

170


◆ For CFA: Do not configure CFA on the Standby system. On the Standby, CFA is automatically enabled upon failover. For a CFA EA configuration, install CUA on both the Active and Standby systems and then configure CFA on the Active system only.

◆ For CFA: You must enable http access from the Celerra to CFA with the IP address of the Active CFA and the IP address of the Standby CFA. Refer to “Phase 3 - Enable HTTP access from CFA” on page 130 for instructions.

Configuring the Active CUA/CFA

To configure Enhanced Availability, you must first configure the Active CUA/CFA:

1. Log in to the Active system using an SSH client and log in as gwsetup using the password EMCCUA.

2. Select C at the Setup Utility menu to open the Configuration menu.

3. Select V at the Configuration menu to begin the configuration. Refer to the following table for system prompts and the values you should enter.

4. Press Enter on your keyboard.

Prompt Enter

Do you want to change this configuration? (y=yes, n=no)[n]:

Y

Is this system an Active CUA?(y=yes, n=no)[y]: y

Y

Will there be a Standby CUA?(y=yes, n=no)[n]: y

Y

Standby CUA name or address: Enter the hostname or IP address of the Standby CUA.

Note: It is highly recommended that you use a DNS hostname rather than an IP address to identify the Standby system. The use of DNS hostnames greatly simplifies failover procedures.



5. After the message Active System configuration completed appears, press Enter.

6. Select Q to exit the Configuration menu and return to the main menu

7. Select X to enter a Bash Shell


tail /var/log/storigen_nms

9. Look for output similar to the following to ensure that the Active CUA is configured correctly:

Oct 28 09:32:05 CUA1 nmslib[603]: [75] [28-Oct-2004 9:32:05.560] [sev 2] [SFSBACKUP] Backup complete, CA is 3PADPTQ8KHOCIe3NGHJ48T4S5VG

Note: Re-execute step 8 if you do not see the Backup Complete message.

Refer to “Viewing event messages” on page 220 for more information on viewing event messages.

Configuring the Standby CUA/CFA

Once the Active CUA/CFA is configured for enhanced availability, the Standby system must be configured:

Note: If there is a time lag between the time you configure the Active CUA and the Standby CUA, it is strongly recommended that you backup the Active CUA before configuring the Standby. Doing a backup will reduce the time it takes for the Standby to come on line.

1. Log in to the Standby system using an SSH client and log in as gwsetup using the password EMCCUA.

2. Select C at the Setup Utility menu to enter the Configuration menu.

Configuring enhanced availability 171

172


3. Select V at the Configuration menu to begin the configuration. Refer to the following table for system prompts and the values you should enter.

4. After the message Active System configuration completed appears, press the Return key on your keyboard.

5. Select Q to exit the Configuration menu and return to the main menu.

6. Select X to enter a Bash Shell.



8. Look for output similar to the following:

Oct 28 09:27:17 CUA2 nmslib[619]: [60] [28-Oct-2004 9:27:17.047] [sev 2] [SFSRESTORE] Restore complete from CA 3PADPTQ8KHOCIe3NGHJ48T4S5VG

Note: Re-execute step 7 if you do not see the Restore Complete message. Depending on the number of files that have been migrated or exist on the CFA, this may command may take a while to complete.

Prompt Enter


Y


N

Enter the name or IP address of the Active CUA:

Enter the hostname or IP address of the Active CUA.

Note: It is highly recommended that you use a DNS hostname rather than an IP address to identify the Active system. The use of DNS hostnames greatly simplifies failover procedures.



Resolving an Active CUA/CFA failureIf there is a catastrophic failure on the Active CUA/CFA, you can restore service by manually transitioning the Standby system to an active state. Figure 15 on page 173 shows the stages of the CUA/CFA failover and failback procedure.

Figure 15 CUA/CFA failover and failback procedure

`

Old Active CUA/CFA

`

Active CUA/CFA

failover

`

Active CUA/CFA

`

Standby CUA/CFA

`

Standby CUA/CFA

`

Active CUA/CFA

`

Active CUA/CFA

`

Standby CUA/CFA(re-installed with CUA software)

failback

`

Active CUA/CFA

`

Standby CUA/CFA

5.

1.

2.

3.

4.

Establish EA configuration

Active CUA/CFA fails

Failover to Standby CUA/CFA

Configure old Active as new Standby

(Optional )Failback to original EA configuration. Re-install CUA on the original Standby.

Resolving an Active CUA/CFA failure 173

174


Table 19 on page 174 describes the stages of CUA/CFA failover and failback and provides links to the specific procedures.

Notes If a default profile is used on the CUA underlying your Active and Standby CFA, the profile must be in place on both the Active and Standby systems. If you are using PEA files instead of access profiles, the PEA file must be copied to the Standby system. Refer to “Using access profiles” on page 88 for more information.

Testing CUA/CFA failover

To test the CUA/CFA failover procedure in a test environment, do the following:

1. Log in to the Active system using an SSH client and log in as gwsetup using the password EMCCUA.

2. Select S at the Setup Utility menu to shutdown the Active system.

3. Connect to the Standby system Setup Utility using an SSH client.


Table 19 Stages of CUA EA configuration and failover

Stage Description Procedure

1 Establish EA configuration by configuring Active and Standby systems.

“Configuring enhanced availability” on page 169.

2 Active system fails. n/a

3 Failover to Standby system. The Standby system becomes an Active standalone system.

• For testing failover, refer to “Testing CUA/CFA failover” on page 174.

• For failing over in a production environment, refer to “Initiating failover in a production environment” on page 175.

4 After failover, the EA configuration must be reconfigured with the old Active system as the new Standby.Note: Reinstall the old Active system if there is no data in the lost+found directory.

“Reconfiguring the old system” on page 183.

5 (Optional) Return to the original EA configuration by failing back (a secondary failover) the current Active system to the new Standby system. Reinstall CUA software on the current Active system and reconfigure the EA.

“Implementing second failovers (failback)” on page 184




tail -f /var/log/storigen_nms

6. Look for output similar to the following:

Oct 28 09:34:39 CUA2 nmslib[619]: [63] [28-Oct-2004 9:34:39.614] [sev 3] [SCAGD] Active CUA (CUA1.centera.lab.emc.com) is [not responding]; pid 6530

Oct 28 09:37:14 CUA2 nmslib[619]: [64] [28-Oct-2004 9:37:14.184] [sev 4] [SCAGD] Active CUA (CUA1.centera.lab.emc.com) is in an [unreachable] state; pid 6530

Only proceed with the failover procedure once you see [unreachable] in the output for the Active system.

7. Perform the procedure in “Initiating failover in a production environment” on page 175.

Initiating failover in a production environment

If the Active system fails, you can transition the Standby system to an active state:

Note: This procedure is the equivalent of step 3 in Figure 15 on page 173.

Note: If failing over an Active CUA/CFA and primary EMC Centera, to a remotely located Standby CUA/CFA and secondary EMC Centera, you must complete the steps in “Using a remote Standby CUA/CFA and EMC Centera” on page 185 before performing the failover.

1. Power down the Active system and disconnect it from the network. If the Active system is running and reachable, the software will refuse to transition the Standby system to an active state.

2. Log in to the Standby system using an SSH client and log in as gwsetup using the password EMCCUA.

3. Select V at the Setup Utility menu to start the failover process.

The Standby CUA/CFA attempts to initialize itself as the Active system. Because the new Active system has a different hostname/IP address than the failed system, you need to change the hostname/IP address of the new Active to the hostname/IP address of the old Active.


176


Note: If the failover completes successfully, the system sends an email notification to the address configured under E-Mail Alert Settings in the service administrative account.

4. Change the hostname/IP address of the new Active system:

a. Select C at the Setup Utility menu to display the Configuration menu.

b. Select C at the Configuration menu to begin the configuration process.

c. Enter the original Active system’s hostname (in FQDN format) and IP Address when prompted.

d. Modify any other specific settings such as the SMTP From and SMTP ReplyTo fields.

e. Press Enter to step through the rest of the options.

f. Reboot the system when prompted.

5. CUA only: If you use CUA with Active Directory, you must also complete the steps in “CUA failover and Active Directory” on page 176.

Note: CUA only: Any files that had not been written from the failed CUA to the EMC Centera at the time of failure will not be available on the newly transitioned system. Refer to “Recovering unwritten data from the old Active CUA” on page 179 for instructions on how to check for and transfer unwritten data.

CUA failover andActive Directory

If you are using CUA with Active Directory Service for CIFS authentication, you need to do the following after failover:


2. Restart the CIFS Daemon by typing the following command:

sudo /etc/init.d/smb restart

3. Adjust the clock skew by typing the following command:

sudo /usr/bin/net time set -S server

Note: Replace server with the name of the Kerberos server.



4. Verify the Active Directory information by typing the following command:

/usr/bin/net ads info

5. Join the Active Directory Domain:

a. Type the following command:

sudo /usr/bin/net ads join -U ADMIN

Note: Replace ADMIN with the name of the Windows NT Domain administrator.

b. Enter the password for the Windows NT Domain administrator.


/usr/bin/wbinfo -p



178


Reconfiguring the old Active systemAfter failover, when the old Active system is powered up, it registers that it is no longer the Active system and carries out the following:

◆ Removes the default TCP/IP gateway address.

◆ Restarts the network.

◆ For CUA: Cleans up /sfs/gateway_cifs and /sfs/gateway_nfs by automatically moving any files that were not yet written to EMC Centera from /sfs/gateway_nfs and /sfs/gateway_cifs to /sfs/serviceinfo/lost+found/.

After the old Active system is again available, you can do the following:

◆ “Recovering unwritten data from the old Active CUA” on page 179 — provides information on transferring unwritten data from the old Active CUA to the new Active CUA. If no data has to be recovered, reinstall the old Active CUA as described in “Installation on a Dell” on page 34 for information on reinstallation.

◆ “Reconfiguring the old system” on page 183 — provides information on reconfiguring the old Active system as the new Standby system.

Note: This procedure contains instructions on changing the IP address and DNS address on the old Active system. You must do this before reconnecting to the network.



Recovering unwritten data from the old Active CUA

Note: This procedure pertains only to the CUA. CFA immediately writes data from the Celerra to the EMC Centera and does not cache files on the CFA system.

At the time of failover, any files that have not been written back to the EMC Centera from the failed CUA are not available on the new Active CUA. After failover, when the old Active CUA is powered up, all files that have not been written back to the EMC Centera are moved to the directory /sfs/serviceinfo/lost+found.

To recover these files from the failed system and place them on the new Active CUA for write back, do the following:

1. Plug in a monitor and keyboard to the back of the failed CUA.

2. Log in to the failed CUA.

3. Reset the default TCP/IP gateway and reconfigure the hostname/IP address of the old Active CUA by doing the following:

a. Select C at the Setup Utility menu to display the Configuration menu.

b. Select C at the Configuration menu to begin the configuration process.

c. When prompted, enter the default TCP/IP gateway address.

d. When prompted, enter the original Standby system’s hostname (in FQDN format) and IP Address.

Note: This step is necessary because as part of the failover, you assigned the failed CUA’s hostname and IP address to the new Active system. You must change the hostname and IP address of the failed CUA before bringing it back on the network.


f. Select S to save the configuration changes.

g. Reboot the system when prompted.

4. Determine if any files need to be recovered:

a. After reboot, log back into the failed CUA.

b. Select X to enter the bash shell.

c. Run the following command:

Reconfiguring the old Active system 179

180


ls /sfs/serviceinfo/lost+found/sfs/

d. If the /sfs/serviceinfo/lost+found/sfs/gateway_nfs directory exists, you need to recover NFS files. Go to Step 5 to recover NFS files.

If the /sfs/serviceinfo/lost+found/sfs/gateway_cifs directory exists, you need to recover CIFS files. Go to Step 6 to recover CIFS files.

e. Exit the bash shell.

5. If you need to recover NFS files:

a. Select C at the Setup Utility menu.

b. Select N to modify the NFS Configuration.

a. Add the following line to /etc/exports:

/sfs/serviceinfo/lost+found/sfs/gateway_nfs <xx.xx.xx.xx>(ro,all_squash,anonuid=0,anongid=0)

where:

<xx.xx.xx.xx> = the CUA’s IP address

Refer to “Transferring the files from the old Active CUA” on page 181 for instructions on moving the files from the old Active to the new Active CUA.

6. If you need to recover CIFS files:

a. Select P at the Setup Utility menu. Change the values to the following:

File :/sfs/serviceinfoUser :gwsetupGroup :gwsetupPermissions :755Recurse down directory tree (Y/N) : Y

b. Select X to enter the bash shell.



c. Create a lost+found directory with the following commands:

mkdir /sf/gateway_cifs/lost+found/ln -s /sfs/serviceinfo/lost+found/sfs/gateway_cifs /sfs/gatway_cifs/lost+found

d. Exit the bash shell and select C at the Setup Utility menu.

e. Select S to modify the CIFS configuration.

f. Add the following to the CIFS configuration file:

[global]security = sharemap to guest = bad userserver string = EMC CIFS Serverworkgroup = EMC_CUA

[lost+found]path = /sfs/gateway_cifs/lost+foundguest account = gateway_cifsforce user = gateway_cifswriteable = noguest ok = yesbrowseable = no

g. Save the file.

h. Run the following command:

sudo /etc/init.d/smb start

Refer to “Transferring the files from the old Active CUA” on page 181 for instructions on moving the files from the old Active to the new Active CUA.

Transferring the filesfrom the old Active

CUA

You can transfer the recovered files from the old Active CUA to the new Active CUA with the following methods:

◆ Transfer via mounted drive or Windows share

◆ Transfer via FTP

Transfer via share or mounted driveFrom a client machine, you need to access both the failed CUA and the new Active CUA and copy the files from the failed CUA to the Active CUA.

Note: The client must be on the same network as the failed CUA and the Active CUA.


182


◆ For CIFS:

1. Create a mapped drive to /lost+found on the failed CUA.

2. Create a mapped drive to /gateway_cifs on the Active CUA.

3. Copy the lost+found files from the failed CUA to the Active CUA.

◆ For NFS:

1. Mount the failed CUA’s /sfs/serviceinfo/lost+found/sfs/gateway_nfs directory.

2. Mount the Active CUA’s /sfs/gateway_nfs directory.

3. Copy the lost+found files from the failed CUA to the Active CUA.

Transfer via FTPWith this method, you tar the lost+found files on the failed CUA and then transfer them to the Active CUA via FTP.

Note: To use this method, FTP access must be enabled on the Active CUA. Refer to “Configuring FTP access” on page 78 for instructions on enabling FTP access.

1. Log into the failed CUA.

2. Select X to enter the bash shell.

3. Navigate to the lost+found directory with the following command:

cd /sfs/serviceinfo/lost+found/sfs/

4. Run the df command to see if /sfs is less than 40%. If /sfs is over 40%, you must transfer the files by mounting/sharing a drive as described in “Transfer via share or mounted drive” on page 181.

5. Tar the gateway_nfs and gateway_cifs directories with these commands:

tar -cvf gateway_nfs.tar ./gateway_nfstar -cvf gateway_cifs.tar ./gateway_cifs

6. FTP the files to the Active CUA with the following commands:

tar -C /sfs -xvf /<path to ftp location>/gateway_nfs.tartar -C /sfs -xvf /<path to ftp location>/gateway_cifs.tar



Reconfiguring the old system

Once the failover is complete, CUA has to be reinstalled on the old Active system before you can reconfigure it as a Standby system. If you want to reconfigure the old Active CUA/CFA as the new Standby system, do the following:

1. Plug in a monitor and keyboard to the back of the system to reconfigure it.

2. Since the Active and the Standby systems need unique hostnames/IP addresses, you must change the hostname/IP address on the old Active before configuring it as a Standby:

Note: If you have already changed the failed CUA’s hostname and IP address as part of “Recovering unwritten data from the old Active CUA” on page 179, you can skip this step.

a. Log in to the new Standby CUA/CFA using an SSH client and log in as gwsetup using the password EMCCUA.

b. Select C at the Setup Utility menu to display the Configuration menu.

c. Select C at the Configuration menu to begin the configuration process.

d. Enter the default TCP/IP gateway, and the original Standby system’s hostname and IP Address when prompted.


f. Reboot the system when prompted.

3. Configure the new EA setup as described in “Configuring the Active CUA/CFA” on page 170.


184


Implementing second failovers (failback)

When the original Active system becomes available again, you may reestablish the original Active/Standby relationship by performing a second failover (also called a failback).

Note: This procedure is the equivalent of step 5 in Figure 15 on page 173.

To execute a second failover to the original Active system, do the following:

1. Follow the procedure in “Testing CUA/CFA failover” on page 174 to failback the new Active system (original Standby) to the new Standby system (original Active).

2. Reinstall the CUA/CFA software on the original Standby system. Refer to “Installation on a Dell” on page 34 for information on installation.

Note: Unless you reinstall the CUA/CFA software on the original Standby system, you will not be able to complete step 3 of this procedure.

3. Re-establish the original Active/Standby relationship by reconfiguring the original Standby system as Standby. Refer to “Configuring the Standby CUA/CFA” on page 171 for instructions.

Note: Make sure both CUA/CFAs are pointing to the same EMC Centera.



Using a remote Standby CUA/CFA and EMC CenteraInitially, both the Active and Standby systems point to the primary EMC Centera as the target. Before failing over to the Standby system, you may need to change which EMC Centera the Standby system targets:

◆ If both the Active CUA/CFA and the primary EMC Centera fail, you need to point the Standby system to the secondary EMC Centera before failing over.

◆ If the Active CUA/CFA fails but not the primary EMC Centera, you can either let the Standby system continue to point to the primary EMC Centera, or you can reconfigure the Standby so that it points to the secondary EMC Centera.

If the secondary EMC Centera is on the same network as the Standby system, you should consider performance and inter-site bandwidth costs when determining which EMC Centera to target. If choose to target the secondary EMC Centera, you loose all disaster recovery protection for the Centera.

Note: If a catastrophic failure occurs to the primary EMC Centera and migrated content has not yet replicated, it could result in partial data unavailability. If you intend to target the secondary EMC Centera during a catastrophic failure to the Active system, the replication queue on the primary EMC Centera should be fully emptied before failing over.

Figure 16 on page 186 shows the failover of a remote CUA/CFA and a remote EMC Centera. One arrow shows the path of communications if only the CUA/CFA fails and the other arrow shows the path of communication if both the CUA/CFA and the primary EMC Centera failed.

Using a remote Standby CUA/CFA and EMC Centera 185

186


Figure 16 Failover of remote Standby system with remote EMC Centera

Centera

FailedCFA

Celerra

`

Client Client

CenteraCentera

Failed CUA

`

Client Client

(EA: Full site failure)

Centera

Standby CUA

(EA: C

UA

only

failu

re)

Remote standby CFA with remote EMC Centera

Remote standby CUA with remote EMC Centera

During failover

(EA: Full site failure)

ActiveCFA

(EA: C

FA

only

fail ur

e)

During failover



Failing over to a Standby CUA/CFA and a secondary EMC CenteraIf you choose to change the target of the Standby system, you must do the following before initiating failover:

1. Log in to the Active CUA/CFA using an SSH client and log in as gwsetup using the password EMCCUA.

2. If the Active CUA/CFA is unreachable, continue to step 3. If you are testing failover and the Active system is still accessible, follow the procedure in “Testing CUA/CFA failover” on page 174 to shut down the Active system. Do not proceed with the failover procedure described in “Initiating failover in a production environment” on page 175.

3. On the Standby system, change the hostname/IP address of the EMC Centera:

a. Select C to run the configuration procedure.

b. Run through the procedure accepting the defaults until you see the following prompt.

Centera Names or IP address

c. At this prompt, enter the IP addresses of the secondary EMC Centera.

d. Continue through the options, accepting the defaults and save the configuration.

After saving the configuration, the system reboots.

4. Initiate the failover per the instructions in “Resolving an Active CUA/CFA failure” on page 173.


188


Swapping your primary site with your remote siteIf you want to swap your primary site (CUA/CFA and EMC Centera) with your secondary site, you need to break the EA relationship between the Active system and the Standby system, and then point both systems at the EMC Centera at the secondary site.

1. Log in to the Active CUA/CFA using an SSH client and log in as gwsetup using the password EMCCUA.

2. Break the EA relationship:

a. On the Active system, select C at the Setup Utility menu to enter the Configuration menu.

b. Select V at the Configuration menu. Refer to the following table for the prompts and the values you should enter.

c. Connect to the Standby system using an SSH client.

d. Select C at the Setup Utility menu to enter the Configuration menu.

e. Select V at the Configuration menu. Refer to the following table for the prompts and the values you should enter.

Prompt Enter


Y


Y


N

Prompt Enter


Y


Y


N



3. Since both the Active and the Standby systems point to the primary EMC Centera, you need to point both systems to the secondary EMC Centera:

a. On the Active system, select C to run the configuration procedure.



c. At this prompt, enter the IP address of the secondary EMC Centera.

Note: For CFA you must use the hostname (in FQDN format) of the secondary EMC Centera.



e. Repeat steps a - d on the Standby system.

4. Reverse the EA relationship between the two systems by changing the Active system to Standby and the Standby system to Active. Refer to “Configuring enhanced availability” on page 169 for instructions.

Note: If a catastrophic failure occurs to the primary EMC Centera and migrated content has not yet replicated, it could result in partial data unavailability. If you intend to target the secondary EMC Centera during a catastrophic failure to the Active system, the replication queue on the primary EMC Centera should be fully emptied before failing over.


190


Using a remote Standby CFA, EMC Centera, and CelerraThis configuration provides the most comprehensive disaster recovery protection for CFA since all devices are protected. All the standby components are located remotely to provide for cross site protection. This configuration is the most complex to deploy and also requires a high speed network between sites.

For this configuration, all migrated content on the replicated file system on the standby Celerra can be accessed by clients; however, no modifications or policies can be run against that particular file system since it is read-only. Figure 17 on page 191 shows the setup of this configuration.

For failover scenarios, you can either have full-site failover, or a partial failover of CFA, EMC Centera, or Celerra. Refer to “Full Site CFA Failover” on page 193 for full site failover. Refer to “Single component CFA failover” on page 194 for the failover procedure for one of the components.



Figure 17 EA configuration with remote standby CFA, remote EMC Centera, and remote Celerra

Centera

ActiveCFA

Celerra

`

Client Client

Centera

Standby CFA

Celerra

Stub files

Data

Logs

Full s ite

fai lover

FailedCentera

FailedCFA

FailedCelerra

`

Client Client

Centera

ActiveCFA

Celerra

CFA full site EA configuration

CFA full site failover

Using a remote Standby CFA, EMC Centera, and Celerra 191

192


System requirements for full site CFA standbyThis configuration requires the following:

◆ Two Celerras running DART OS v5.4 or greater.

◆ Two EMC Centeras.

◆ Two CFA systems. Both of the CFA systems in the configuration must be running on the same hardware family and class (in other words, a Dell 2850 Active must use Dell 2850 Standby, and a Gen4 Active must use a Gen4 Standby). The CFA systems must be running CFA v 4.0 or greater.

◆ Centera Replication License.

◆ Celerra Replicator.

◆ A network with cross site network communication.

Note: CFA can use HTTP instead of NFS for file stubbing on the Celerra. In an EA configuration, HTTP stubbing along with DNS round robin, allows the Celerra to retrieve data from either the Active or Standby system. This read redundancy eliminates the need for failover if the Active CFA is unavailable. It is strongly recommended that you use HTTP stubbing and DNS round robin for your CFA EA configurations. Refer to “Using DNS round robin with CFA” on page 195 for a description of this type of configuration.

Failover for full remote CFA configuration

In a full remote EA configuration where all components are replicated, there are a few failure scenarios to consider:

◆ Full site failure — “Full Site CFA Failover” on page 193

◆ Single component failure — “Single component CFA failover” on page 194

◆ Recovery protection



Full Site CFA Failover In this scenario, the primary Celerra, primary EMC Centera, and the Active CFA are all down.

1. Log in to the Active CFA using an SSH client and log in as gwsetup using the password EMCCUA.

2. On the Standby CFA, change the hostname of the EMC Centera:

a. Select C to run the configuration procedure.



c. At this prompt, enter the hostname (in FQDN format) of the secondary EMC Centera.



3. Initiate the failover per the instructions in “Resolving an Active CUA/CFA failure” on page 173.

4. Initiate failover on the secondary Celerra to turn on the write capabilities for the file system. Refer to Celerra FileMover documentation for failover instructions. No new FileMover connections need to be established if DNS is used.

Data loss warning Depending on the state of the secondary Celerra and EMC Centera, data could be lost if it was not replicated. The worst case scenario is if data was not replicated from the primary Celerra to the secondary Celerra, or if a migration was done to a file on the primary Celerra, the stub is successfully replicated to the secondary Celerra, and the primary EMC Centera has not replicated its data to the secondary when a catastrophic disaster occurs.

Using a remote Standby CFA, EMC Centera, and Celerra 193

194


Single componentCFA failover

The following sections tell what to do in the event of the failure of any component of the CFA system.

Active CFA FailoverInitiate failover on the Standby CFA and adjust the DNS entry to associate the DNS address with the new Active CFA's IP address. Refer to “Initiating failover in a production environment” on page 175 for instructions on failing over CFA.

Primary Celerra FailoverIf the primary Celerra fails, failover should be initiated on the secondary Celerra. Users can continue to connect to the Active CFA and primary Centera. Refer to Celerra FileMover documentation for failover instructions.

Data loss warning Any data that has not been replicated from primary to secondary Celerra before the failure is lost.

Primary EMC Centera FailoverIf the primary EMC Centera fails, reconfigure the Active CFA's target EMC Centera to point to the secondary EMC Centera:

1. Log in to the Active CFA using an SSH client and log in as gwsetup using the password EMCCUA.

2. Since the Active CFA points to the failed primary EMC Centera, you need to point the Active CFA to the secondary EMC Centera:

a. On the Active CFA, select C to run the configuration procedure.



c. At this prompt, enter the hostname (in FQDN format) of the secondary EMC Centera.



Note: Any data that hasn't been replicated from the primary Centera to the secondary Centera is lost in the event of a catastrophic failure.



Using DNS round robin with CFAThis version of CFA gives you the option of using DNS round robin for resolving the IP addresses of the Active and Standby CFA systems. With DNS round robin, a single DNS name points to two CFA systems. For example, cfa.xxx.com would point to the systems 10.0.0.10 and 10.0.0.11.

Figure 18 CFA Reads with DNS Round Robin

DNS round robin is valuable in EA configurations because it allows reads from either CFA system. This read redundancy eliminates the need for failover if the Active CFA is unavailable. If the Active CFA becomes unavailable, the Celerra automatically contacts the Standby CFA for data.

Note: The read redundancy provided by DNS round robin only effects data retrieval. If you want to migrate files with the Standby CFA, you must failover the Standby CFA to Active status. Refer to “Initiating failover in a production environment” on page 175 for instructions on converting a Standby CFA to an Active CFA.

DNS Round Robin Entriescfa.xxx.com IN A 10.0.0.10cfa.xxx.com IN A 10.0.0.11

CenteraCelerra

DNS Server

cfa.xxx.com

Active CFA10.0.0.10

Standby CFA10.0.0.11

Active

connection

Connection during failure of Active

Using DNS round robin with CFA 195

196


This feature requires the use of HTTP for stubbing files on the Celerra rather than NFS. This option is only available if:

◆ All data movers used in the CFA configuration are Celerra version 5.5.27-5 and greater Data Movers.

◆ An HTTP connection is used for the connection from the Celerra to the CFA.

Note: It is important to understand that DNS round robin provides read continuity if the Active CFA fails but does not provide load balancing. IP addresses are cached on the Celerra until they become non-responsive; at that time, the Celerra contacts the DNS server for a new IP address.

Enabling DNS round robin

To enable HTTP stubbing and DNS round robin, do the following:

1. Select L at the EMC Centera FileArchiver Configuration menu to enable DNS round robin. Refer to the following table for system prompts and the values you should enter.

Note: Wait about 5 minutes before establishing the HTTP connection. It takes approximately this long for the updated CFA configuration file to be copied to the Standby CFA.

2. Establish the HTTP connection from the Celerra to the CFA:

a. In the FileArchiver configuration file, define at least one file system connection to the Data Mover from which you are enabling HTTP connections. Refer to “Defining file systems” on page 139 for information.

Note: For an EA configuration, wait a short period before performing Step b. It may take a up to a minute for the new configuration to be transferred to the standby CFA.

Prompt Enter

Is the Centera FileArchiver in a DNS domain?

Y

Use DNS round robin name when stubbing files?: N

Y

DNS round robin name: Enter the full DNS round robin name and press Enter.



b. From the Celerra CLI, enter the following command:

fs_dhsm -connection file_system_name -create -type http -read_policy_overrride passthrough -secondary http://DNS_RR_hostname -cgi n

where:filesystemname = the filesystem on the Celerra with FileMover enabled.DNS_RR_hostname = the FQDN of the DNS round robin entry that resolves to the IP addresses of the Active and Standby CFAs.

Note: You must use the -read_policy_override passthrough option. CFA may not function properly without this option.

3. Convert all the NFS stubbed files to HTTP stubs:

IMPORTANT!This is an irreversible action. Once you convert the stubs to HTTP, they cannot be reconverted to NFS.

a. Select D at the Setup Utility menu to enter the EMC Centera FileArchiver Configuration menu.

b. Select O to convert the Celerra stubs to HTTP format. The following message appears:

WARNING: Once converted to use HTTP for secondary file access the stubs can not be converted back to NFS.

Do you want to continue? (Y/N)

c. Enter Y and press Enter.

Note: If you change the round robin name or a node name, you must rerun the HTTP conversion.

Using DNS round robin with CFA 197

198



6Invisible Body Tag

This chapter describes how to log in to the EMC Centera Universal Access (CUA) Graphical User Interface (GUI), monitor system status, view event messages, and change the GUI user password. This section also describes the different ways in which file sizes are represented on CUA.


◆ Storage presentation ........................................................................ 200◆ Accessing the GUI............................................................................ 201◆ Monitoring system status ............................................................... 203◆ Getting detailed node status........................................................... 208◆ Monitoring services ......................................................................... 215◆ Monitoring a standby CUA ............................................................ 216◆ Viewing event messages ................................................................. 220◆ Disk monitoring ............................................................................... 223◆ Changing the GUI password.......................................................... 224

Monitoring CUA Status

Monitoring CUA Status 199

200


Storage presentationTo verify file size and storage utilization on the CUA, use the following guidelines when accessing the information from different sources.

From a NFS mount point

The Linux command du returns the sum of the sizes for all files currently available on the CUA at the specified directory level.

The command ls returns the size of the file independent of how much data is cached on the CUA.

The command df returns statistics related to the cached data, which is the storage utilization for the CUA’s disks.

The command df -i returns the number of available, used, and free i-nodes on the CUA. This information is also available through the GUI on the Monitor => Node Detail page (refer to “Accessing the GUI” on page 201 for more information).

From a CIFS share Windows Explorer shows file size independent of how much data is cached on the CUA. This presentation is similar to the NFS ls command.

The properties size figure for a share in Windows Explorer is equivalent to the NFS df command.

Using the Monitor GUI

The percentage utilized, free disk space in Gigabytes, and total capacity in Gigabytes is presented both for the CUA’s local disk cache and the EMC Centera cluster similar to the NFS df command.



Accessing the GUIAccess the CUA GUI through a standard web browser. Use the following URL, which uses the HTTPS (HTTP Secure) protocol:

https://machine:7227/

Replace machine with the host name or IP address of CUA.

Example The following example connects to a CUA named demo1:

https://demo1:7227/

Once connected to the CUA GUI, the browser displays the Login page shown in Figure 19 on page 201.

Figure 19 Login page

Enter a valid user name and password for the system monitor account, then click Login. The default account is named gwsetup and has the default password, EMCCUA.

Note: Username and password entries are case-sensitive. Type them exactly as shown.

Accessing the GUI 201

202


After logging in, a page similar to the one shown in Figure 20 on page 202 appears. This is the Home Page of the CUA GUI.

Figure 20 Home page



Monitoring system statusThe CUA GUI provides access to information that gives useful data on the state of the system and its capabilities and statistics on the individual accounts and subaccounts.

Under the Monitor tab:

◆ Node Summary — gives an overall summary of system health.

◆ Node Detail — gives a detailed picture of individual system components.

◆ Event Log — provides access to system information, errors, and warnings.

◆ Services — displays all installed services, corresponding states and ports.

The Node Summary page shows the high-level current status of the CUA. This page updates itself periodically, as specified in the Refresh pull-down menu.

Note: The Refresh menu does not appear for newer browsers that can update their displays without a complete page reload; these browsers update the data only every five seconds.

The following path reaches this page:

Monitor ⇒ Node Summary

Monitoring system status 203

204


Figure 21 Node summary

Field descriptions Following is a description of the fields on this page:

SubsystemThis is the subsystem, or system element being summarized.

StateThis is the overall state of the subsystem:

◆ NORMAL (green icon) indicates good health.

◆ DEGRADED (yellow icon) indicates that one or more of the installed services is either reporting non-fatal errors or has FAILED.

◆ FAILED (red icon) indicates the service is down.

◆ UNKNOWN indicates that the service status is not reaching the UI.

◆ UNMOUNTED (for File System and Storage only) indicates that the SL/FS is not mounted.



These subsystems are DEGRADED when they meet or exceed certain thresholds.

The default values are:

◆ 95 % CPU usage

◆ 90 % Memory usage

◆ Packets/dropped — for every 5 one dropped

◆ Packets/collisions — for every 5 there is one collision

◆ Packets/send errors — for every 5 there is one send error

◆ Packets/receive errors — for every 5 there is one receive error

◆ File system usage — threshold is 97% on /var, /sfs, /, and /usr

MetricsThis shows relevant metrics for the subsystem.

The Metrics are described below for each subsystem:

Services

This is the group of processes responsible for data delivery to clients.

◆ Web — Shows the status of the web server. The status can either be NORMAL or FAILED.

◆ Stream — Shows the status of the streaming delivery application. The status can be NORMAL, FAILED, or it can be DEGRADED meaning that the client can connect to the port but the service is off.

◆ File — Shows the status of the file delivery application (CIFS and NFS). The status can either be NORMAL or FAILED.

◆ Other — Shows the status of CUA to the EMC Centera. The status can either be NORMAL or FAILED.

Network

This is the connection to the IP network.

◆ Send BW (Mbps) — The total amount of bandwidth transmitted over the network interface(s), in megabits per second (Mbps). One megabit per second is 1,000,000 bits per second. This bandwidth includes overhead for IP and other networking protocols, so it is always greater than the bandwidth reported for the File System.


206


◆ Receive BW (Mbps) — The total amount of bandwidth received into the network interface(s).

File system

This is the Storigen File System, which holds and serves the file data.

◆ Read

• Bandwidth (Mbps) is the current rate of read operations at the file system. One megabit per second (Mbps) is 1,000,000 bits per second.

• Data (GB) is the total amount of data (in Gigabytes) read from the file system by end-users.

• Active is the number of file streams currently open for reading.

◆ Write

• Bandwidth (Mbps) is the current rate of write operations onto the file system, in Mbps.

• Data (GB) is the total amount of data (in Gigabytes) read from the file system by end-users. One GigaByte (GB) is 1,073,741,824 bytes (UNIX). Note that there is another computation for one GB: when computing disk hardware, one GB is 1,000,000,000 bytes.

• Active is the number of file streams currently open for writing.

◆ Rejects — The percentage of total file requests that the CUA rejected based on its quality of service settings. Rejects are caused by oversubscription of account bandwidth.

Computing

These are the CPUs in the current node.

◆ CPU Util — Are the percentages for CPU used on this node:

• Total is the total usage, combining the next two values.

• System is the percentage used for Linux kernel processing (for example, file retrieval from the SL/FS software).

• User is the percentage used for non-kernel processing, such as the HTTP server processing an end-user file request.

◆ CPUs — The number of CPUs configured for this node.



Note: A Dell 2650/2850 contains 2 physical CPUs but, due to hyperthreading technology, you may see 4 logical CPUs listed. A Dell 2950 has a dual core processor and will appear as 2 logical CPUs. Gen3 and Gen4 nodes also appear as 2 logical CPUs.

Memory

This is the machine memory.

◆ Util — The percentage of memory currently in use.

◆ Free — The free memory in Megabytes (MB).

◆ Total — The total memory in MB.

◆ Swap Util — The percentage of swap space currently in use.

◆ Swap Total —The total swap space available in MB.

Cached content

This is the disk space for CUA.

◆ Util — The percentage of disk storage used on the CUA.

◆ Free (GB) — The number of available Gigabytes on the CUA.

◆ Total (GB) — The total storage on the CUA.

◆ Disks — The number of disks on this node.


208


Getting detailed node statusUse the Node Detail page to see a detailed display of the current node’s status. The statistics on this page are static but you can refresh them at any time by clicking Refresh Now in the upper-right corner of the page.

Note: The Node Detail page uses the term Elapsed Time instead of the more common Uptime to indicate how long the system has been functional.


Monitor ⇒ Node Detail

Figure 22 Node detail



Field descriptions Following is a description of the fields on this page.

SubsystemThis is the subsystem, or system element being summarized.

State

This is the overall state of the subsystem in this node:

◆ NORMAL (green icon) indicates good health.

◆ DEGRADED (yellow icon) indicates that one or more of the installed services is either reporting non-fatal errors or has FAILED.

◆ FAILED (red icon) indicates the service is down.

◆ UNKNOWN indicates that the service status is not reaching the UI.

◆ UNMOUNTED (for File System and Storage only) indicates that the SL/FS is not mounted.

These subsystems are DEGRADED when they meet or exceed certain thresholds.

The default values are:

◆ 95 % CPU usage

◆ 90 % Memory usage

◆ Packets/dropped - for every 5 one dropped

◆ Packets/collisions - for every 5 there is one collision

◆ Packets/send errors - for every 5 there is one send error

◆ Packets/receive errors - for every 5 there is one receive error

◆ File system usage - threshold is 97% on /var, /sfs, /, and /usr

Status

Show the current rates, cumulative performance over time, and analytical calculations for the subsystem.

Click Refresh Now to refresh the page immediately. The time of the most-recent refresh is noted at the bottom of the page.

Click Disk Detail to see a map of the disks in the node along with the status of each disk.

The subsections below describe the data for each subsystem.

Getting detailed node status 209

210


Services

This is the group of processes responsible for data delivery.

◆ Web — Shows the status of the web server. The status can either be NORMAL or FAILED.

◆ Stream — Shows the status of the streaming delivery application. The status can be NORMAL, FAILED, or it can be DEGRADED meaning that the client can connect to the port but the service is off.

◆ File — Shows the status of the file delivery application (CIFS and NFS). The status can either be NORMAL or FAILED.

◆ Other — Shows the status of CUA to the EMC Centera. The status can either be NORMAL or FAILED.

Network

Is the node’s connection to the IP network.

◆ Bandwidth (Mbps) — Shows the combined bandwidth used on all the node’s network interfaces. The bandwidth is measured in megabits (1,000,000 bits) per second. This is the bandwidth going over the network interface(s), as opposed to the bandwidth coming off of the SL/FS™ file system.

• Send is the bandwidth sent out to end users.

• Recv is the bandwidth received from providers.

◆ Packets/sec — Shows network bandwidth in packets per second:

• Send is the bandwidth sent out to end users.

• Recv is the bandwidth received from providers.

◆ Delivered (MB) — Shows total MegaBytes (MB) of data that have traveled through the node’s network interface(s).

• Send is the data sent out to end users.

• Recv is the data received from providers.

◆ Analysis — Shows details for each interface in the node. Each interface is shown in one row, with the following data columns:

• Name is the name of the interface.

• Send: Errors is the number of send errors logged by this interface.

• Send: Dropped is the number of packets dropped by the interface instead of sent.



• Send: BW (Mbps) is the bandwidth being used to send data out over the interface.

• Recv: Errors is the number of receive errors logged by this interface.

• Recv: Dropped is the number of packets dropped by the interface after being received.

• Recv: BW (Mbps) is the bandwidth being used to receive packets at the interface.

File system

This is the Service-Level File System (SL/FS™), which holds and serves the file data. These measures apply to SL/FS activity on the local node only.

◆ Bandwidth (Mbps) — Shows the rate of file-system traffic in megabits (1,000,000-bit) per second. This is the bandwidth going through the SL/FS software, as opposed to the bandwidth going through the HTTP servers or the network interfaces.

• Send is the bandwidth sent out to end users from this node’s SL/FS software.

• Recv is the bandwidth received from providers to this node’s SL/FS software.

• These numbers should be roughly equal to the sum of the Cache Bandwidth (under Memory) and the Bandwidth under Storage.

◆ Delivered (MB) — Shows the megabytes of data transferred onto and off of this node’s SL/FS.

• Send is the megabytes sent out to end users from this node’s SL/FS.

• Recv is the megabytes received into this node’s SL/FS.

◆ Active Read — This is the number of currently active read streams on this node’s SL/FS.

◆ Active Write — This is the number of currently active write streams on this node’s SL/FS.

◆ Read Requests — This is the total number of read requests received by SL/FS software on this node.

◆ Write Requests — This is the total number of write requests received by SL/FS software on this node.


212


◆ Open Files — This is the number of currently open files on this node’s SL/FS.

◆ Rejects — This is the percentage of total file requests that the CUA rejected. Rejects are caused by oversubscription of account bandwidth by end-users.

◆ Avg File Size — This is the average size of files residing on this node’s SL/FS, in blocks. A block is a single unit of disk-storage space. The size of a block appears in the Block Size field under Storage.

◆ Total Inodes — Is the total number of i-node entries available on the Storigen file system.

◆ In use Inodes — Is the number of i-node entries currently being used by the file system.

Enhanced availability

If the Enhanced Availability option is configured this section provides information on the CUA Type (Active or Standby), its node name, and state. The state can be any of the following:

◆ restore — the Standby CUA is being configured or restoring itself from an Active CUA backup

◆ init — system is coming up

◆ ready — Standby CUA is ready to assume the active role

◆ reachable — the system is responding to healthcheck heartbeat messages

◆ suspend— the system is about to fail over

◆ not responding — heartbeats missed but the system is not yet considered unreachable

◆ unreachable — system is no longer reachable



Computing

This is the hardware for this node.

Utilization — Are the percentage statistics for current CPU utilization:

◆ Free percentage of CPU cycles.

◆ Used CPU percentage. Used is divided into two classes of CPU user:

• Sys is the CPU used by system-overhead processes (the kernel, including kernel calls from non-kernel programs).

• User is the CPU used by application processes such as the HTTP server and the UI.

Note: A Dell 2650 or 2850 contains 2 physical CPUs but you may see 4 logical CPUs listed. This is due to hyperthreading technology.

◆ Usage (cpu-sec) — Are the total CPU times for each of the Utilization categories. These are measured in seconds of the CPU’s time. They measure CPU use since the time of the last system reboot.

◆ Analysis — Midplane Temp. Readings (F). This is the temperature of the node in Fahrenheit. The remaining fields show details for each CPU in the node. Each CPU is in its own row, with the following data columns:

• CPU is the name and type of CPU.

• Util is the percentage of this CPU being used.

Memory

This is s the machine memory.

◆ Cache Bandwidth (Mbps) — Shows the rate of disk-cache memory usage for this node. The rate is measured in megabits (1,000,000-bit) per second. These numbers are usually lower than the Storage rates below because the CUA generally bypasses disk-cache for improved performance.

• Send is the bandwidth sent out to end users from disk cache.

• Recv is the bandwidth received into disk cache from the SL/FS software.


214


◆ Cache Delivered (MB) — Shows the data transferred through this node’s disk cache. The data is measured in MegaBytes (MB).

• Send is the total MB sent out to end users from disk cache.

• Recv is the total MB received into disk cache from the SL/FS software.

◆ In Use — Is MB of memory currently being used on this node. This includes all types of memory in use, including the disk cache described above.

◆ Mem Free — Shows the node’s available memory in two forms: percentage free and MB free. This is total free memory, including the disk cache described above.

◆ Mem Cached — Is the MB of memory used for this node’s various caches (disk cache, page cache, and other memory buffers).

Storage

This represents the hard disks of CUA.

◆ Bandwidth (Mbps) — Shows the rate of disk traffic for this node. The disk traffic is measured in megabits (1,000,000-bit chunks) per second. This is the bandwidth going to and from the hard disks themselves, as opposed to the bandwidth going through the network interface.

• Send is the total bandwidth sent out to end users from this node’s disks.

• Recv is the total bandwidth placed onto this node’s disks. Note that each file is mirrored to two disks.

◆ Delivered (MB) — Shows the MegaBytes (1,048,576-byte chunks) of data transferred onto and off of the node’s disks.

• Send is the total MegaBytes sent out to end users from disk.

• Recv is the total MegaBytes received into disk.

◆ Block Size — Is the size of each disk block. This unit is used in other fields on this screen.



Monitoring servicesUse the Services page to show the status of applications and services on the server.


Monitor ⇒ Services

Figure 23 Monitor services

Field descriptions The following is a description of the fields on this page:

CategoryThis is the type of installed service.

Service NameThis is the name of the CUA service.

StateThis provides an indication of the current condition of the service, which can be: NORMAL (green bar), FAILED (red six-sided icon), or DEGRADED, (yellow triangle).

PortsThis is the ID of each server port currently in use.

Monitoring services 215

216


Monitoring a standby CUAThere are two ways to monitor the status of a Standby CUA:

◆ Using the CUA GUI.

◆ Using the CUA Setup Utility.

Monitoring a standbyCUA using the GUI

Use your browser to connect to the GUI on the Standby CUA and log in using the sysmon account.


Monitor ⇒ Node Detail

Figure 24 Node detail page

Monitoring a standbyCUA using the Setup

Utility

You can also browse the Standby CUA and view its file list log by using the Setup Utility. Refer to “Advanced Configuration” on page 53 for complete instructions.

1. Connect to the CUA Setup Utility using an SSH client.



2. Select option B at the Setup Utility menu. The following output appears:

Browsing Metadata

Enter a file or directory name to browse (<ENTER> to return to main menu)?

3. Enter the path name of the file or directory you want to browse.

Note: You can only browse the metadata of the subdirectories of /sfs.

The following output appears:

Browsing Metadata

Enter a file or directory name? /sfs/logs

drwxrwxr-x 0 gwsetup gwsetup 1024 30-May-2007 23:10:20 /sfs/logsdrwxrwxr-x 0 gwsetup gwsetup 1024 30-May-2007 23:21:24 /sfs/logs/fileArchiverdrwxrwxr-x 0 gwsetup gwsetup 1024 30-May-2007 23:21:24 /sfs/logs/fileArchiver/migrateddrwxrwxr-x 0 gwsetup gwsetup 1024 30-May-2007 23:21:25 /sfs/logs/fileArchiver/migrated/1-rwxrwxr-x 0 gwsetup gwsetup 5155 30-May-2007 23:21:25 /sfs/logs/fileArchiver/migrated/1/fileArchiver_migrated-2007-05-30-23-21-25.log-rwxrwxr-x 0 gwsetup gwsetup 24664 30-May-2007 23:21:26 /sfs/logs/fileArchiver/migrated/.seqdrwxrwxr-x 0 gwsetup gwsetup 1024 30-May-2007 23:10:20 /sfs/logs/writebackdrwxrwxr-x 0 gwsetup gwsetup 1024 30-May-2007 23:10:21 /sfs/logs/writeback/1-rwxrwxr-x 0 gwsetup gwsetup 155 30-May-2007 23:10:21 /sfs/logs/writeback/1/writeback-2007-05-30-23-10-21.log-rwxrwxr-x 0 gwsetup gwsetup 24664 30-May-2007 23:10:20 /sfs/logs/writeback/.seq Browse another directory? (default = N):

Note: The output for a path name, such as /sfs/logs shown above, is a recursive listing of all files in the directory and subdirectories. The output may fill many screens. Use the space bar to go to the next screen. Use Ctrl-C at any time to exit.

Monitoring a standby CUA 217

218


Viewing writeback audit trailsCUA contains an audit log of the files that have been written to EMC Centera and when the write occurred. This ’audit trail’ gives applications a way to verify that files have been written to EMC Centera by CUA. This enables applications to make informed decisions regarding the disposition of the content still managed by the application.

Log files are themselves written to EMC Centera when they reach 5MB in size and a new log file is created. A system reboot also causes a new log file to be created. The log files may be obtained by either logging in to the CUA and using the Setup Utility menu or by HTTP access to the CUA.

The log files are written to the directory /sfs/logs/writeback. The names of the log file conform to the following rule:

writeback + date + .log


When the number of log files reaches 200, the files will be put into a different subdirectory starting with the name of 1 and incrementing by 1 for each new subdirectory.

Example /sfs/logs/writeback/1/writeback-2004-12-20-10-37.log

The contents of the file are a space separated list of the following data:

◆ Timestamp

◆ Filename

◆ Content Address

◆ File size

◆ Boolean specifying if blob was updated

◆ Boolean specifying if clip was updated

◆ Boolean specifying if retention was updated

◆ The number of days that retention is set



◆ Retention class

◆ Access profile name

By default, CUA allows HTTP access to the log files:

Example http://server/logs/writeback/1/writeback-2004-12-20-10-37.log

By default, the /logs directory is the only directory that is accessible by HTTP. The administrator may modify the /etc/apache2/vhosts.d/cua_httpd.conf file to allow access to other directories. If required, the administrator may also modify the cua_httpd.conf to turn off access to the log file directory.

The CUA will not automatically delete these logs, as they are there for auditing purposes. The gwsetup account has the rights to delete them if desired.

Viewing writeback audit trails 219

220


Viewing event messagesThe Event Messages page is used to view the events being generated by software on the CUA node.


Monitor ⇒ Event Log

Figure 25 Event log

Event messages are stored on disk in /var/log/messages and rotated weekly, each Monday in the early morning. Rotating the logs on disk has no effect on the Event Log Viewer. You can view the most-recent 100 events, or you can view all events generated in a specific range of time.

Click the Most Recent 100 button to see the most recent 100 events. Use the From and To fields to choose a time range, then click Submit to see all events generated between those two times. Click Reset to return to the times set when you reached this screen.



Field descriptions Following is a description of the fields on this page.

TimeThis is the exact time and date of the event.

CodeThis identifies the event message. Each Source has a unique set of event message codes.

SeverityThis indicates the severity of the event. This shows one of three settings:

◆ information

◆ warning

◆ error

SourceThis is the software component that generated the message.

MessageThis is the event message text.

Use the Most Recent 100 button to refresh the event messages in the table and display the most recent 100 event messages. The note under the table indicates the most-recent refresh time.

Table Table 20 on page 221 provides a description of the acronyms that may appear in the Source column.

Table 20 Source column acronyms (1 of 2)

Source Identification

SCSI SCSI disk controller

SYSTEM System event

NIC Network Interface Card

SEVL System Event Logger

SFS Storigen File System

SBA Block Access

FLTD Fault Detector Daemon

Viewing event messages 221

222


NMS Network Management System

DC Data Collector which monitors the system’s bandwidth

FD Fault Detector

SRMGRD Recovery Manager Daemon

SFSRD Storigen File System Recovery Daemon — responsible for drive replacement, drive rebuild, mirror repair and file restriping (future)

SFSM Storigen File System Management Library

CRASH Crash utility which copies crash dumps during system reboot

LBA-SCSI Local Block Access (SCSI drives)

LBA Local Block Access

SAFTE SCSI Accessed Fault Tolerant Enclosures — used to monitor power supplies, fans, and disk subsystem.

ESM Data Replication

SFSPROXYD File Proxy

XMD Extended Metadata

Table 20 Source column acronyms (2 of 2)

Source Identification



Disk monitoringThe Repair page contains applications which allow you to monitor disk repair.


⇒ Repair

Figure 26 Monitoring disk repair

The Status field for each disk indicates the current state.

◆ Active — The disk is operational.

◆ Offline — The disk is currently offline and needs to be replaced.

◆ Disk Not Detected — The disk for this slot has not been detected.

◆ Recovery In Progress — The disk is currently being re-mirrored. You can monitor the percentage complete by looking at the Active Disk Actions on the right half of the page. You will see this status on both the drive being re-mirrored to and the drive being mirrored from.

◆ User Action Required — This drive is healthy but the system partitions for this drive have lost their mirror. This indicates that this drive is currently the only drive containing some of the system partitions.

Disk monitoring 223

224


Changing the GUI passwordUse the Change Password page to change the password for the current user account.

Note: The Change Password screen on the GUI changes the password for access to the GUI, but does not change it for shell access through SSH.


Password ⇒ Change Password

Figure 27 Changing an account password

Enter the new password twice and then click Change Password.

New passwordThis is a case-sensitive password for the current user account. The password is masked as you type it.

For increased security, we recommend a minimum password length of six characters. We also recommend combining numbers, mixed-case letters, and punctuation (such as “.”, “-”, and “_”) in your passwords.



Re-type passwordConfirms the password. Type the same text you typed above. This is also masked as you type it.

Changing the GUI password 225

226



7

This chapter will show you how to upgrade CUA, replace a disk drive, replace an EMC Centera node, and connect to CUA with a modem.

The major sections are:

◆ Upgrading CUA on a Dell .............................................................. 228◆ Upgrading CUA on an EMC Centera node ................................. 230◆ Upgrade error messages ................................................................. 231◆ Changing CUA hardware ............................................................... 233◆ Replacing disk drives and nodes................................................... 235◆ Connecting to CUA remotely......................................................... 240

Servicing the CUA

Servicing the CUA 227

228

Servicing the CUA

Upgrading CUA on a DellThis upgrade procedure requires that the user is familiar with EMC Centera Universal Access (CUA), the Linux operating system, and has unrestricted access to the Dell. This upgrade requires either SSH access or a monitor and keyboard attached to the CUA.

Notes◆ In an Enhanced Availability environment, the Standby CUA must

be upgraded first.

◆ Refer to the customer’s Dell documentation for detailed information on the Dell PowerEdge 2650, Dell PowerEdge 2850, and Dell PowerEdge 2950.

Note: Dell 2650s with 73 GB drives are no longer supported. Dell 2650s with 146 GB drives are only supported for fresh installations.

◆ If you changed the default passwords for any account and then upgraded to a later version of CUA, passwords are set to the default values for the new version.

◆ After an upgrade, it is very important that you verify that the CUA/CFA can send email alerts to the EMC Customer Service Center. Follow the procedure in “Testing email alerts” on page 97 to test email home functionality.

Upgrading a Dell 2650 from v3.6 SP1 to v4.0.1

Upgrades on Dell 2650s with 146 GB drives are not supported. In order to upgrade a Dell 2650, you must backup your data, perform a fresh installation of CUA v4.0.1, and then restore your data. For back up and restore information, refer to “Restoring a CUA configuration” on page 81.

Upgrading a Dell 2850 from CUA v3.6 SP1 to 4.0.1

Upgrade a Dell 2850 from CUA version 4.0.1, by doing the following:

1. Log in to the CUA as user gwsetup using the password EMCCUA.

2. Select U at the Setup Utility menu to enter the Upgrade procedure.

3. Insert the CUA/CFA CD into the CD drive of the CUA.


Servicing the CUA

4. Select C to begin the upgrade procedure.

The upgrade script will automatically check the status of the system before attempting the upgrade. The script will check for things such as the appropriate version of CUA, the existence of backups, and available disk space.

For EA configurations, the system checks to see if the Standby system has already been upgraded.

Refer to “Upgrade error messages” on page 231 for a list of possible upgrade error messages and their solutions. All upgrade status information is displayed on the screen.

The system will reboot before the upgrade completes. Monitor the output for information about the status of the upgrade (refer to “Monitoring CUA Status” on page 199 for more information).

Note: Rolling back the upgrade—If the upgrade is interrupted for any reason, the CUA can roll back the upgrade to the prior version. This happens automatically. If this happens, you must reperform the upgrade.

Upgrading CUA on a Dell 229

230

Servicing the CUA

Upgrading CUA on an EMC Centera nodeTo upgrade a CUA on an EMC Centera node, contact your EMC representative.


Servicing the CUA

Upgrade error messagesThe CUA upgrade process runs a number of checks before and during the upgrade. You may see the following error messages during the upgrade. In addition, if the network is down or the EMC Centera is unreachable during the upgrade, the upgrade will fail.

If the upgrade fails, CUA automatically rolls back the upgrade and restores the previous version.

Table 21 Upgrade error messages (1 of 2)

Error Type Error Messages Remedy

Wrong CUA/CFA version

Unable to determine CUA version; must be CUA 3.6 SP1 or laterUnrecognized CUA version <CUA version> must be CUA 3.6 SP1 or laterUnrecognized CUA version <CUA version>; must be CUA 3.6 SP1 or laterInvalid CUA version <CUA version>; must be CUA 3.6 SP1or later

Update CUA/CFA to version 3.6 SP1 and then do 4.0.1 upgrade.

Hardware Problems

Upgrading a Dell PowerEdge 2650 is not supportedDevice <disk ID> is not online <disk state>The following md device(s) are not synced:The following md device(s) are not complete:

Upgrade to CUA/CFA 4.0.1 supported on Dell 2850/2950, and Gen3/Gen4 nodes. Drives must be healthy (online, not re-syncing or re-building).

No Backup No backups found. CUA must be backed-up before upgrading.Unable to determine RFI of last backup.Failed to validate backup RFI Unable to determine time of last backupUnable to convert time of last backup to secondsUnable to determine RFI of database backupUnable to determine RFI of configuration files backupUnable to retrieve time from CenteraUnable to convert time from Centera to secondsLast backup was taken too long ago <backup time>Failed to verify the integrity of the configuration file backupUnable to retrieve size of the last backupUnable to retrieve complete table listing from the databaseUnable to count the numbers of rows in the <table name> tableLast database backup appears grossly undersized

Must have a recent, valid CUA/CFA backup.

Upgrade error messages 231

232

Servicing the CUA

Insufficient space Your database partition appears to be XX% full.You could run out of space for your database.Do you wish to continue (y/n)?

The upgrade may fail due to lack of disk space. You can go ahead with the upgrade but be aware that it may fail. If the upgrade fails, contact the EMC Global Services Customer Support Center.

Table 21 Upgrade error messages (2 of 2)

Error Type Error Messages Remedy


Servicing the CUA

Changing CUA hardwareChange the hardware on which the CUA is installed, by doing the following:



3. Ensure that all writes are complete by typing the following command:

sudo caginfo

Note: If any files are listed in the output, you must discontinue writing files to the EMC Centera and wait until all files are written back. Continue to issue the command until all files are written back to EMC Centera and no files appear in the output.

4. Type exit to return to the Setup Utility.

5. Select W at the Setup Utility to run a backup.

6. Once the output Backup complete appears, select X at the Setup Utility to enter a Bash Shell



8. Look for output similar to the following and take a note of the Content Address (the long string of letters and numbers).

Oct 28 09:32:05 CUA1 nmslib[603]: [75] [28-Oct-2004 9:32:05.560] [sev 2] [SFSBACKUP] Backup complete, CA is 3PADPTQ8KHOCIe3NGHJ48T4S5VG

Refer to “Viewing event messages” on page 220 for more information on viewing event messages.

Note: It is very important that you make a copy of the Content Address in order to be able to restore the CUA on the new hardware.

9. Carry out a complete installation of the CUA on the new hardware (refer to “Installation and Configuration” on page 27).

Changing CUA hardware 233

234

Servicing the CUA

10. Connect to the CUA Setup Utility on the new hardware using an SSH client and log in as gwsetup using the password EMCCUA.

11. Select R at the Setup Utility menu to run a restore using the CA provided in step 8 (refer to “Restoring a CUA configuration” on page 81).

12. Once the restore is complete, the CUA is ready to use on the new hardware.


Servicing the CUA

Replacing disk drives and nodes

Determining the drive to replace

There are two ways to determine which disk needs replacing on a CUA.

◆ CUA Graphical User Interface (GUI)

◆ CUA Setup Utility

Only use the CUA Setup Utility if access to the GUI is not available as the GUI is more reliable.

Using the CUA GUI To use the CUA GUI to determine which disk needs replacing, do the following:

1. Access the CUA GUI (refer to “Accessing the GUI” on page 201 for more information on accessing the CUA GUI).

2. Click on the Repair tab.

3. At the bottom of the Disk Management panel, each disk is represented by a small box showing the Disk ID, Bus/Target, and Status.

4. Disks that need replacing are indicated with the Status Offline (refer to Figure 28 on page 236 and Figure 29 on page 236, and Figure 30 on page 237).

5. Once you have determined which drive needs replacing, refer to “Disk replacement on a Dell” on page 238 or “Disk replacement on an EMC Centera Gen3 or Gen4 node” on page 239.

Replacing disk drives and nodes 235

236

Servicing the CUA

Figure 28 Disk failure on a Dell 2850

Figure 29 Disk failure on a Dell 2650


Servicing the CUA

Figure 30 Disk failure on an EMC Centera node

Using the CUA SetupUtility

To identify the correct disk to replace on the CUA Node using the Setup Utility, do the following:

1. Connect to the CUA Setup Utility an SSH client and log in as gwsetup using the password EMCCUA.


3. Change to the /var/log directory, by typing the following command:

cd /var/log

4. Search the bottom of the storigen_nms file for offline disks (search for OFFLINE). You will see output similar to one of the following:

Disk's state (location = 00/00/01/01.0) has been set OFFLINE due to write errors

Disk's state (location = 00/00/01/01.0) has been set OFFLINE due to read errors

Disk (location_id=00/00/001/01.0) has gone offline

Physical disk (location = 00/00/01/01.0, serial number = , device = ) has gone offline


238

Servicing the CUA

The information between the () shows you the location of the disk. The information below shows the location of each disk depending on the type of CUA Node:

Dell PowerEdge 2950◆ 00/00/00/00.0 = top left

◆ 00/00/00/01.0 = bottom left

◆ 00/00/00/02.0 = top middle

◆ 00/00/00/03.0 = bottom middle

◆ 00/00/00/04.0 = top right

◆ 00/00/00/05.0 = bottom right

Dell PowerEdge 2850◆ 00/00/00/00.0 = bottom left

◆ 00/00/00/01.0 = top left

◆ 00/00/00/02.0 = bottom middle

◆ 00/00/00/03.0 = top middle

◆ 00/00/00/04.0 = bottom right

◆ 00/00/00/05.0 = top right

Dell PowerEdge 2650◆ 00/00/00/00.0 = top left

◆ 00/00/00/01.0 = bottom middle

◆ 00/00/00/02.0 = top middle

◆ 00/00/00/03.0 = bottom right

◆ 00/00/00/04.0 = top right

Once you have determined which drive needs replacing, refer to “Disk replacement on a Dell” on page 238.

Disk replacement on a Dell

If it becomes necessary to replace a disk drive on a Dell system, contact your Dell service representative. If replacing a drive on a Dell 2950, the Dell service representative needs to perform the following procedure to ensure that the RAID configuration on the drive is correct. Provide this procedure to your Dell service representative if they are replacing a drive on a Dell 2950.


Servicing the CUA

CAUTION!This procedure is to be performed only by an authorized Dell service representative and is provided here for informational purposes only.

Drive replacement onDell 2950

To replace a drive on Dell 2950:

1. Shutdown and power off the system.

2. Replace the drive.

3. Power up the system.

4. Press CTRL-R when PowerEdge Expandable RAID Controller initializes and starts scanning devices to enter the controller setup.

5. On the VD MGMT tab highlight Controller.

6. Press F2 to get Available Operations.

7. Scroll to Foreign Config and select Clear.

8. Confirm Foreign Config Clear.

9. Highlight Controller again.

10. Press F2 to get Available Operations.

11. Select Create New VD.

12. In the Physical Discs window select an available drive and then select OK.

If you are replacing the boot drive, you must also complete the following steps:

1. Go to the Ctrl Mgmt page.

2. Highlight the Select Bootable VD field.

3. Select one of the good VDs as a bootable drive.

4. Apply the changes.

Disk replacement on an EMC Centera Gen3 or Gen4 node

To replace a disk on an EMC Centera Gen 3 or Gen 4 node, contact your EMC representative.


240

Servicing the CUA

Connecting to CUA remotelyThis section contains instructions on how to connect to a CUA/CFA from a remote location.


Servicing the CUA

Connecting to CUA with a modem

EMC Centera nodes — By default, a Gen3 node is configured for a modem, however Gen4 nodes are not. The serial line on a Gen4 CUA node is set for serial console, not modem.

Dell systems — Dell 2650s have two serial ports, and the second serial port is by default configured to use a modem. Dell 2850s and 2950s are by default configured to use a serial console, not a modem.

For Gen4 nodes, Dell 2850 and Dell 2950 systems, you need to modify parameters to support a modem.

To use a modem on CUA, do the following:

1. Log in to the CUA as gwsetup.

2. Select X from the menu to enter the Bash shell.

3. Enter the following command:

sudo /usr/local/storigen/set_serial_mode modem

4. Exit the Bash prompt.

Note: To switch the serial port back for use as a serial console, run this command:/usr/local/storigen/set_serial_mode console

Connecting to CUA remotely 241

242

Servicing the CUA


8

This chapter will help you understand the cause of an event that triggers an email notification and provides information for resolving the issue.

The major sections are:

◆ Enabling email alert notification.................................................... 244

Alerts andTroubleshooting

Alerts and Troubleshooting 243

244

Alerts and Troubleshooting

Enabling email alert notificationTo enable the Email Alert feature, refer to “Phase 2: Entering CUA/CFA configuration information” on page 40 for instructions. Enabling this feature notifies you if the CUA meets or exceeds certain predefined system default values or performance thresholds.

The notification uses the following format in the subject line of each email message:

Subject: EMC CUA filesystem alert from CUA1.emc.com

An example of the content of an email alert message is shown below:

<alert><error><parameter name="type" value="node" /><parameter name="Device" value="filesystem" /><parameter name="Message" value="File system /var usage 92% exceeds threshold 90%" /> <parameter name="Release" value="CUA4.0_3937" /> </error> <gateway><parameter name="NODE_SERIAL" value="" /> <parameter name="NodeName" value="CUA1.emc.com" /><parameter name="FPclusterID" value="f8e9bbc8-1dd1-11b2-8d3e-d38dfa138df3" /><parameter name="siteid" value="" /> <parameter name="clusterserial" value="APM00040402602" /><parameter name="IP_EXTERNAL" value="192.168.1.1" /> </gateway> </alert>

Alert parameters The Email Alert program tracks the following system and network faults and parameters:

◆ File system usage exceeds enabled thresholds (default threshold 92% full) for the following filesystems:

• /var

• /sfs

• /

• /usr

◆ CPU usage exceeds enabled thresholds (default threshold 95%).

◆ Monitored Network Interface down.

◆ Cannot communicate with EMC Centera.



◆ EMC CUA temporary key EMC CUA File System expired 1 day ago on hostname.

◆ Delivery application is down.

◆ Disk errors.

Responding to an email alert

Take the following actions to resolve the cause of the alert. Examples of these alerts are shown after Table 22 on page 245.

Table 22 CUA email alerts (1 of 2)

If you see this message... Do this...

Active CUA (1) is in an [unknown] state Standby CUA cannot get the state of the Active CUA. Check both CUA states and connectivity.

Active CUA (1) is in an [unreachable] state

Standby CUA cannot communicate with Active CUA. Check both CUA states and connectivity.

Application healthcheck failed Start the service GUI interface and select Monitor ⇒ Node Summary to verify whether the EMC Centera or other applications are in a "Failed" state. If it is, contact the EMC Global Services Customer Support Center.

Array mdX disk failure on partitionName; removing partitionName from array

Contact the EMC Global Services Customer Support Center.

CPU usage 98 exceeds threshold 95 Contact the EMC Global Services Customer Support Center.

Current directory count x exceeds the supported 1,000,000 limit.

Reduce the number of directories by changing your directory structure or adding an additional CUA.

Note: You should have no more than 30,000 directories for every parent directory

Directory count is at 85 percent of the supported 1,000,000 limit.

Reduce the number of directories by changing your directory structure or adding an additional CUA.

Note: You should have no more than 30,000 directories for every parent directory.

Disk 3 (mount point /sfs) has been removed


Enabling email alert notification 245

246


EMC CUA temporary key EMC CUA File System expired 1 day ago on hostname

To install a permanent license key, refer to “Entering permanent licensing information” on page 98.

Entry count x for directory y is at 85 % of the supported 100,000 limit.

Rework your directory structure to reduce the number of files per directory.

Note: EMC recommends that you should have no more than 5-10,000 files per directory.

Entry count for x for directory y exceeds the supported 100,000 limit.

Rework your directory structure to reduce the number of files per directory.

Note: EMC recommends that you should have no more than 5-10,000 files per directory.

Error processing activity log (/var/log1) on CUA

CUA encountered an error while processing the activity log. Contact the EMC Global Services Customer Support Center.

File system /var usage 92% exceeds threshold 90%


Gateway Backup failed at CUA1 Contact the EMC Global Services Customer Support Center.

Monitored Network Interface eth2 is down


Standby CUA (CUA2) failover for Active CUA

Informational message to indicate that failover was initiated.

Standby CUA (1) is in an [unreachable] state

Active CUA cannot communicate with Standby CUA. Check both CUA states and connectivity.

Table 22 CUA email alerts (2 of 2)

If you see this message... Do this...



Active CUA (1) is in an [unknown] state

<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"> <alert><error><parameter name="type" value="cag-dr" /><parameter name="Device" value="backup" /> <parameter name="symptomCode" value="1.1.10.6.01.03" /> <parameter name="creationDateTime" value="1177618531" /> <parameter name="Message" value="Active CUA (1) is in an [unknown] state; pid 12345" /> <parameter name="Release" value="CUA4.0_4725" /> </error> <gateway> <parameter name="NODE_SERIAL" value="" /> <parameter name="NodeName" value="siqa-cua1" /> <parameter name="FPclusterID" value="44dc8634-1dd2-11b2-9635-8937f2e27cda" /> <parameter name="siteid" value="" /> <parameter name="clusterserial" value="APM00031700200" /> <parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /> </gateway> </alert> </ConnectHome>

Active CUA (1) is in an [unreachable] state<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"> <alert> <error> <parameter name="type" value="cag-dr" /> <parameter name="Device" value="backup" /> <parameter name="symptomCode" value="1.1.10.6.01.02" /> <parameter name="creationDateTime" value="1177618528" /> <parameter name="Message" value="Active CUA (1) is in an [unreachable] state; pid * Data missing *" /> <parameter name="Release" value="CUA4.0_4725" /> </error> <gateway> <parameter name="NODE_SERIAL" value="" /> <parameter name="NodeName" value="siqa-cua1" /> <parameter name="FPclusterID" value="44dc8634-1dd2-11b2-9635-8937f2e27cda" /> <parameter name="siteid" value="" /> <parameter name="clusterserial" value="APM00031700200" /> <parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /> </gateway> </alert>


248


</ConnectHome>

Application healthcheck failed<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"> <alert> <error> <parameter name="type" value="healthcheck" /> <parameter name="Device" value="healthcheck" /> <parameter name="Message" value="Service Centera is not running" /> <parameter name="Release" value="CUA4.0_4281.p1" /> </error> <gateway> <parameter name="NODE_SERIAL" value="" /> <parameter name="NodeName" value="l4-cua1.centera.lab.emc.com" /> <parameter name="FPclusterID" value="6e3107ec-1dd2-11b2-9070-89ab1ffbfb86" /> <parameter name="siteid" value="" /> <parameter name="clusterserial" value="" /> <parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /> </gateway> </alert> </ConnectHome>

Array mdX disk failure on partitionName; removing partitionName from array<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"><alert><error><parameter name="type" value="node" /><parameter name="Device" value="filesystem" /><parameter name="Message" value="Array mdX disk failure on partitionName; removing partitionName from array" /><parameter name="Release" value="CUA4.0_4281.p1" /></error><gateway><parameter name="NODE_SERIAL" value="" /><parameter name="NodeName" value="l4-cua1.centera.lab.emc.com" /><parameter name="FPclusterID" value="6e3107ec-1dd2-11b2-9070-89ab1ffbfb86" /><parameter name="siteid" value="" /><parameter name="clusterserial" value="" /><parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /></gateway></alert></ConnectHome>



CPU usage 98 exceeds threshold 95<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"> <alert> <error> <parameter name="type" value="node" /> <parameter name="Device" value="cpu" /> <parameter name="Message" value="CPU usage 95 exceeds threshold 90" /> <parameter name="Release" value="CUA4.0_4281.p1" /> </error> <gateway> <parameter name="NODE_SERIAL" value="" /> <parameter name="NodeName" value="l4-cua1.centera.lab.emc.com" /> <parameter name="FPclusterID" value="6e3107ec-1dd2-11b2-9070-89ab1ffbfb86" /> <parameter name="siteid" value="" /> <parameter name="clusterserial" value="" /> <parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /> </gateway> </alert> </ConnectHome>

Current directory count x exceeds the supported 1,000,000 limit<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"><alert><error><parameter name="type" value="node" /><parameter name="Device" value="filesystem" /><parameter name="Message" value="Current directory count 1010317 exceeds the supported 1000000 limit" /><parameter name="Release" value="CUA4.0_4685" /></error><gateway><parameter name="NODE_SERIAL" value="" /><parameter name="NodeName" value="cagblue6" /><parameter name="FPclusterID" value="9cf8d11e-1dd1-11b2-8427-e51fdab72fae" /><parameter name="siteid" value="12345" /><parameter name="clusterserial" value="APM00041703888" /><parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /></gateway></alert></ConnectHome>


250


Directory count is at x percent of the supported 1,000,000 limit<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"><alert><error><parameter name="type" value="node" /><parameter name="Device" value="filesystem" /><parameter name="Message" value="Directory count is at 90 percent of the supported 1000000 limit" /><parameter name="Release" value="CUA4.0_4685" /></error><gateway><parameter name="NODE_SERIAL" value="" /><parameter name="NodeName" value="cagblue6" /><parameter name="FPclusterID" value="9cf8d11e-1dd1-11b2-8427-e51fdab72fae" /><parameter name="siteid" value="12345" /><parameter name="clusterserial" value="APM00041703888" /><parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /></gateway></alert></ConnectHome>

Disk 3 (mount point /sfs) has been removed<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"><alert><error><parameter name="type" value="node" /><parameter name="Device" value="filesystem" /><parameter name="Message" value="Disk 3 (mount point /sfs) has been removed" /><parameter name="Release" value="CUA4.0_4281.p1" /></error><gateway><parameter name="NODE_SERIAL" value="" /><parameter name="NodeName" value="l4-cua1.centera.lab.emc.com" /><parameter name="FPclusterID" value="6e3107ec-1dd2-11b2-9070-89ab1ffbfb86" /><parameter name="siteid" value="" /><parameter name="clusterserial" value="" /><parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /></gateway></alert></ConnectHome>



EMC CUA temporary key EMC CUA File System expired 1 day ago on hostname<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"> <alert> <error> <parameter name="type" value="license" /> <parameter name="Device" value="nic" /> <parameter name="Message" value="EMC CUA temporary key CUA expired 2 days ago on CUA1" /> <parameter name="Release" value="CUA4.0_4281.p1" /> </error> <gateway> <parameter name="NODE_SERIAL" value="" /> <parameter name="NodeName" value="l4-cua1.centera.lab.emc.com" /> <parameter name="FPclusterID" value="6e3107ec-1dd2-11b2-9070-89ab1ffbfb86" /> <parameter name="siteid" value="" /> <parameter name="clusterserial" value="" /> <parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /> </gateway> </alert> </ConnectHome>

Entry count x for directory y is at x % of the supported 1,000,000 limit<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"><alert><error><parameter name="type" value="node" /><parameter name="Device" value="filesystem" /><parameter name="Message" value="Entry count 98010 for directory /sfs/gateway_cifs/data is at 98 percent of the supported 100000 limit" /><parameter name="Release" value="CUA4.0_4685" /></error><gateway><parameter name="NODE_SERIAL" value="" /><parameter name="NodeName" value="cagblue6" /><parameter name="FPclusterID" value="9cf8d11e-1dd1-11b2-8427-e51fdab72fae" /><parameter name="siteid" value="12345" /><parameter name="clusterserial" value="APM00041703888" /><parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /></gateway></alert></ConnectHome>


252


Entry count for x for directory y exceeds the supported 1,000,000 limit<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"><alert><error><parameter name="type" value="node" /><parameter name="Device" value="filesystem" /><parameter name="Message" value="Entry count 100010 for directory /sfs/gateway_cifs/data exceeds the supported 100000 limit" /><parameter name="Release" value="CUA4.0_4685" /></error><gateway><parameter name="NODE_SERIAL" value="" /><parameter name="NodeName" value="cagblue6" /><parameter name="FPclusterID" value="9cf8d11e-1dd1-11b2-8427-e51fdab72fae" /><parameter name="siteid" value="12345" /><parameter name="clusterserial" value="APM00041703888" /><parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /></gateway></alert></ConnectHome>

Error processing activity log (/var/log1) on CUA<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"> <alert> <error> <parameter name="type" value="cag-dr" /> <parameter name="Device" value="backup" /> <parameter name="symptomCode" value="1.1.10.6.01.06" /> <parameter name="creationDateTime" value="1177618543" /> <parameter name="Message" value="Error processing activity log (/var/log1) on CUA (CUA1); pid 1234" /> <parameter name="Release" value="CUA4.0_4725" /> </error> <gateway> <parameter name="NODE_SERIAL" value="" /> <parameter name="NodeName" value="siqa-cua1" /> <parameter name="FPclusterID" value="44dc8634-1dd2-11b2-9635-8937f2e27cda" /> <parameter name="siteid" value="" /> <parameter name="clusterserial" value="APM00031700200" /> <parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /> </gateway> </alert> </ConnectHome>



File system /var usage 92% exceeds threshold 90%<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"><alert><error><parameter name="type" value="node" /><parameter name="Device" value="filesystem" /><parameter name="Message" value="File system /var usage 92% exceeds threshold 90%" /><parameter name="Release" value="CUA4.0_4281.p1" /></error><gateway><parameter name="NODE_SERIAL" value="" /><parameter name="NodeName" value="l4-cua1.centera.lab.emc.com" /><parameter name="FPclusterID" value="6e3107ec-1dd2-11b2-9070-89ab1ffbfb86" /><parameter name="siteid" value="" /><parameter name="clusterserial" value="" /><parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /></gateway></alert></ConnectHome>

Gateway Backup failed at CUA1<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"> <alert> <error> <parameter name="type" value="cag-dr" /> <parameter name="Device" value="backup" /> <parameter name="symptomCode" value="1.1.10.6.01.01" /> <parameter name="creationDateTime" value="1177618494" /> <parameter name="Message" value="Gateway Backup failed at CUA1 for CUBE3" /> <parameter name="Release" value="CUA4.0_4725" /> </error> <gateway> <parameter name="NODE_SERIAL" value="" /> <parameter name="NodeName" value="siqa-cua1" /> <parameter name="FPclusterID" value="44dc8634-1dd2-11b2-9635-8937f2e27cda" /> <parameter name="siteid" value="" /> <parameter name="clusterserial" value="APM00031700200" /> <parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /> </gateway> </alert> </ConnectHome>


254


Monitored Network Interface eth2 is down<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"> <alert> <error> <parameter name="type" value="node" /> <parameter name="Device" value="nic" /> <parameter name="Message" value="Monitored Network Interface eth2 is down" /> <parameter name="Release" value="CUA4.0_4281.p1" /> </error> <gateway> <parameter name="NODE_SERIAL" value="" /> <parameter name="NodeName" value="l4-cua1.centera.lab.emc.com" /> <parameter name="FPclusterID" value="6e3107ec-1dd2-11b2-9070-89ab1ffbfb86" /> <parameter name="siteid" value="" /> <parameter name="clusterserial" value="" /> <parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /> </gateway> </alert> </ConnectHome>

Standby CUA (CUA2) failover for Active CUA<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"> <alert> <error> <parameter name="type" value="cag-dr" /> <parameter name="Device" value="backup" /> <parameter name="symptomCode" value="1.1.10.6.01.05" /> <parameter name="creationDateTime" value="1177618540" /> <parameter name="Message" value="Standby CUA (CUA2) failover for Active CUA (CUA1)" /> <parameter name="Release" value="CUA4.0_4725" /> </error> <gateway> <parameter name="NODE_SERIAL" value="" /> <parameter name="NodeName" value="siqa-cua1" /> <parameter name="FPclusterID" value="44dc8634-1dd2-11b2-9635-8937f2e27cda" /> <parameter name="siteid" value="" /> <parameter name="clusterserial" value="APM00031700200" /> <parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /> </gateway> </alert> </ConnectHome>



Standby CUA (1) is in an [unreachable] state<?xml version="1.0"?><ConnectHome xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" SchemaVersion="1.0.6"> <alert> <error> <parameter name="type" value="cag-dr" /> <parameter name="Device" value="backup" /> <parameter name="symptomCode" value="1.1.10.6.01.04" /> <parameter name="creationDateTime" value="1177618536" /> <parameter name="Message" value="Standby CUA (1) is in an [unreachable] state" /> <parameter name="Release" value="CUA4.0_4725" /> </error> <gateway> <parameter name="NODE_SERIAL" value="" /> <parameter name="NodeName" value="siqa-cua1" /> <parameter name="FPclusterID" value="44dc8634-1dd2-11b2-9635-8937f2e27cda" /> <parameter name="siteid" value="" /> <parameter name="clusterserial" value="APM00031700200" /> <parameter name="IP_EXTERNAL" value="10.xxx.xx.xx" /> </gateway> </alert> </ConnectHome>


256



E

Index

AAccess profiles

backup/restore profile 91CFA 136 to 137configuring 89default 90gateways 91mapping 90permissions 88search order 92using 88

Active CUA/CFA 170Activity logs 153Advanced configuration 53 to 112Audience 15

CCelerra 131Centera

Gen3 node 30Gen4 node 30

CFAaccess profiles 136 to 137Configuration Checklist 127configuration file 138configuration menu 128configuration options 139configure DNS settings 135configuring 134defining file systems 139defining migration 144defining migration policies 138, 141digest access authentication 134

enabling 126, 127, 131enabling filename translation 128establishing connection from Celerra 131filename translation 128HTTP access 130HTTP connection 132installation 126introduction 124manually starting and stopping a policy run

149migration policies 138NFS permissions 130round robin 195sample configuration file 145scheduling a policy run 147simulating a policy run 150viewing CFA statistics 152

CFA, policies 124Change Password page 224CIFS

adding AD access control 65adding IP access control 60adding NT access control 62connecting to CUA 59mapping a drive 71

Computing performance statsNode Detail page 213Node Summary page 206

Configure SNMP Parameters page 107CUA

advanced configuration 53 to 112basic configuration 38 to 45configuration information 40verifying functionality 46 to 51


258

Index

CUA accessCIFS 59FTP 78HTTP 79NFS 75

Custom metadata 95

DDell systems

PowerEdge 2650 30PowerEdge 2850 29PowerEdge 2950 29replacing a disk 238

Digest Access Authentication 134Disaster recovery 22Disk

determining drive to replace 235monitoring 223replacement on a Dell 238replacement on a node 239replacing 235

DNSdomain 41round robin 195settings 135

EEmail alerts

configuring 41enabling 244panel 111parameters 244responding 245testing 96

Enable Auth Trap for SNMP 108Enabling filename translation 128Enhanced availability

configuring 169configuring the Active system 170configuring the Standby system 171failover 174initiating failover 175introduction 160recovering files 179resolving an Active system failure 173types of configurations 163

Error messages, upgrade 231Event messages 220Event Messages page 104, 220

FFailover

initiating 174, 175testing 174

File protection 86File stubbing 132File System performance stats

Node Detail page 211Node Summary page 206

Filename translation 128Files, recovering 179FTP access 78

GGUI

CUA 57Monitoring 200

HHardware Problems 231Hardware specifications 28Hostname 41HTTP access 79, 130HTTP Delivery Service performance stats

Node Summary page 206HTTP stubbing 132, 195

IInstallation 27

bonding NICs 39Dell 34 to 37node 37 to??required network information 32

Insufficient space 232Interface IP addresses 41

LLicenses 98Limitations, CUA/CFA 26Logs


E

Index

Activity logs 153

MMemory performance stats


Metadata, custom 95MIB-2 objects 108Migration policies, defining 138Modems 240Monitor GUI 200Monitoring

disks 223node status 208services 215standby CUA/CFA 216system status 203

NNetwork real-time performance stats


NFSaccess 75adding access control 76adding mount points 75CFA connection 132mount example 75stubbing 132

NIC bonding 39No Backup 231Node Summary page 203Nodes

monitoring status 208replacing 235replacing disks 239

Ooperating system 15Orphans

deleting 157excluding from deletion 157finding 155

PPasswords 33

changing 103GUI, changing 224

PEA files 88, 92ping command 44Policies

CFA 124Policy run

manually starting and stopping 149scheduling 147simulating 150

Product Overview 20

RReal-time monitoring


Remote connections 240Restoring a configuration 81Restricted Host Address

to access the SNMP agent 109Retention

classes 25periods 100 to 102

Round robin 195

SServices real-time performance stats

Node Detail page 210Services, monitoring 215Setup utility, CUA 54Site requirements 32smb.conf file 71SNMP 107

SNMP Trap Destinations panel 110Standby CUA/CFA 171, 216Storage performance stats

Node Detail page 214Subsystem state

DEGRADED 204FAILED 204NORMAL 204UNKNOWN 204UNMOUNTED 204


260

Index

valid states 204System

health, summary 203monitoring status 203time 104

System components, report 203System Contact MIB-2 object 108System Description MIB-2 object 108System Information panel 108System Location MIB-2 object 108System Name MIB-2 object 108

TTime, system 104

UUpgrade error messages 231Upgrading CUA/CFA

Dell 228EMC Centera node 230

User interfaces 54Usernames and passwords 33UTF-8 filename support 112

VVerify

CUA/EMC Centera communication 46email home functionality 46invalidation functionality 50read functionality 51writeback functionality 49

WWriteback 49

audit trails 218Wrong CUA/CFA version 231


EMC Centera Universal Access/FileArchiver · wrongly, and Richard Stallman agrees with me, when...

Documents

Transcript of EMC Centera Universal Access/FileArchiver · wrongly, and Richard Stallman agrees with me, when...