SOL90713_Quest_NetVault_SmartDisk_1.5.1_Installation_Upgrade_Guide_Rev1_0.pdf

NSD-101-1.5.1-EN-02 12/14/11

Quest NetVault SmartDisk

version 1.5.1

Installation/Upgrade Guide

2011 Quest Software, Inc.ALL RIGHTS RESERVED.This guide contains proprietary information protected by copyright. The software described in this guide is furnished under a software license or nondisclosure agreement. This software may be used or copied only in accordance with the terms of the applicable agreement. No part of this guide may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording for any purpose other than the purchasers personal use without the written permission of Quest Software, Inc.

DisclaimerThe information in this document is provided in connection with Quest products. No license, express or implied, by estoppel or otherwise, to any intellectual property right is granted by this document or in connection with the sale of Quest products. EXCEPT AS SET FORTH IN QUEST'S TERMS AND CONDITIONS AS SPECIFIED IN THE LICENSE AGREEMENT FOR THIS PRODUCT, QUEST ASSUMES NO LIABILITY WHATSOEVER AND DISCLAIMS ANY EXPRESS, IMPLIED OR STATUTORY WARRANTY RELATING TO ITS PRODUCTS INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTY OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OR NON-INFRINGEMENT. IN NO EVENT SHALL QUEST BE LIABLE FOR ANY DIRECT, INDIRECT, CONSEQUENTIAL, PUNITIVE, SPECIAL OR INCIDENTAL DAMAGES (INCLUDING, WITHOUT LIMITATION, DAMAGES FOR LOSS OF PROFITS, BUSINESS INTERRUPTION OR LOSS OF INFORMATION) ARISING OUT OF THE USE OR INABILITY TO USE THIS DOCUMENT, EVEN IF QUEST HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. Quest makes no representations or warranties with respect to the accuracy or completeness of the contents of this document and reserves the right to make changes to specifications and product descriptions at any time without notice. Quest does not make any commitment to update the information contained in this document.

If you have any questions regarding your potential use of this material, contact:

Quest Software World HeadquartersLEGAL Dept5 Polaris WayAliso Viejo, CA 92656email: [email protected]

For regional and international office information, refer to the web site (www.quest.com).

TrademarksQuest, Quest Software, the Quest Software logo, and NetVault are trademarks and registered trademarks of Quest Software, Inc. in the United States of America and other countries. For a complete list of Quest Softwares trademarks, see http://www.quest.com/legal/trademarks.aspx. Other trademarks and registered trademarks are property of their respective owners.

PatentsThis product is protected by U.S. Patents # 5,250,687; 7,814,260; 7,913,043; and 7,979,650. Protected by Japanese, E.U., French, and UK patents 1615131 and 05250687.0, and German patent DE602004002858. Additional patents pending.

Chapter 1: Introducing Quest NetVault SmartDisk 7

Quest NetVault SmartDisk At a Glance ........................................................................ 7 Key Benefits ...................................................................................................................... 8 Feature List ....................................................................................................................... 9 Target Audience ................................................................................................................ 9 Recommended Additional Reading ................................................................................ 9 About Quest Software ...................................................................................................... 9

- Contacting Quest Software ............................................................................................................ 10- Quest Support ................................................................................................................................ 10

Chapter 2: Understanding NVSD Architecture 11

NVSD Architecture An Overview ................................................................................ 11 Components .................................................................................................................... 12

- NVSD Instance .............................................................................................................................. 12- NVSD Storage Pools ..................................................................................................................... 12- Manifests ........................................................................................................................................ 13- Processes ...................................................................................................................................... 13

Nondeduplicated Backup Process ................................................................................ 14 Nondeduplicated Restore Process ............................................................................... 15 Deduplicated Backup Process ...................................................................................... 16 Deduplicated Restore Process ...................................................................................... 18 Garbage Collection Process .......................................................................................... 19

Chapter 3: Planning Your NVSD Deployment 21

NVSD Deployment Strategy An Overview ................................................................. 22 Defining Which Data to Deduplicate ............................................................................. 22

- Identifying the Ideal Targets for Deduplication ............................................................................... 22- Understanding the Cost of Restoring Deduplicated Data .............................................................. 23

Table of Contents- Identifying Targets for Nondeduplication ....................................................................................... 23- Deduplicating Similar Data Together ............................................................................................. 23- Separating vs. Combining Deduplicated Data in NVSD Instances ................................................ 23

4 Determining the Size of the NVSD Deployment ...........................................................25- Calculating NVSD License Capacity .............................................................................................. 25- Calculating the Required Number of NVSD Instances .................................................................. 28- Calculating the Total Required Physical Disk Space for Licensed Capacity ................................. 32

Choosing a Deployment Location for an NVSD Instance ............................................36- Reviewing Deployment Options for NVSD Nondeduplicated Instances ........................................ 36- Reviewing Deployment Options for NVSD Deduplicated Instances .............................................. 41- Reviewing Deployment Options for Disaster Recovery ................................................................. 42

Defining an NVSD Storage Pool Strategy .....................................................................44- Understanding Space Allocation for NVSD Storage Pools ............................................................ 44- Understanding the Simplified Management Strategy ..................................................................... 46- Understanding the Optimal Performance Strategy ........................................................................ 46

Chapter 4: Installing NVSD 49

Installing NVSD An Overview ......................................................................................49 System Requirements .....................................................................................................49

- Software Requirements ................................................................................................................. 49- Hardware Requirements ................................................................................................................ 50- Authentication Requirements ......................................................................................................... 51

Installing the NVSD Software .........................................................................................53- Important Considerations ............................................................................................................... 53- Installing on Linux or UNIX (Excluding Solaris) ............................................................................. 54- Installing on Solaris (SPARC, x86, or x86-64) ............................................................................... 57- Installing on Windows .................................................................................................................... 60- Installing on Mac OS X .................................................................................................................. 68

Chapter 5: Completing the Installation Process 75

Updating the Storage Pool Directory Structure ............................................................75 Adding NVSD Devices to Your Backup Product ..........................................................75 Protecting the NVSD Unique Identity ............................................................................76

Chapter 6: Licensing NVSD 77

NVSD License Keys An Overview ...............................................................................77 Obtaining the NVSD Machine ID ....................................................................................77 Requesting a License Key ..............................................................................................78

Installing a License Key on Linux, UNIX, or Mac OS X ................................................78

Quest NetVault SmartDiskInstallation/Upgrade Guide

5

Installing a License Key on Windows ........................................................................... 79

Chapter 7: Uninstalling NVSD 81

Preserving Data Before Removing NVSD ..................................................................... 81 Removing NVSD from Linux, UNIX (Excluding Solaris), or Mac OS X ...................... 81 Removing from Solaris (SPARC, x86, or x86-64) ......................................................... 82 Removing from Windows ............................................................................................... 82

Chapter 8: Reinstalling NVSD 85

Reinstalling NVSD An Overview ................................................................................. 85 Prerequisites for Reinstalling NVSD ............................................................................. 85 Reinstalling NVSD if identity.cfg is Available ............................................................... 86 Reinstalling NVSD if identity.cfg is Not Available ........................................................ 88

Chapter 9: Upgrading NVSD 91

Checking the Status and Taking the Device Offline .................................................... 91 Upgrading the NVSD Software ...................................................................................... 92

Chapter 10: Troubleshooting NVSD 93

Corrupted or Missing Installfiles Subdirectory Impedes Removal or Upgrade on Windows .......................................................................................................................... 93


7

Chapter 1:

INTRODUCING QUEST NETVAULT SMARTDISK

Quest NetVault SmartDisk At a Glance Key Benefits Feature List Target Audience Recommended Additional Reading About Quest Software

Contacting Quest Software Quest Support

1.1.0 Quest NetVault SmartDisk At a GlanceQuest NetVault SmartDisks (NVSDs) disk-based backup and data-deduplication option reduces storage costs with its powerful byte-level, variable-block-based software deduplication, which packs up to 12 times more protected data into the same storage area for a 92% reduction in storage footprint. Hardware costs are reduced by using existing storage infrastructure instead of requiring specific drives or appliances. Seamless integration with other Quest products, such as NetVault Backup (NVBU) and vRanger, speeds deployment while simplified ease-of-use reduces the level of storage expertise required to perform deduplicated disk-based backups.

To shrink backup windows with no additional impact on protected-server resources, you can schedule NVSDs post-process deduplication option outside the backup window.

With NVSD, administrators have increased choices, including the ability to deploy multiple NVSD Instances to improve load balancing and performance; copy or move data between NVSD Instances, to VTL, or to tape-based devices to place redundant backups in offsite locations for failover and disaster-recovery purposes.

Chapter 1Introducing Quest NetVault SmartDisk

8

1.2.0 Key Benefits Reduces Storage Costs NVSD reduces storage costs with its byte-level,

variable-block-based deduplication option. This option packs up to 12 times more protected data into the same storage area with the ability. It accomplishes this by analyzing smaller amounts of data, thereby avoiding the problem that occurs when fixed blocks mask changes and store duplicate data as unique.NVSD is hardware-agnostic and works on most file systems, so hardware costs are reduced by using the existing storage infrastructure instead of requiring you to acquire specific drives or appliances. The ability to add additional file-system paths easily to NVSD Storage Pools reduces costs by deferring storage expenditures into new budget periods when costs are lower or budget restraints may be eased and also ensures that storage does not sit unused.Data stored in NVSD can have individual retention policies, which gives administrators the flexibility to store the most recent copies on disk and move older data to an offsite location or to less expensive storage. This provides a foundation for Lifecycle Management, which further reduces costs with tiered-storage deployments.

Shrinks Backup Windows To shrink backup windows with no additional impact on protected-server resources, you can schedule NVSDs post-process deduplication option outside the backup window.Even if you do not use the optional NVSD deduplication feature, you will still receive all the advantages of disk-based backup. For example, you can perform more simultaneous backups with disk than you can by writing backups to tape in a serial process. Completing your backups faster ensures that the performance of the application, database, and files servers is not impacted during business hours by overrunning backup windows.

Maximizes Flexibility Administrators are able to deploy multiple NVSD Instances to distribute network, disk, and deduplication resources, which improves load balancing and performance. Shared-storage support provides the ability to share disk-based backup storage across heterogeneous platforms, which provides a single data-protection solution.

Speeds Restores NVSD improves restore times because the data is coming from random-access storage without having to find and mount tapes, and then search for the correct tape position. The majority of restore requests come in the first 14 to 30 days after backup. With disk-based backup, you can minimize downtime with shorter Recovery Time Objectives (RTOs), and you can focus on other critical tasks.


9

1.3.0 Feature List Simplified disk-based backup Byte-level, variable-block software-based deduplication 12-to-1 deduplication ratio Hardware agnostic Post-process deduplication Rapid deployment with existing infrastructure Seamless integration with other products, such as NVBU and vRanger Job-level deduplication Shared-storage support Easily extendible Retention-policy support Heterogeneous-platform support Web-based Distributed Authoring and Versioning (WebDAV)-authenticated

access

1.4.0 Target AudienceThis guide is intended for backup administrators and other technical personnel who are responsible for designing and implementing a backup strategy for the organization. A good understanding of the operating system (OS) on which NVSD is running is assumed.

1.5.0 Recommended Additional ReadingThe following documentation is also available:

Quest NetVault SmartDisk Administrators Guide This guide describes how to configure and work with NVSD.

Quest NetVault SmartDisk Supported Platforms This guide lists the operating systems and versions supported by NVSD.

https://support.quest.com/Search/SearchDownloads.aspx

1.6.0 About Quest SoftwareQuest Software simplifies and reduces the cost of managing IT for more than 100,000 customers worldwide. Quests innovative solutions make solving the toughest IT management problems easier, enabling customers to save time and

money across physical, virtual, and cloud environments. For more information about Quest, go to www.quest.com.

1 Chapter 1Introducing Quest NetVault SmartDisk

0

1.6.1 Contacting Quest Software

For regional and international office information, refer to the web site.

1.6.2 Quest SupportQuest Support is available to customers who have a trial version of a Quest product or who have purchased a Quest product and have a valid maintenance contract. Quest Support provides unlimited 24x7 access to SupportLink, Quest Supports self-service portal. Visit SupportLink at http://support.quest.com.

From SupportLink, you can:

Retrieve thousands of solutions from the online Knowledge Base. Download the latest releases and service packs. Create, update, and review support cases.

For a detailed explanation of support programs, online services, contact information, policies. and procedures, view the Global Support Guide. This guide is available at: http://support.quest.com.

Email: [email protected]: Quest Software, Inc.

World Headquarters5 Polaris WayAliso Viejo, CA 92656USA

Web site: www.quest.com


11

Chapter 2:

UNDERSTANDING NVSD ARCHITECTURE

NVSD Architecture An Overview Components

NVSD Instance NVSD Storage Pools Manifests Processes

Nondeduplicated Backup Process Nondeduplicated Restore Process Deduplicated Backup Process Deduplicated Restore Process Garbage Collection Process

2.1.0 NVSD Architecture An OverviewNVBU is used as an example throughout this chapter; however, you can use any product that works with NVSD.

Figure 2-1:Instance NetVault SmartDisk Instance

NetVault SmartDisk Storage Pools

1 Chapter 2Understanding NVSD Architecture

2

2.2.0 Components

2.2.1 NVSD InstanceA NetVault SmartDisk Instance is comprised of one or more Storage Pools and a set of processes that perform disk-based backups and byte-level, variable-block software deduplication. You can deploy a single NVSD Instance on a dedicated server, or another server or client, and it can accept data streams from heterogeneous platforms. In addition, you can deploy multiple NVSD Instances to distribute network, disk, and deduplication resources, which improves load balancing and performance. When multiple NVSD Instances are deployed, Storage Pools and processes are not shared across the multiple NVSD Instances.

2.2.2 NVSD Storage PoolsA NetVault SmartDisk Storage Pool is comprised of one or more file-system volumes, which you can easily extend by adding additional file-system paths. Each Storage Pool is associated with roles that it can perform. When configuring the Storage Pool, you can specify the preferred roles or the roles not allowed. Available roles include the Content Index (also referred to as Disk Index), Staging Store (also referred to as Staging), Chunk Index, and Chunk Store (also referred to as Storage). For example, Storage Pool A will be used for the Content Index and the Chunk Index while Storage Pool B will only be used for the Staging Store and Chunk Store. Storage Pool A can be comprised of fault-tolerant disks with good random-access performance, while Storage Pool B is comprised of fault-tolerant disks with good streaming performance. When NVSD is analyzing the available Storage Pools, it chooses the optimal Storage Pool to use based on available space, the defined roles, and the other current activity in the NVSD Instance.

2.2.2.a Storage-Pool Roles Content Index The NetVault SmartDisk Content Index role is the index

where the NVSD Instance tracks the data it is currently protecting and whether it is stored in the Staging Store or the Chunk Store. The Content Index is relatively small and should reside in a Storage Pool comprised of fault-tolerant disks with good random-access performance.

Staging Store The NetVault SmartDisk Staging Store role is where all nondeduplicated data streams are stored, whether they are awaiting post-processing deduplication or were not selected for deduplication. Data streams that were selected for deduplication are deleted from the Staging Store after the post-processing deduplication is completed. The Staging Store should reside in a Storage Pool comprised of fault-tolerant disks with

good streaming performance.


13

Chunk Index The NetVault SmartDisk Chunk Index role is a list of the unique Chunks and where each Chunk is stored in the Chunk Store. There is one Chunk Index per NVSD Instance. The Chunk Index should reside in a Storage Pool comprised of fault-tolerant disks with good random-access performance.

Chunk Store The NetVault SmartDisk Chunk Store role is where the unique Chunks are stored. The Chunk Store should reside in a Storage Pool comprised of fault-tolerant disks.

2.2.3 ManifestsStored in an NVSD Storage Pool, a NetVault SmartDisk Manifest is an ordered list of the Chunks that comprise the data stream. It is used during the reconstruction of the data stream to ensure that the Chunks are reconstructed in the correct order. There is one Manifest per data stream.

2.2.4 Processes Storage Service The NetVault SmartDisk Storage Service listens on a

user-defined port for data flowing to and from NVSD. There is one Storage Server process per data stream.

Deduplicator The NetVault SmartDisk Deduplicator performs byte-level, variable-block software-based deduplication by breaking the data streams into unique Chunks. The Deduplicator uses the Chunk Store Daemon (CSD) to query the Chunk Index to determine if the Chunk is unique. When a Chunk is found in the Chunk Index, the Chunk Store Daemon stores the unique Chunk in the Chunk Store. There is one Deduplicator process per data stream.

Chunk Store Daemon The NetVault SmartDisk Chunk Store Daemon is the only process that writes to the Chunk Index and Chunk Store. There is only one CSD per NVSD Instance. The CSD also manages the Garbage Collection process.

percolatormonitor The percolatormonitor is a privileged process that launches and monitors other NVSD Processes.

percolatorslave The percolatorslave is responsible for communication routing. It routes messages between NVSD Processes on the local machine and between NVSD and other remote servers.

logd The logd is the logging daemon that creates the NVSD logs and stores them in an internal database. It also sends them to the applicable system location, such as the Application log in the Windows Event Viewer.

superserver The superserver is the high-level manager of all other processes. It coordinates operations, balances the load between file

systems, and schedules deduplication and Garbage Collection.


4

winservicerunner On Windows, the winservicerunner runs the Windows Service and launches the percolatormonitor.

2.3.0 Nondeduplicated Backup ProcessFigure 2-2:

Back up data without

deduplicating it

1. After the user initiates a nondeduplicated NVBU Backup that is targeted to an NVSD Instance, the NVBU Client streams data on the user-defined port to Storage Service.

2. The Storage Service stores the backup stream from the NVBU Client in the Staging Store.

3. The Storage Service updates the Content Index to indicate where the backup is stored in the Staging Store.

4. The Storage Service updates the NVBU Client to indicate that the backup stream has been successfully stored.

5. The NVBU Client updates the NVBU Server with a Backup Completed job status.

NetVault SmartDisk Instance



15

2.4.0 Nondeduplicated Restore ProcessFigure 2-3:

Restore data that was not

deduplicated

1. After the user initiates a restore of a nondeduplicated NVBU Backup that is targeted to an NVSD Instance, the NVBU Client requests data on the user-defined port from the Storage Service.

2. The Storage Service queries the Content Index to determine whether the backup is stored in the Staging Store for nondeduplicated backups or the Chunk Store for deduplicated backups.

3. The Storage Service retrieves the backup stream from the Staging Store and streams the data to the NVBU Client.

4. On completion of restoring the backup stream, the NVBU Client updates the NVBU Server with a Restore Completed job status.




6

2.5.0 Deduplicated Backup ProcessFigure 2-4:

Back up data and

deduplicate it

1. After the user initiates a deduplication-enabled NVBU Backup that is targeted to an NVSD Instance, the NVBU Client streams data on the user-defined port to the Storage Service.

2. The Storage Service stores the backup stream from the NVBU Client in the Staging Store.

3. The Storage Service updates the Content Index to indicate where the backup is stored in the Staging Store.

4. The Storage Service updates the NVBU Client to indicate that the backup stream has been successfully stored.

5. The NVBU Client updates the NVBU Server with a Backup Completed job status.

A During the user-defined deduplication window, backup data is retrieved from the Staging Store and sent to the Deduplicator.




17

B The Deduplicator performs byte-level, variable-block software-based deduplication by breaking data streams into unique Chunks and sending them to the Chunk Store Daemon.

C The Deduplicator creates a backup-specific Manifest, which is an ordered list of Chunks that comprise the backup.

D The Chunk Store Daemon sends the unique Chunks to the Chunk Store for storage and updates the Chunk Index to indicate where each Chunk is stored in the Chunk Store.

E The Content Index is updated to indicate that the backup is stored in the Chunk Store.

F The backup data stream is removed from the Staging Store.


8

2.6.0 Deduplicated Restore ProcessFigure 2-5:

Restore data that was

deduplicated

1. After the user initiates a restore of a deduplicated NVBU Backup that is targeted to an NVSD Instance, the NVBU Client requests the data on the user-defined port from the Storage Service.

2. The Storage Service queries the Content Index to determine whether the backup is stored in the Staging Store for nondeduplicated backups or the Chunk Store for deduplicated backups.

3. The Storage Service queries the backups Manifest for the first or next batch of Chunk Names in the backup.

4. The Storage Service queries the Chunk Store Daemon for the location of the Chunks in the Chunk Store.

5. Using Chunk Names from the Manifest, the Chunk Store Daemon queries the Chunk Index for pointers to the Chunks in the Chunk Store

6. The Chunk Store Daemon provides the Storage Service with the location of the Chunks in the Chunk Store.




19

7. The Storage Service retrieves the Chunks from the Chunk Store and streams the Chunk to the NVBU Client.Step 4 through Step 7 repeat for each batch of Chunk Names in the Manifest.

8. On completion of restoring all Chunk Names in Manifest, the NVBU Client updates the NVBU Server with a Restore Completed job status.

2.7.0 Garbage Collection ProcessWhen backups targeted to NVSD are retired, the data is removed from the backup database; however, the unique chunks are not automatically removed from the Chunk Store. The NVSD Garbage Collection process is designed to remove orphaned chunks from the Chunk Store and reclaim disk space.Garbage Collection cannot run at the same time as deduplication. During the configured Garbage Collection window, data waiting for deduplication will be queued until Garbage Collection has completed. By default, the Garbage Collection window is from 13:00 through 19:00 on Monday through Thursday; however, you can revise this (for more information, refer to the Quest NetVault SmartDisk Administrator's Guide). Outside the Garbage Collection window, Garbage Collection is only started if no data is being deduplicated or is queued for deduplication. If data becomes available for deduplication outside the Garbage Collection window and Garbage Collection is active, Garbage Collection is stopped and deduplication proceeds.

Garbage Collection does add additional disk-access overhead to NVSD operations, which reduces performance while it runs. Quest Software recommends scheduling Garbage Collection on a regular basis to avoid delays associated with collecting large amounts of data.

Note: If a backup job that uses deduplication is aborted, be aware that the Chunk Store space used by segments that have already been deduplicated is not reclaimed until the next Garbage Collection (GC) occurs. Space used by the Staging Store and license enforcement is reclaimed immediately.


0


21

Chapter 3:

PLANNING YOUR NVSD DEPLOYMENT

NVSD Deployment Strategy An Overview Defining Which Data to Deduplicate

Identifying the Ideal Targets for Deduplication Understanding the Cost of Restoring Deduplicated Data Identifying Targets for Nondeduplication Deduplicating Similar Data Together Separating vs. Combining Deduplicated Data in NVSD Instances

Determining the Size of the NVSD Deployment Calculating NVSD License Capacity Calculating the Required Number of NVSD Instances Calculating the Total Required Physical Disk Space for Licensed Capacity

Choosing a Deployment Location for an NVSD Instance Reviewing Deployment Options for NVSD Nondeduplicated Instances Reviewing Deployment Options for NVSD Deduplicated Instances Reviewing Deployment Options for Disaster Recovery

Defining an NVSD Storage Pool Strategy Understanding Space Allocation for NVSD Storage Pools Understanding the Simplified Management Strategy Understanding the Optimal Performance Strategy

2 Chapter 3Planning Your NVSD Deployment

2

3.1.0 NVSD Deployment Strategy An OverviewAn NVSD Deployment Strategy is comprised of multiple components, which you define prior to installing NVSD. To ensure a successful deployment, complete the following steps, which are described in detail in subsequent sections:

1. Define the data that you want to deduplicate.2. Calculate the size of your NVSD Deployment.3. Define the deployment location of the NVSD Instance.4. Define a strategy for the NVSD Storage Pools.

NVBU is used as an example throughout this chapter; however, you can use any product that works with NVSD.

3.2.0 Defining Which Data to DeduplicateThe first step in defining your NVSD Deployment Strategy is determining which data to deduplicate. Not all data benefits from deduplication; therefore, it is important to determine which data will be targeted to use NVSDs Deduplication Option to reduce its storage footprint.

3.2.1 Identifying the Ideal Targets for DeduplicationHigh deduplication ratios are achieved when multiple generations of the same data are targeted for the same NVSD Instance. Data that is an ideal target for deduplication includes the following:

Structured databases, such as Oracle and SQL Server databases that are protected by Quest NetVault Backup (NVBU) Plug-ins.

Unstructured file-system data, such as data that is stored on File Servers protected by NVBU Plug-in for FileSystem.

Workstation data, such as desktops and laptops protected by NVBU Workstation Client.

Virtual Machine (VM) images from the same OS and similar applications, such as those protected by vRanger or NVBU Plug-ins for VMware and Hyper-V.

Email servers, such as Exchange and Domino systems that are protected by NVBU Plug-ins for Exchange and Domino.

Note: While email is an ideal target for deduplication, Single Instance Store (SIS), such as that provided by Exchange, will reduce deduplicated data; therefore, the deduplication ratio for the email server will not be as high as other types of data.


23

3.2.2 Understanding the Cost of Restoring Deduplicated DataWhile data deduplication reduces storage costs by reducing the storage footprint, there is a cost incurred during the restore processes. During the restoration of a deduplicated backup, NVSD has to reassemble the chunks as it restores the data. This reassembly process, also called rehydration, lengthens the time to restore the data. Therefore, if the Recovery Time Objective (RTO) is important for a specific database, email, or file system, you must consider the trade-off between reducing storage costs and increasing RTO when you identify which data to deduplicate.

3.2.3 Identifying Targets for NondeduplicationData that does not deduplicate well and should not be deduplicated includes: Encrypted data (this is because the data stream is unique) Data with demanding RTOs

3.2.4 Deduplicating Similar Data TogetherYou can increase deduplication ratios by targeting backups from the same database, file system, or application to the same NVSD Instance. When a backup is deduplicated and a previous backup from the same database, file system, or application has already been deduplicated by the NVSD Instance, only the unique or new chunks that did not exist in the previous backup will have to be stored in the Chunk Store. If a previously deduplicated backup does not exist in the NVSD Instance, the majority of the backup is considered unique data, which increases the number of unique chunks that have to be stored in the Chunk Store.When targeting backups to NVSD Instances, deduplication ratios will decline if backups are targeted to random NVSD Instances. Quest Software recommends that you target backups from the same database, file system, or application to the same NVSD Instance.

3.2.5 Separating vs. Combining Deduplicated Data in NVSD InstancesAn entire NVSD is licensed as either nondeduplicated or deduplicated capacity. When the data stored in NVSD is a combination of nondeduplicated and deduplicated data, two deployment options are available:

Create separate NVSD Instances, one for nondeduplicated data and one for deduplicated data.

Create a single NVSD Instance in which nondeduplicated and deduplicated data reside together in the Storage Pools.


4

3.2.5.a Using Separate NVSD Instances for Nondeduplicated and Deduplicated DataWhen the data targeted for NVSD is a combination of nondeduplicated and deduplicated data, you can create multiple NVSD Instances: one for nondeduplicated data and one for deduplicated data. In this case, the Nondeduplicated Instance is licensed as NVSD Capacity and the Deduplicated Instance is licensed as NVSD Capacity with Deduplication Option.This option lets you purchase the Deduplication Option for only the capacity that will be deduplicated. In this type of deployment, it is your responsibility to ensure that the correct NVSD Instance is targeted for backup. This ensures that data intended for deduplication is deduplicated and that data not intended for deduplication is stored as nondeduplicated data.

Advantages The Deduplication Option only has to be purchased for deduplicated NVSD

capacity, which is more cost effective. Nondeduplicated NVSD Instances have lower hardware requirements.

Disadvantages You must deploy and manage multiple NVSD Instances. Backups must explicitly target a deduplicated or nondeduplicated NVSD

Instance or Device to ensure that data gets deduplicated or not.

3.2.5.b Using a Single NVSD Instance for Nondeduplicated and Deduplicated DataWhen the data targeted for NVSD is a combination of nondeduplicated and deduplicated data, you can create a single NVSD Instance.

This option requires that you purchase the Deduplication Option for the entire NVSD Instance, including the capacity for the backups that are not deduplicated.

Advantages Simplifies deployment by reducing the number of NVSD Instances. Backups can target the same NVSD Instance whether or not the backup is

enabled for deduplication.

Disadvantage The NVSD Deduplication Option must be purchased for both

nondeduplicated and deduplicated NVSD Capacity.


25

3.3.0 Determining the Size of the NVSD DeploymentThe second step in defining your NVSD Deployment Strategy is to determine the size of your NVSD Deployment. To determine the size of your deployment, complete the following steps, which are described in detail in subsequent sections:

1. Calculate the NVSD Licensed Capacity.2. Calculate the required number of NVSD Instances.3. Calculate the total required physical disk space for licensed capacity.

3.3.1 Calculating NVSD License CapacityNVSD is licensed based on the Logical Capacity or Total Size (before deduplication) of all backups stored in NVSD.NVSD provides the ability to deploy multiple instances, which can be targeted for backup from multiple servers. When calculating the licensed capacity for NVSD, the capacity across all deployed NVSD Instances must be included. For example, if two servers are targeting backups to one or more NVSD Instances, the capacity of the backups from both servers must be included in the calculation.

NVSD Licensed Capacity is not based on the following: The actual size of the Storage Pool, Staging Store, or Chunk Store. The actual size of the backups after deduplication.

Before calculating the required NVSD Capacity, ensure that the backups being targeted to NVSD are compatible with NVSD. For more information, refer to the Quest NetVault SmartDisk Release Notes.

Calculating NVSD Capacity requires the collection of the following values for each server that will target backups to NVSD. When inserting these values into the calculations be sure to include the sum of the values across all servers that will target backups to NVSD. For example, if NVBU Server 1s Size of Weekly Backups is 40GB and NVBU Server 2s is 60GB, use 100GB as the Size of the Weekly Full Backups.

If separate NVSD Instances for Nondeduplicated and Deduplicated Data will be used, perform the following calculations separately for the backups that will not be deduplicated versus the backups that will be deduplicated. This enables you to determine how much NVSD Nondeduplicated capacity needs to be licensed and how much capacity needs to be licensed for the NVSD Deduplication Option.

If a single NVSD Instance will be used for both nondeduplicated and deduplicated data, include both the values for nondeduplicated and deduplicated data in the single calculation.Note: Be aware of the differences between Growth Rate and Change Rate. Change Rate refers to the frequency with which blocks change in a volume.


6

Growth Rate refers to the frequency at which new blocks are added to a volume. Although Change Rate includes all blocks associated with Growth Rate, Growth Rate does not necessarily include all blocks associated with Change Rate.

Size of Weekly Full Backups This is the size of all Full Backups that will be targeted to NVSD. For example, if SQL Server, Exchange, Oracle, and File System backups will be targeted for NVSD, calculate the total size of all the Full Backups for SQL Server, Exchange, Oracle, and File System.

Weekly Full Backup Growth Rate This is the average weekly growth rate of the Full Backups that are included in the Size of Weekly Full Backups calculation. Weekly Full Backup Growth Rate is critical to determining the size the NVSD Deployment. For example, the average of all Full Backups is growing 10% each week. Another method for calculating this is to base it on annual data growth. For example, an annual growth rate of 100%, or an annual doubling of data, represents a weekly growth rate of 1.333%.

Weekly Full Backup Retention Period in Weeks This is the number of weeks that Full Backups are retained in NVSD. Restores are faster when performed from disk-based media than when performed from tape-based media. Increasing the amount of protected data that is available on disk-based media increases the number of restore scenarios that can be performed from disk-based media. Requiring tapes to be located, retrieved, and loaded to perform a restore slows down the restore speed and increases downtime.Additionally, when performing deduplication, the longer the data is retained in NVSD, the better the deduplication ratios will be because more duplicate chunks are found thereby enabling the ability to pack more data into the same storage footprint. This enables even more protected data to be available via disk-based media.To obtain the most ideal deduplication ratios, Quest Software recommends a retention period of 12 weeks or more.

Size of Daily Backups This is the size of all Daily Backups that will be targeted to NVSD. For example, if SQL Server, Exchange, Oracle, and File System backups will be targeted for NVSD, calculate the total size of all Incremental Backups for SQL Server, Exchange, Oracle, and File System.Daily Backups are interim backups created between the Weekly Full Backups. These are anticipated to be Incremental or Differential Backups, and as such, will generally be much smaller than the Weekly Full Backups.

Daily Backup Growth Rate This is the average daily growth rate of Daily Backups that are included in the Size of Daily Backups. For example, the average of All Daily Backups is growing 1% each day. Number of Daily Backups between Full Backups This is the number of Daily Backups that are performed between Full Backups. For example, if Full


27

Backups are performed on Sunday and Daily Backups are performed every day, Monday through Saturday, the Number of Daily Backups is six.

Daily Backup Retention Period in Weeks This is the number of weeks that Daily Backups will be retained in NVSD. Daily Backups provide the ability to perform fixed-point-in-time restores, thereby reducing the RTO. Daily Backups can be retained for the same period as Full Backups, which enables fixed-point-in-time restores for the entire Weekly Full Backup Retention Period. Fixed-point-in-time restores are typically performed using backups taken within the last four weeks. As such, it is possible to retain Weekly Full Backups for one retention period and Daily Backups for a shorter period. This allows you access to Daily Backups for a set period, such as four weeks, while providing you access to Full Backups to a longer period, such as 12 weeks.

3.3.1.a Using the Logical-Capacity FormulaUse the following formula to calculate the total amount of NVSD Capacity that is required for all servers that will target backups to NVSD.You can determine the Logical Capacity by calculating the Full Backup Capacity, calculating the Daily Backup Capacity, and adding the totals.

Total NVSD Capacity to License = Full Backup Capacity + Daily Backup Capacity

Full Backup Capacity

Daily Backup Capacity

Alternatively, you can use the Quest NetVault SmartDisk Sizing Calculator to perform this calculation. The Quest NetVault SmartDisk Sizing Calculator is available from your Quest Software representative.

(Size of Weekly Full Backups)

* ((1+ Weekly Full Backup Growth Rate) ^ (Weekly Full Backup Retention Period + 1)

- 1.0) / (Weekly Full Backup Growth Rate)

(Size of Daily Backups)

* (((1 + Daily Backup Growth Rate) ^ ((Daily Backup Retention Period + 1)

* Number of Daily Backups between Weekly Backups)) - 1.0) / (Daily Backup Growth Rate)


8

Logical Capacity Calculation Example100GB = Size of Weekly Full Backups

10% = Weekly Full Backup Growth Rate

12 = Weekly Full Backup Retention Period in Weeks

10GB = Size of Daily Backups

1% = Daily Backup Weekly Growth Rate

6 = Number of Daily Backups between Full Backups

4 = Daily Backup Retention Period in WeeksFull Backup Capacity

Daily Backup Capacity

In the preceding example, 2.72TB of NVSD Capacity is required. Because NVSD Capacity is sold in 1TB increments, 3TB of NVSD Capacity must be licensed.

3.3.2 Calculating the Required Number of NVSD Instances

3.3.2.a Determining the Number of Deduplicated NVSD InstancesThree factors determine the number of NVSD Instances with the NVSD Deduplication Option enabled. If any of these factors are true, you must deploy multiple NVSD Instances. To determine the total number of NVSD Instances for the licensed NVSD Capacity with Deduplication Option, perform the following calculations:

Unique Data Size > 15TB for 64-bit OS and > 5TB for 32-bit OS Ingest Rate > Maximum Sustained Network Bandwidth

(100) * (((1+ .10) ^ (12 + 1) - 1.0) / (.10)) =

(100) * (((1.1) ^ (13) - 1) / (0.1)) =

(100) * (((3.45) - 1) / (0.1)) =

(100) * (2.45 / 0.1) = 2450 GB

(10) * (((1 + .01) ^ ((4 + 1) * 6)) - 1.0) / (.01) =

(10) * (((1.01) ^ ((5) * 6)) - 1) / (0.01) =

(10) * (((1.01) ^ 30) - 1) / (0.01) =

(10)*((1.34)-1) / (0.01) =

(10)*(0.34/0.01) = 340GB Deduplication Rate < Ingest Rate


29

The maximum result for the three calculations determines the number of NVSD Instances that must be deployed. For example, if the Unique Data Size calculation and the Deduplication Rate indicates that only one NVSD Instance is required, but the Ingest Rate calculation indicates that two NVSD Instances are required, you must deploy two NVSD Instances.

Unique Data Size > OS Bit LimitThe Unique Data Size is the amount of unique data that is stored inside the NVSD Instance. An additional NVSD Instance must be deployed for every 15TB of unique data on a 64-bit OS and for every 5TB of unique data on a 32-bit OS. Use the following calculation to determine the number of NVSD Instances based on the Unique Data Size:

Example:100GB = Size of Weekly Full Backups

10% = Weekly Change Rate



4 = Daily Backup Retention Period in Weeks


If the OS is 64-bit:460/15360 = .029

Rounded Up to Next Whole Number = 1 NVSD Instance If the OS is 32-bit:

460/5120 = .089

Size of Weekly Full Backups + ((Size of Weekly Full Backups * Weekly Change Rate)

* Weekly Full Backup Retention Period)

+ (Size of Daily Backups * (Number of Daily Backups between Weekly Full Backups

* Daily Backup Retention Period))

100 + ((100 * 0.1) * 12) + (10 * (6*4)) =

100 + ((10) * 12) + (10 * (24)) =

100 + (120) + (240) = 460GBRounded Up to Next Whole Number = 1 NVSD Instance


0

Ingest Rate > Network BandwidthThe NVSD Ingest Rate is the rate at which data can be streamed into NVSD. Because data streams into the NVSD Instance through a user-defined port, the network bandwidth will limit the speed at which data can be streamed. To ensure that all backups targeted to an NVSD Instance can complete during the backup window, you must divide the required ingest rate by the Maximum Sustained Network Bandwidth to determine the number of required NVSD Instances. Use the following calculation to determine the number of NVSD Instances based on the NVSD Ingest Rate and Network Bandwidth:


8 = Number of Hours in Backup Window

100 = Maximum Sustained NetWork Bandwidth (MB/Sec)

Rounded Up to Next Whole Number = 1 NVSD Instance

Deduplication Rate < Ingest RateThe NVSD Deduplication Rate is the rate at which NVSD deduplicates data. If NVSD deduplication is configured to run during the backup, the NVSD Deduplication Rate must be greater than or equal to the Ingest Rate to ensure that the deduplication process completes before the next weekly backup window begins.

Use the following to estimate the Deduplication Rates:

1. Determine the Single Core Deduplication Rate:

((Size of Weekly Full Backups * 1024) / (Number of Hours in Backup Window* 60 * 60))

/ Maximum Sustained Network Bandwidth

((100 *1024) / (8 * 60* 60)) / 100 =

((102400) / (28,800)) / 100 =

3.55 / 100 = .035

Hardware Speed24 to 36 Month Old Intel Core 1, AMD Athlon or Intel Celeron 10 to 20 MB/sec

12 to 24 Month Old Core 2 Duo, AMD Quad Core Opteron or Dual Core Xeon

20 to 40 MB/sec

0 to 12 Month Old Desktop or Server Core 2 Duo, Quad Core Xeon 40 MB/sec or more

or Quad Core AMD


31

2. Multiply the Single Core Deduplication Rate by the number of Cores.Use the following calculation to determine the number of NVSD Instances based on the Deduplication Rate and Ingest Rate:


10% = Weekly Full Backup Growth Rate



24 = Number of Hours in Daily Deduplication Window

50 = Deduplication Rate (MB/Sec)


Determining the Required Number of NVSD InstancesThe maximum result for the following three calculations determines the number of NVSD Instances that must be deployed:

Unique Data Size > 15TB for 64-bit OS and > 5TB for 32-bit OS Ingest Rate > Maximum Sustained Network Bandwidth Deduplication Rate < Ingest Rate

Example: Unique Data Size > OS Bit Limit = 1 NVSD Instance Ingest Rate > Network Bandwidth = 1 NVSD Instance Deduplication Rate < Ingest Rate = 1 NVSD Instance

The maximum result is one NVSD Instance; therefore, only one NVSD Instance is required in this example.

(((Size of Weekly Full Backups) + (Size of Daily Backups * Number of Daily Backups between Weekly Full Backups)) * 1024)

/ (Number of Seconds in Deduplication Window for 6 Days) / Deduplication Rate

(((100) + (10 * 6)) * 1024) / (24 * 6 * 60 * 60)) / 50 =

(((100) + (60)) * 1024) / (518400)) / 50 =

((160*1024)/518400)/50) =

(163840/518400)/50) = .0063


2

3.3.2.b Determining the Number of Nondeduplicated NVSD InstancesThe number of Nondeduplicated NVSD Instances is determined by a single factor:

Ingest Rate > Maximum Sustained Network Bandwidth

Ingest Rate > Network BandwidthThe NVSD Ingest Rate is the rate at which data can be streamed into NVSD. Because data streams into the NVSD Instance through a user-defined port, the network bandwidth will limit the speed at which data can be streamed. To ensure that all backups targeted to an NVSD Instance can complete during the backup window, you must divide the required ingest rate by the Maximum Sustained Network Bandwidth to determine the number of required NVSD Instances. Use the following calculation to determine the number of NVSD Instances based on the NVSD Ingest Rate and Network Bandwidth:


8 = Number of Hours in Backup Window

100 = Maximum Sustained NetWork Bandwidth (MB/Sec)


3.3.3 Calculating the Total Required Physical Disk Space for Licensed CapacityNVSD is licensed based on the Logical Capacity of the data that it can store. However, in Deduplicated NVSD Instances, Logical Capacity does not match Physical Capacity or physical disk space. This is because the NVSD Deduplication Option packs up to 12 times more protected data into the same storage area for a 92% reduction in the storage footprint.

3.3.3.a Determining the Capacity for Deduplicated NVSD Instances

((Size of Weekly Full Backups * 1024) / (Number of Hours in Backup Window* 60 * 60))

/ Maximum Sustained Network Bandwidth

((100 *1024) / (8 * 60* 60)) / 100 =

((102400) / (28,800)) / 100 =

3.55 / 100 = .035A Deduplicated NVSD Instance can have a combination of both Deduplicated and Nondeduplicated data. In this configuration, calculating the total Physical Capacity


33

or physical disk space is achieved by calculating the Physical Capacity for the Deduplicated Backups, calculating the Physical Capacity for the Nondeduplicated Backups, and adding the sums.

Deduplicated BackupsThe Physical Capacity or physical disk space required for Deduplicated Backups in Deduplicated NVSD Instances is equal to the Size of Weekly Full Backups plus the Unique Data Size. Use the following formula to calculate it:

Example100GB = Size of Deduplicated Weekly Full Backups across all servers

10% = Weekly Change Rate


10GB = Size of Deduplicated Weekly Daily Backups



Nondeduplicated BackupsUse the following formula to calculate the Physical Capacity or physical disk space required for Nondeduplicated Backups in a Deduplicated NVSD Instance:

(Size of Weekly Full Backups) +

(Size of Weekly Full Backups + ((Size of Weekly Full Backups * Weekly Change Rate)


+ (Size of Daily Backups * (Number of Daily Backups between Weekly Full Backups

* Daily Backup Retention Period)))

100 + (100 + ((100 * 0.1) * 12) + (10 * (6*4))) =

100 + (100 + ((10) * 12) + (10 * (24))) =

100 + (100 + (120) + (240)) = 560GB

(Size of Nondeduplicated Weekly Full Backups * Weekly Full Backup Retention Period)

+ (Size of Nondeduplicated Daily Backups * (Number of Daily Backups between Full Backups * Daily Backup Retention Period in Weeks))


4

Example:10GB = Size of Nondeduplicated Weekly Full Backups


1GB = Size of Nondeduplicated Daily Backups



Total Required Disk Space = Deduplicated Backup Disk Space + Nondeduplicated Backup Disk Space

560GB + 144GB = 704GB

The Total Required Disk Space is divided into the Staging Store and the Chunk Store. If you intend to use different file systems or disks for the Staging Store and the Chunk Store, it is important to know how much of the Total Required Disk Space will be allocated to the Staging Store versus the Chunk Store. Use the following calculations to determine this.

Calculating Staging Store SizeThe amount of time a Deduplicated Backup remains in the Staging Store is different than the amount of time a Nondeduplicated Backup remains in the Staging Store. Therefore, to determine the total disk space required for the Staging Store, you must calculate the size required for both Deduplicated Backups and Nondeduplicated Backups.

Total Staging Store Disk Space = Deduplicated Backup Disk Space + Nondeduplicated Backup Disk SpaceDeduplicated Backups

During the NVSD Deduplicated Backup process, backups are initially streamed to the Staging Store where they remain until the deduplication process is finished. The Staging Store must be large enough to hold all the Deduplicated Backups while they await deduplication.

If the deduplication window overlaps the backup window, deduplication and backups will occur at the same time. Each backup starts deduplicating after the backup job is finished. Additionally, large backups will be broken into 400GB segments. Each segment starts to deduplicate after it is received, even if the backup stream is still being streamed to the Staging Store. Therefore, the

(10 * 12) + (1 * (6 * 4)) =

(120) + (1*(24)) =

120 * 24 = 144GBStaging Store only requires space to hold some portion of the total backup size. If


35

the Staging Store runs out of space, backups will pause until space becomes available, such as when a deduplication operation finishes.

The maximum disk space required for the Staging Store is equal to the Size of the Weekly Full Backups.Nondeduplicated BackupsBackups that are not deduplicated are streamed into the Staging Store, where they remain until they are retired. Therefore, calculating the size of the Staging Store for Nondeduplicated Backups is identical to calculating the required physical disk space for Nondeduplicated Backups in a Deduplicated NVSD Instance.

Example:10GB = Size of Nondeduplicated Weekly Full Backups


1GB = Size of Nondeduplicated Daily Backups



Calculating Chunk Store SizeThe amount of physical disk space required for the Chunk Store is equal to the Unique Data Size. To calculate the Unique Data Size or the size of the Chunk Store, use the following formula:

(Size of Nondeduplicated Weekly Full Backups * Weekly Full Backup Retention Rate)

+ (Size of Nondeduplicated Daily Backups * (Number of Daily Backups between Full Backups * Daily Backup Retention Period in Weeks))

(10 * 12) + (1 * (6 * 4)) =

(120) + (1*(24)) =

120 * 24 = 144GB

Size of Weekly Full Backups + ((Size of Weekly Full Backups * Weekly Change Rate)


+ (Size of Daily Backups * (Number of Daily Backups between Weekly Full Backups* Daily Backup Retention Period))


6






3.3.3.b Determining the Capacity for Nondeduplicated NVSD InstancesThe Physical Capacity or physical disk space required for Nondeduplicated NVSD Instances is equal to the Logical Capacity because deduplication has been disabled. The Nondeduplicated NVSD Instance will consume the same amount of disk space for the Staging Store as Logical Capacity that is being licensed.

3.4.0 Choosing a Deployment Location for an NVSD InstanceThe third step in defining your NVSD Deployment Strategy is deciding where you want to deploy your NVSD Instances. Whether the NVSD Instance is enabled for deduplication dictates the available deployment options.

3.4.1 Reviewing Deployment Options for NVSD Nondeduplicated InstancesYou can deploy NVSD Instances that do not have deduplication on the server, a heterogeneous client, or a dedicated server, and they can accept data streams from heterogeneous platforms. When multiple NVSD Instances are deployed, Storage Pools and processes are not shared across the multiple NVSD Instances. However, you cannot deploy multiple NVSD Instances on the same machine, whether it is a backup server, backup client, or a dedicated NVSD Server.

The following sections describe several deployment options that you might consider for deploying NVSD Nondeduplicated Instances. This is not a comprehensive list. These descriptions use NVBU only as an example.

100 + ((100 * 0.1) * 12) + (10 * (6*4)) =

100 + ((10) * 12) + (10 * (24)) =

100 + (120) + (240) = 460GB


37

3.4.1.a NVSD Nondeduplicated Instance Deployed on a Single NVBU ServerIn the simplest deployment, when only one NVSD Nondeduplicated Instance is required for an NVBU Domain with a single NVBU Server, the NVSD Instance is deployed on the single NVBU Server that uses file-system paths accessible by the NVBU Server for the NVSD Storage Pools. The single NVBU Servers Clients that target backups to NVSD will stream backup data on a user-defined port to the NVSD Instance where the backup is stored in the NVSD Storage Pools.

Figure 3-1:Single NVSD

Instance in single NVBU

Server environment

NetVault BackupServer

NetVault BackupClient

= Backup/Restore Data Path


NetVault BackupSmartClient

LAN

NetVault SmartDiskStorage Pools

NVSD Nondeduplicated

Instance


8

3.4.1.b NVSD Nondeduplicated Instance Deployed in a Multiple NVBU Server EnvironmentIn NVBU environments where only one NVSD Instance is required for an NVBU Domain with multiple NVBU Servers, a single NVSD Instance can accept data streams from multiple NVBU Servers. In the following example where a single NVSD Instance is deployed for multiple NVBU Servers, the NVBU Clients from both NVBU Servers will stream backup data on a user-defined port to the NVSD Instance that is deployed on one of the NVBU Servers.


Instance in multiple NVBU

Server environment

In NVBU environments where multiple NVSD Instances are required for an NVBU Domain with multiple NVBU Servers, you can deploy one NVSD Instance on each of the NVBU Servers, or you can deploy one NVSD Instance on one NVBU Server and on one or more NVBU Clients. In the following example where two NVSD Instances are required and there are two NVBU Servers, a single NVSD Instance is deployed on each of the NVBU Servers. The NVBU Clients from both NVBU Servers can stream backup data on a user-defined port to either NVSD Instance that is deployed.

Note: You cannot deploy multiple NVSD Instances on the same dedicated NVSD Server.









LAN



NVSDNondeduplicated

Instance


39

Figure 3-3:Multiple NVSD

Instances in multiple NVBU

Server environment









LAN


Instance



NVSDNondeduplicated

Instance


0

3.4.1.c NVSD Nondeduplicated Instance Deployed on a Dedicated ServerIn NVBU environments where only one NVSD Instance is required for an NVBU Domain with one or more NVBU Servers and you want to have the backups across all the NVBU Servers stored together, you can deploy a single NVSD Instance on a dedicated server that has no NVBU Server or Client software installed. In the following example, data from all NVBU Clients will stream backup data on a user-defined port to the single NVSD Instance that is deployed on the dedicated server.


Server in multiple NVBU

Server environment




NetVault SmartDiskServer






LAN




Instance


41

3.4.2 Reviewing Deployment Options for NVSD Deduplicated InstancesIn environments where only one NVSD Instance is required for a domain with one or more backup servers and you want the backups from all backup servers deduplicated together, you must deploy a single NVSD Instance on a dedicated server that has no backup server or client software installed.

Important: You cannot deploy a deduplicated instance on the same machine on which NVBU Server or Client software is installed.

In the following example, data from all NVBU Clients will stream backup data on a user-defined port to the single NVSD Instance that is deployed on the dedicated server.


Server in multiple NVBU

Server environment

In environments where multiple NVSD Instances are required for a domain, you must deploy each NVSD Instance on its own dedicated server. In the following example, two NVSD Instances are required and they are deployed on two dedicated NVSD Servers. The NVBU Clients from both NVBU Servers can stream backup data on a user-defined port to either deployed NVSD Instance.

Note: You cannot deploy multiple NVSD Instances on the same machine, whether it be a backup server, backup client, or dedicated NVSD Server.










LAN



NVSD Deduplicated

Instance


2Figure 3-6:

Multiple NVSD Servers in

multiple NVBU Server

environment

3.4.3 Reviewing Deployment Options for Disaster RecoveryThe data stored in an NVSD Instance is best protected from disaster if you deploy multiple NVSD Instances in a disaster-recovery scenario.

The primary NVSD Instance will be deployed on the backup server, backup client, or dedicated NVSD Instance. The secondary NVSD Instance will be deployed on a dedicated backup server and connected to the primary backup server as a second NVSD Device. The secondary NVSD Instance will typically be deployed in an offsite location to provide maximum protection.

The backups that are targeted to the primary NVSD Instance will be copied to the secondary NVSD Instance as the target for a secondary copy backup or as an independent Data Copy backup that specifies the primary NVSD Instance as the source and the secondary NVSD Instance as the destination.

Important: Third-party replication or mirroring technology is not supported for the replication of the NVSD Storage Pools. Data that is replicated using third-party technology, including deduplication appliance replication technology, is not recoverable.

If a failure of only the primary NVSD Instance occurs, you can restore backups from the secondary NVSD Instance. Primary backups can also be targeted to the secondary NVSD Instance until the primary NVSD Instance is ready to resume the primary role.









LAN


NetVault SmartDiskServers



NVSD Deduplicated

Instance

NVSD Deduplicated

Instance


43

Figure 3-7:Multiple Non-deduplicated

NVSD Instances

deployed in disaster-recovery scenario

Figure 3-8:Multiple

Deduplicated NVSD

Instances deployed in

disaster-recovery scenario

If a complete failure of the production site occurs or a failure of the primary backup server occurs, a backup server with the identical Machine Name as the primary backup server in the production site will need to be deployed in the disaster-recovery site. After the backup server with the identical Machine Name is running in the disaster-recovery site, the secondary NVSD Instance can be added to the

LAN

DisasterRecovery

Site

Production Site

PrimaryNetVault Backup

Server

StandbyNetVault Backup

Server



NetVault Backup Client

NetVault Backup Workstation

Client



= Backup/Restore Data Path= DR Data Path

NVSDNondeduplicated

Instance

NVSDNondeduplicated

Instance

LANDisasterRecovery

SiteProduction

SitePrimary


StandbyNetVault Backup

Server




NetVault Backup Client

NetVault Backup Workstation

Client



= Backup/Restore Data Path= DR Data Path

NVSDDeduplicated

Instance

NVSDDeduplicated

Instancebackup server as an NVSD Device and scanned. After the backups have been scanned, they can be restored and the secondary NVSD Device can become a


4

target for primary backups until the primary backup server is returned to working order.

Important: An NVSD Device can only be scanned into a backup server that has the identical Machine Name as the original backup server that performed the backups.

After the primary backup server is ready to resume the production role, the secondary backup server is shut down. If backups were targeted to the secondary NVSD Instance while it performed the role of the primary NVSD Instance, it will need to be scanned back into the primary backup server.

Note: The primary and secondary backup servers cannot be running at the same time.

3.5.0 Defining an NVSD Storage Pool StrategyThe fourth step in defining your NVSD Deployment Strategy is defining your strategy for the NVSD Storage Pools.

3.5.1 Understanding Space Allocation for NVSD Storage PoolsBefore choosing a strategy for your NVSD Storage Pools, consider the following with regard to how NVSD allocates space in the NVSD Storage Pools:

You can add an unlimited number of Storage Pools to NVSD. You can add an unlimited number of volumes to each NVSD Storage Pool. Each NVSD Storage Pool is associated with roles that it can perform. When configuring a NVSD Storage Pool, you can identify it as a favorite for

specific roles (favour), and you can indicate that it should never be used for specific roles (deny).

There is no quota specified per-volume of the absolute amount of space NVSD can use (that is, you cannot say use exactly 100GB). However, you can specify an amount of physical disk-storage that you want to set aside that cannot be used by NVSD. This simplifies NVSDs internal space-allocation policies so that Staging Store and Chunk Store processes can efficiently share volumes. Each role independently implements and honors the configured thresholds, which allows efficient parallel-processing without a requiring a single, centralized space allocator.

Using the volume-size parameter, you can set aside a portion of the volume so that it cannot be used for NVSD. After the volume is full with NVSD or non-

NVSD data, NVSD will not store anything else on the volume until space is made available.


45

NVSD uses storage space in each volume up to a threshold amount that includes space directly set aside using the volume-size parameter, as well as the Last Resort Threshold (LRT) amount set by NVSD internally.

NVSD first uses a volume for its intended role (for example, Chunk Store or Staging Store), but it will also use it for other purposes if necessary. Even if you supply a threshold, NVSD implements an LRT that stops NVSD from using a volume after there is less than a certain amount of space available and thus prevents the disk from becoming completely full. The LRT is calculated to include both a fixed amount of space (1GB) and the space that would be required to make Garbage Collection possible (that is, the gc_reserve_bytes parameter). This means that the figure varies in size depending on the amount of data held in the Chunk Store (typically, you can expect the gc_reserve_bytes parameter to be at least 1.7GB).

If the LRT is large enough, NVSD properly allocates space when multiple Staging Store and Chunk Store processes are both writing to a volume that becomes too full. NVSD maintains an overall idea of how much space is used both by the Staging Store and the Chunk Store. This allows it to enforce the licensed amount of capacity and to enforce the LRT and the volume-size limits.

Licensing is not affected by user thresholds or NVSD's LRT. Licensing measures the amount of data accepted for protection by the NVSD Instance, regardless of whether the data has been deduplicated.

The following figure shows a graphical example of how space is organized in NVSD. This figure assumes all volumes are mapped to one file system. Additionally, the space reserved for LRT is reserved on a per-volume basis; if your configuration uses multiple volumes, more LRT space might be allocated.

Figure 3-9:Space

allocation for NVSD

Unused space available for NVSD

Chunk Store

Staging Store

Garbage Collection Reserve

LRT

((100-X)/100)*T

Space currently used by OS, users, and non-NVSD applications

T=Total space available on an empty file system

X=Volume-size parameter


6

3.5.2 Understanding the Simplified Management StrategyIf the goal is simplified management, configure a single volume for all required storage. For example, if you are using a single RAID array, configure a single RAID 5 or RAID 6 volume, and map all the NVSD Storage Pool roles to a single file system that is mapped to the RAID array. This provides adequate performance while minimizing management complexity.

3.5.3 Understanding the Optimal Performance StrategyConsider the following guidelines when assigning NVSD Storage Pool Roles:

Content Index - Will be relatively small and should reside in a Storage Pool comprised of fault-tolerant disks with good random-access performance.

Staging Store Should reside in a Storage Pool comprised of fault-tolerant disks with good streaming performance.

Chunk Index Should reside in a Storage Pool comprised of fault-tolerant disks with good random-access performance.

Chunk Store Should reside in a Storage Pool comprised of fault-tolerant disks.

If the goal is optimal performance, using additional disks to increase Input/Output Operations per Second (IOPS) and aggregate disk bandwidth improves NVSD performance more than using additional memory.

For example, if you are using a single RAID array, consider configuring four separate RAID 1 volumes, one for each Storage Pool Role. This separates the IO workload for each volume across independent RAID volumes. You can tune performance further by restricting the number of independent disk operations (reads, writes, and deletes) allowed per volume to two operations, which avoids disk thrashing. In addition, configure the RAID array to support more bandwidth in and out of Staging Store.This strategy is appropriate for NVSD Instances where staging, deduplication, and restores might occur simultaneously.

Note: RAID recommendations also apply to Storage Area Network (SAN) and Network Attached Storage (NAS).

Additionally, Linux, UNIX, and Mac OS X systems support use of the noatime feature for file systems, usually as a mount option. Quest Software recommends that you enable this feature on your system to improve NVSD performance, especially for the Chunk Index and Chunk Store volumes. Using this feature can reduce the number of metadata writes needed to update read-access times for files.On Windows platforms, you can disable the New Technology File System (NTFS) Last Access Update feature, which can reduce disk accesses and increase


47

performance. For instructions on disabling this feature, refer to the documentation for your specific OS.

3.5.3.a Optimizing Performance While Protecting Against Data Loss During Power FailuresAlthough enabling disk-write caches improves NVSD performance, power failures that occur before modified disk-cache contents have been written to non-volatile magnetic storage can potentially cause data loss in NVSD. Because of this risk, it is critical that you understand how your underlying disk technology caches writes to disk. You can turn off disk-write caching completely, but due to the improvement in performance that write-caching offers, it is increasingly used despite the risk, and the risk is mitigated through the use of additional technology. A common mitigation technique is ensuring that power does not go off. In high-end server environments, with their uninterruptible power supplies (UPSs) and redundant power supplies, having unfilled cached writes is less of an issue.

Additionally, drives that employ write-caching have a write-flush feature that instructs the drive to send pending writes from the cache to the disk immediately. This command is usually sent before UPS batteries run out (if the system detects a power interruption) or just before the system is shut down for any other reason.

Finally, most disk array systems use non-volatile random access memory (NVRAM) to protect data written to disk in case of a power failure. Quest Software recommends that you review your server and storage vendors product documentation to understand what steps are taken to ensure that disk writes are written to non-volatile magnetic storage in case of a power failure.


8


49

Chapter 4:

INSTALLING NVSD Installing NVSD An Overview System Requirements

Software Requirements Hardware Requirements Authentication Requirements

Installing the NVSD Software Important Considerations Installing on Linux or UNIX (Excluding Solaris) Installing on Solaris (SPARC, x86, or x86-64) Installing on Windows Installing on Mac OS X

4.1.0 Installing NVSD An OverviewInstalling NVSD involves the following main steps:

Verifying that the system requirements are in place. Installing the software. Updating the Storage Pool structure, if applicable. Adding the NVSD Devices to your backup product. Protecting the NVSD identify. Installing the license key.

4.2.0 System Requirements

4.2.1 Software RequirementsNVSD has the following software requirements:

Windows platforms NVSD has the following requirements on Windows: Microsoft Visual C++ If it is not already installed, NVSD automatically

installs the applicable portions of the Microsoft Visual C++ 2005 SP1 Redistributable Package. No additional steps are required by you.

5 Chapter 4Installing NVSD

0

System-memory requirements Make sure that your Windows system has at least 2 gigabytes (GB) plus 400 megabytes (MB) for every terabyte (TB) of data that the Chunk Store must address. If you use a 32-bit platform, be aware that you are limited to a Chunk Store of 2.5TB of unique data.

Linux platforms NVSD has the following library requirements on Linux. Installation might fail if these packages are not installed on the system. For instructions on installing the required packages, refer to the relevant OS documentation. Asynchronous I/O Library Depending on the Linux distribution, make

sure that the libaio.so.1 package is installed in the following directory: 32-bit OS /lib or /usr/lib 64-bit OS /lib64 or /usr/lib64Some distributions of Linux do not install the Asynchronous I/O Library package by default. If required, install this package.

Standard C++ Library NVSD Linux x86_Hybrid and Linux IA64 Builds Make sure that

the libstdc++.so.6 package is installed in the /usr/lib directory. If required, install the Standard C++ Library package.

NVSD Linux x86_Pure64 Build Make sure that the libstdc++.so.5 package is installed in the /usr/lib64 directory. If required, install the Compatibility Standard C++ Library package.

4.2.2 Hardware RequirementsNVSD has the following hardware requirements:

Deduplicated NVSD Instances If the NVSD Deduplication Option is enabled, make sure that the following requirements are met: CPU requirements CPU requirements vary based on the required

NVSD Deduplication Rate (the rate at which NVSD deduplicates data) the higher the Deduplication Rate requirements, the higher the CPU requirements for the machine where NVSD software is installed.

The NVSD Deduplication Rate must be greater than or equal to the Ingest Rate to ensure that the deduplication process completes during the backup window or before the next weekly backup window begins. For more information, see Calculating the Required Number of NVSD Instances on page 28.

Memory requirements: Required minimum 2GB

Recommend minimum An additional 400MB per 1TB of NVSD-

unique data


51

Example10TB of NVSD Unique Data2GB + (400MB * 10) = 2GB + 4000MB = 6GBFor more information, see Calculating the Required Number of NVSD Instances on page 28.

Nondeduplicated NVSD Instances If the NVSD Deduplication Option is not enabled, make sure that the following requirements are met: CPU requirements Use a machine with a 24-to-36-month-old Intel Core

1, AMD Athlon, or Intel Celeron processor. Quest Software recommends that you use faster processors for best performance.

Memory requirements The minimum requirement is 512MB. Network requirement for all NVSD Instances TCP/IP name resolution.

After the machine that will host the NVSD software has been identified, make sure that the machine can resolve its own host name.

Storage Pool requirements For Storage Pool requirements based on whether your goal is simplified management or optimal performance, see Defining an NVSD Storage Pool Strategy on page 44.

For additional information, refer to the Quest NetVault SmartDisk Supported Platforms document.

4.2.3 Authentication Requirements Root-level or administrator account for installation Root-level or

administrator credentials are only required for use during the installation process, removal process, and, if necessary, manually restarting the NVSD Services.

NVSD Non-privileged Group and User Security group with full NVSD configuration and monitoring rights.To create the NVSD Non-privileged Group and Users, follow these guidelines: For Linux and UNIX (the following examples use sdusr and sdgrp for

the user and group, respectively):1. If the non-privileged group does not exist, create it.

2. If the non-privileged user does not exist, create it and add it to the non-privileged group.

# groupadd sdgrp

# useradd -G sdgrp sdusr

5 Chapter 4Installing NVSD

2

3. Define the password for the non-privileged user.

4. If the non-privileged user does exist, add it to the non-privileged group.

For Windows:On Windows, you are not required to create the non-privileged user prior to installation. You can enter the applicable information during the installation process. The installer will verify that the account meets all applicable system requirements before continuing with the installation process.

For Mac OS X: Create the non-privileged user by clicking Apple > System

Preferences > Accounts. On the System Preferences dialog, click Accounts. Use the Accounts dialog to create a non-privileged user.

Use dseditgroup to add the non-privileged user to an applicable group. The following example uses sdusr and sdgrp for the user and group, respectively.

For Linux, UNIX, and Mac OS X platforms: If your OS does not create a corresponding non-privileged group by

default, create one and add the non-privileged user to it. After you have created the non-privileged user and group, add your

root-level or administrator account to the same non-privileged group. UNC path and credentials if using a network share on Windows You

can specify a path that includes a network share, whether it is mapped or using Universal Naming Convention (UNC). To do so, the parent directory of the target directory must exist (the target directory does not have to exist yet) and you must have the access credentials.

4.2.3.a Important Notes Regarding Windows and the Non-Privileged UserIf you are installing NVSD on Windows, be aware that the NVSD installer grants read- and write-access permission to the specified non-privileged user during the installation process. This access, which lets NVSD report Windows Events, only applies to the Application log.

# passwd sdusr

# usermod -a -G sdgrp sdusr

sudo dseditgroup -o create sdgrp

sudo dseditgroup -o edit -a sdusr -t user sdgrp


53

If you change the non-privileged user (via smartdisk.bat password), the same rights are granted to the new user, and the rights are retained for the original user because the original user might be assigned to multiple services on the same system. For the same reason, this access is also retained if you remove NVSD.

The modified registry key is:

HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\ Eventlog\Application

Therefore, if you know that a non-privileged user is not needed for other services, you should manually update the registry key and user account.

4.3.0 Installing the NVSD SoftwareYou can install NVSD either via the installation CD or the package downloaded from the web site.

Throughout this document, the path to where you installed NVSD is represented by an ellipsis (). Replace the ellipsis with the applicable information.

4.3.1 Important Considerations If you are using NVBU, do not install the product in the same directory

SOL90713_Quest_NetVault_SmartDisk_1.5.1_Installation_Upgrade_Guide_Rev1_0.pdf

Documents

Transcript of SOL90713_Quest_NetVault_SmartDisk_1.5.1_Installation_Upgrade_Guide_Rev1_0.pdf