Nimsoft Monitor oracle Guidedocs.nimsoft.com/prodhelp/en_US/Probes/Catalog/oracle/4.4/oracle.pdf ·...

oracle Guide v4.4 series

Nimsoft® Monitor™

Legal Notices Copyright © 2012, Nimsoft Corporation

Warranty

The material contained in this document is provided "as is," and is subject to being changed, without notice, in future editions. Further, to the maximum extent permitted by applicable law, Nimsoft Corporation disclaims all warranties, either express or implied, with regard to this manual and any information contained herein, including but not limited to the implied warranties of merchantability and fitness for a particular purpose. Nimsoft Corporation shall not be liable for errors or for incidental or consequential damages in connection with the furnishing, use, or performance of this document or of any information contained herein. Should Nimsoft Corporation and the user have a separate written agreement with warranty terms covering the material in this document that conflict with these terms, the warranty terms in the separate agreement shall control.

Technology Licenses

The hardware and/or software described in this document are furnished under a license and may be used or copied only in accordance with the terms of such license.

No part of this manual may be reproduced in any form or by any means (including electronic storage and retrieval or translation into a foreign language) without prior agreement and written consent from Nimsoft Corporation as governed by United States and international copyright laws.

Restricted Rights Legend

If software is for use in the performance of a U.S. Government prime contract or subcontract, Software is delivered and licensed as "Commercial computer software" as defined in DFAR 252.227-7014 (June 1995), or as a "commercial item" as defined in FAR 2.101(a) or as "Restricted computer software" as defined in FAR 52.227-19 (June 1987) or any equivalent agency regulation or contract clause. Use, duplication or disclosure of Software is subject to Nimsoft Corporation’s standard commercial license terms, and non-DOD Departments and Agencies of the U.S. Government will receive no greater than Restricted Rights as defined in FAR 52.227-19(c)(1-2) (June 1987). U.S. Government users will receive no greater than Limited Rights as defined in FAR 52.227-14 (June 1987) or DFAR 252.227-7015 (b)(2) (November 1995), as applicable in any technical data.

Trademarks

Adobe®, Acrobat®, Acrobat Reader®, and Acrobat Exchange® are registered trademarks of Adobe Systems Incorporated.

Intel® and Pentium® are U.S. registered trademarks of Intel Corporation.

Java(TM) is a U.S. trademark of Sun Microsystems, Inc.

Microsoft® and Windows® are U.S. registered trademarks of Microsoft Corporation.

Netscape(TM) is a U.S. trademark of Netscape Communications Corporation.

Oracle® is a U.S. registered trademark of Oracle Corporation, Redwood City, California.

UNIX® is a registered trademark of the Open Group.

Contact Nimsoft

For your convenience, Nimsoft provides a single site where you can access information about Nimsoft products.

At http://support.nimsoft.com/, you can access the following:

■ Online and telephone contact information for technical assistance and customer services

■ Information about user communities and forums

■ Product and documentation downloads

■ Nimsoft Support policies and guidelines

■ Other helpful resources appropriate for your product

Provide Feedback

If you have comments or questions about Nimsoft product documentation, you can send a message to [email protected].

http://support.nimsoft.com/

mailto:[email protected]

mailto:[email protected]

Contents 5

Contents

Chapter 1: oracle 4.4 7

Oracle Requirements.................................................................................................................................................. 10

Supported Oracle versions and clients ................................................................................................................ 10

Specific Notes ...................................................................................................................................................... 11

Known issues ....................................................................................................................................................... 11

Installation notes ................................................................................................................................................ 12

Upgrade from Oracle Probe v3.9x to v4.01 or Later ........................................................................................... 13

Oracle Configuration .................................................................................................................................................. 13

The Setup tab ...................................................................................................................................................... 14

The Connections tab ........................................................................................................................................... 17

The Profiles tab ................................................................................................................................................... 19

The Templates tab .............................................................................................................................................. 24

The Status tab ..................................................................................................................................................... 25

The Groups tab .................................................................................................................................................... 26

Creating a custom checkpoint .................................................................................................................................... 29

Thresholds ........................................................................................................................................................... 38

Schedules ............................................................................................................................................................ 39

Editing a checkpoint ................................................................................................................................................... 42

Thresholds ........................................................................................................................................................... 46

Schedules ............................................................................................................................................................ 47

Checkpoints metrics ................................................................................................................................................... 48

oracle Metrics ..................................................................................................................................................... 48

String matching methods ........................................................................................................................................... 55

Glossary 57


Chapter 1: oracle 4.4

This description applies to oracle probe version 4.4x.

The Oracle monitoring probe monitors local or remote Oracle database instances. This probe will periodically scan through a configurable set of monitoring profiles and apply checks to the local or remote Oracle instances. It is possible to adapt the configuration according to specific monitoring needs/requirements. The probe does not modify any tables in Oracle, and works with a standard Oracle installation.

The following information is extracted and monitored:

■ Database Availability

■ Tablespace Free

■ Tablespace Allocated Free

■ Fragmented Segments

■ Buffer cache-hit ratio

■ Dictionary cache-hit ratio

■ Library Cachehit Ratio

■ User Memory Consumtion

■ Number of rollback segments

■ Objects that are unable to allocate next extent

■ Invalid objects in database

■ Chained rows in database

■ Datafile Availability

■ Concurrent Users

■ Enqueue Timeouts

■ Redo Log contention

■ MTS Response

■ MTS Wait

■ SGA Memory Usage

■ Free SGA Memory

■ Sort Ratio

■ Index status

■ Tablespace Availability

Oracle Requirements

8 oracle Guide

■ Rollback Segment Availability

■ Check if next_extent is bigger than available space

■ Remaining number of extents

■ MTS UGA memory

■ User Locks

■ Number of Locked Users

■ User Buffer Cachehit Ratio

■ Session Waits counter

■ System Statistics

■ System Statistics Delta

■ System Waits counter

■ Database File I/O

■ Lock Wait Time

■ Global Cache Service Utilization

■ Global Cache Consistent Read Timeouts

■ Global Cache Convert Timeouts

■ Global Cache Blocks Lost

■ Global Cache Blocks Corrupt

■ Average Global Lock Get Time

■ Fusion Writes Ratio

■ Number of Long Running Queries

■ Tablespace Size

■ Database Size

■ Resource Utilization %

■ Dataguard Status

■ Dataguard Gap

■ Dataguard Timegap

■ Tablespace Temp Free

■ Active Users

■ Flash recovery area memory free

■ Active Connection Ratio

This section contains the following topics:

Oracle Requirements


Documentation Changes (see page 10) Oracle Requirements (see page 10) Oracle Configuration (see page 13) Creating a custom checkpoint (see page 29) Editing a checkpoint (see page 42) Checkpoints metrics (see page 48) String matching methods (see page 55)

Oracle Requirements

10 oracle Guide

Documentation Changes

This table describes the version history for this document.

Version Date What's New?

4.4x October 2011 Initial Release

Oracle Requirements

The minimal requirement of the oracle probe is:

■ Oracle Client Software

Supported Oracle versions and clients

The matrix below summarizes client and server combinations that are supported. The probe is built using 10.2.0 libraries.

Server Version

Client Version

11.2.0 11.1.0 10.2.0 10.1.0 9.2.0 9.0.1

11.2.0

11.1.0 #3

10.2.0 #3

10.1.0 (#4)

9.2.0 #3 #3

9.0.1

Key:

Yes Supported

No Not Supported

Oracle Requirements


Specific Notes

■ #1 - An ORA-3134 error is correctly reported when attempting to connect to this version.

■ #2 - There are problems connecting from a 10g client to 9i where one is EBCDIC based.

■ #3 - For connections between 10.2/11.1/11.2 (or higher) and 9.2, the 9.2 end MUST be at 9.2.0.4 or higher. Connections between 10.2 (or higher) and 9.2.0.1, 9.2.0.2 or 9.2.0.3 are not supported.

Known issues

The Oracle RAC system statistics used for 'gc_' checkpoints are, starting with Oracle 10.1 no more available. Therefore, these checkpoints will return no value if activated.

On Windows 64bit platform, the probe cannot be installed into the default directory - "Program Files (x86)". A bug in Oracle Client is causing connection errors, if the application home directory names includes special characters, like "(" (Oracle Bug 3807408).

Error ORA-12705 together with log entry "OCIEnvCreate failed with rc = -1" can happen, if environmental variable NLS_LANG is set. Solution is to set this variable to empty space in the controller environment.

On 64 bit Linux, user may get a warning message of insufficient access rights when connection test is performed, even if all the required access rights are provided. The connection can still be used to schedule the profile. Please make sure all the required access rights are provided to the user.

In custom checkpoints, if query tries to fetch data from a table with more than 32 columns, probe will limit the number of columns to 32.

Oracle Requirements

12 oracle Guide

Installation notes

Before installing the Oracle monitoring probe, you must ensure that you add Oracle libraries in the system PATH environment variable and configure the Oracle client appropriately.

To deploy the Oracle monitoring probe:

1. Obtain the latest version of the oracle.zip package from Internet Updates.

2. Install oracle.zip on the computer that has the Oracle client (drag from the Nimbus Archive).

3. Double-click the oracle probe.

The probe property sheet appears,

4. Configure the probe according to your requirements.

5. Update the following environment variables by using the controller GUI or by editing the nimbus script, which is located at /etc/init.d. You must restart the robot after updating the nimbus script.

On Windows platforms:

1. Set the ORACLE_HOME path to the directory where the client is installed. For example, C:\oracle\product\10.2.0\client_1.

On UNIX or Linux platforms:

1. Set the ORACLE_HOME path to the directory where the client is installed. For example, /u01/app/oracle/product/10.2.0/client_1.

2. Set the ORACLE_SID path to the service identifier name that you configured by using the Oracle client.

3. Set the LD_LIBRARY_PATH path to the Oracle library directory. For example, /u01/app/oracle/product/10.2.0/client_1/lib.

Note: On Linux platforms, you must ensure that the Current Enforcing Mode is Permissive.

To set the Current Enforcing Mode to Permissive:

1. Open SELinux Management.

The SELinux Administration window appears.

2. In the Current Enforcing Mode list, click Permissive.

3. Close the SELinux Administration window.

Oracle Configuration


Upgrade from Oracle Probe v3.9x to v4.01 or Later

1. Custom alarm messages for following checkpoints must be reconfigured.

tablespace_free : Custom alarm messages needs to be modified for b,kb,mb,gb

(free.n replaced with freesp.b)

tablespace_alloc_free : Custom alarm messages needs to be modified for

b,kb,mb,gb (free.n replaced with freesp.b)

database_size : Custom alarm messages needs to be modified for b,kb,mb,gb

(data.n replaced with datasize.b)

memory_usage : Custom alarm messages needs to be modified for b,kb,mb,gb

(mem_used.n replaced with value.b)

sga_memory : Custom alarm messages needs to be modified for b,kb,mb,gb

(memory.n replaced with bytes.b)

sga_memory_free : Custom alarm messages needs to be modified for b,kb,mb,gb

(memory.n replaced with bytes.b)

tablespace_deficit : Custom alarm messages needs to be modified for b,kb,mb,gb

(free.n replaced with free_bytes.b)

mts_uga_memory : Custom alarm messages needs to be modified for b,kb,mb,gb

(memory.n replaced with memory.b)

2. Not all variables from V3 version are supported in Oracle V4.xx version. The reconfiguration of following checkpoints is required, if new QoS or alarms are added or modified either in the template section or in profiles.

buf_cachehit_ratio : $hit_tot, $logical_reads

buf_cachehit_ratio : $hit_tot, $logical_reads

lib_cachehit_ratio : $hit_tot

rollback_segments : $waits_tot, $hitratio_tot

sort_ratio : $sortratio_total

redo_logs : $misses_ratio_tot

buf_cachehit_ratio_users : $hit_tot

dbfile_io : $pctIn_tot, $pctOut_tot

gc_service_util : $cache_tot

gc_av_lock_get_time : $av_time_tot

gc_fusion_wrt_rto : $rto_tot

lock_waits : $tot_t,$count,$time


Initial configuration of the oracle probe is done by using the GUI, defining connections and profiles for monitored instances. The probe may be running locally on the database server, or it may be configured to run as a remote client. Every profile will run in its own thread.

Note that all windows containing lists are resizable.

Also note that you click the Apply button to activate any configuration modifications done.


14 oracle Guide

The Setup tab

This property-sheet will set the general run-time parameters regarding the oracle probe.

Field Description

Setup tab

Generate status only Instructs the probe to only generate status, not to issue an alarm when a threshold is breached.

Select the Status tab to see the status for the different checkpoints.

Clear alarms on restart Allows the user to clear alarms on restart. By default, this checkbox is selected.



Field Description

Alarm severity filter Sets a "filter" on which severity levels to be considered as alarms.

The oracle probe is capable of checking many areas of the Oracle databases. Some events that are generated are vital and key to the performance and availability of the database. As a database administrator, you may want to pass the important events on to the operations centre or helpdesk, so the event can trigger pagers, email etc. The Alarm severity filter will consider the events matching the selected severity level and higher as alarms, and pass these on whenever the Generate status only option is unchecked.

If you e.g. set this to be major, then only messages with severity major and upward are considered as alarms.

Status Auto-Update A checkbox lets you activate/deactivate the Status Auto-Update functionality described below:

The "Status Auto-Update" parameter (number of seconds) specifies the automatic refresh interval of the Status Window on the Status tab. Setting this parameter to a value higher than 0 and then selecting a profile on the Status tab, the status will be automatically updated every x seconds. The checkpoints of the selected profile will be displayed until selecting another profile.

Note that this parameter is a "dialog" value - it means it is not saved in the configuration file, but in the machine running the dialog (same as for example widows size).

Log Size Set the size of the probe’s log file to which probe-internal log messages are written. The default size is 100 KB.

When this size is reached, the contents of the file are cleared.

Log level Sets the level of details written to the log file. Log as little as possible during normal operation, to minimize disk consumption.

Message pool tab This tab contains a list of all alarm messages available. You select messages from this list when editing the properties for a checkpoint.

Right-clicking in the list, allows you to add, edit, copy or delete messages.


16 oracle Guide

Defining a message, using the Message pool

Select the Message pool tab, right-click in the list and select New.

You must specify a name in the small dialog popping up.

Tips: Use the name of the checkpoint for which you create the alarm message as name. That makes it easier to find the alarm message when selecting an alarm message in the properties dialog for the checkpoint.



Select the checkpoint for which you create the alarm message in the drop-down list, and all variables available for that check-point will be listed in the right part of the dialog. Type the message and pick the variables you need. Click the OK button when finished. The new message should now appear in the message pool.

The Connections tab

This list contains the various connections to instances that the oracle probe will monitor. You need to specify user name, password and service name you want to use to connect to the instance. The password information is encrypted and placed into the configuration file. A connection can be used by more than one profile.


18 oracle Guide

The list contains one predefined connection that you may modify to your preferences. You may add, edit delete and copy connections.

Selecting the connection and choose Edit from the right popup-menu, opens the connection property window for editing.

Field Description

Description Short description of the connection.

User User-Id with authorization to read v$… views.

Password Valid password.

Service name Service name, as defined in the tsnames.ora file.

Retry attempts Number of attempts the probe should try to repeat connection in case of failure. "0" means only the initial connection will be done.

Retry delay The time the probe will wait between two connection attempts.

Test button Clicking this button will test if the connection can be made. If success, it will return the instance name and its version number. If not, an Oracle error message will be returned.



The Profiles tab

The list contains a sample profile that you may modify to your preferences. Every profile will run as a separate thread, and multiple profiles can be used to monitor one instance. This way the probe can be configured to deploy available resources the best way and allows independent monitoring of several instances simultaneously.

Icons in the profile list

■ Green icon in the profile line means the profile is active and running.

■ Yellow icon means the profile is active but suspended (the ‘suspend’ /’resume’ button in the profile properties dialog allows stopping / starting profile monitoring dynamic, without deactivating /activating the probe).

■ Black icon shows the profile is inactive.


20 oracle Guide

You may add, edit delete and copy profiles.

Selecting the profile and choose Edit from the right popup-menu, opens the profile property window for editing.

The ‘suspend’ /’resume’ commands allows stopping / starting profile monitoring dynamic, without deactivating /activating the probe.

The upper part of the window shows general profile properties and defaults. At the bottom, you will find a list of available checkpoints.

Field Description

Description Short description of the profile.

Heartbeat Defines the interval, at which all profile checkpoints schedules will be tested and trigger eventual checkpoint execution. This number should be common denominator to all used check interval values. The higher the value the lower is the profile overhead.



Field Description

Connection Connection used in this profile. It has to be defined in "Connections" dialog before crating a profile.

Check interval Default value for check interval in the profile. Will be used if nothing else is defined in the checkpoint and overwrites the default checkpoint list setting.

Clear message Message name for clear alarm.

SQL timeout Every checkpoint query run asynchronously. In case the query reaches the SQL timeout, the checkpoint processing will be terminated and the next checkpoint will be started. Alarm is issued.

Message Message name used for SQL timeout alarm

Profile timeout Defines the maximum processing time for all checkpoints in the profile. If this timeout is reached, the interval processing is finished and the probe waits for next heartbeat to evaluate any checkpoint schedules. Alarm message is issued.

Message Message name used for profile timeout alarm.

Timeout severity Severity for timeout messages.

Group Allows user to select the defined groups from the Groups tab.

Suspended/Resumed (indicator)

This indicator is green when the profile is activated. The indicator changes to yellow when the profile is suspended and to black when deactivated.

Alarm Source This option lets you override the source name of the alarm

Profile checkpoints At the bottom, you will find a list of available checkpoints. When defining a new profile, all checkpoints available (listed under the Checkpoints tab) will be listed here. Select the checkpoints you want for your new profile. The global and default checkpoint settings will be used, unless you modify the settings locally for your profile (see Note below this table).

Alarm Source Possibility to change the source for issued alarms. If not used, default is assumed (robot IP).


22 oracle Guide

Note on checkpoint types

Defining a profile, you can use two different strategies how to handle Checkpoints in a profile. You can decide to use checkpoint templates dynamic, which means that the checkpoints are defined globally (under the "Templates" tab) and represent the default settings. Every time you change the template value, it will reflect on all profiles using dynamic templates strategy.

If you want to have specific settings valid just for one profile, you right-click the checkpoint in the list and select Change to static.



Now you can double-click the checkpoint to modify the properties, and the settings will be valid for this profile only.

Note: If attempting to modify a template checkpoint in the Profile dialog without changing it to static as described above, you will get a warning like this:

Of course, there can be both "template" and "static" checkpoints mixed in one profile.

If a checkpoint is managed as static, the checkpoint name will appear in the list with a blue color, and it will be marked as static in the column Type.

Conclusion:

■ Static.

To manage the properties for a checkpoint locally, "change" the checkpoint to static in your profile before modifying it. When modified, the new settings will be valid for this profile only.

■ Template.

To edit the properties for a checkpoint template, double-click the checkpoint in the profile list or Templates tab. When modified, the new settings will be valid for all profiles, unless overruled by static settings in the profile.

See the section Editing Checkpoints for a description of the checkpoint properties.

When deciding which checkpoints to activate/deactivate for a profile, see the section Checkpoints Description for a description of the different checkpoints.


24 oracle Guide

The Templates tab

The list contains the predefined set of checkpoints that you may use in your profiles. These checkpoints can be modified to your preferences.

By default, most checkpoints are active with a reasonable default threshold value. The checkpoint properties may be used in a profile either dynamic, using the template values, or they can be added to the profile and managed static in the profile.

■ Static.

To edit the properties for a checkpoint locally for a profile, right-click the profile in the Checkpoints list in the profile dialog and change it to static. Then double-click the checkpoint to modify it. When modified, the new settings will be valid for this profile only.

■ Template.


See the section Editing Checkpoints for a description of the checkpoint properties.



The Status tab

The status is presented in a hierarchal fashion, with a profile name nodes and one or more checkpoint nodes (only active checkpoints are considered here). The highest status is propagated. Select the checkpoint in the navigation tree (to your left) to bring up the corresponding events.

Changing the individual values for checkpoints:

The properties for an individual checkpoint object can also be modified here. Select a profile and a monitored checkpoint in the left pane. Then double-click an object in the right pane.

If the object belongs to a template object, you will be warned that a modification will make the checkpoint static for the selected profile.

See the section Editing a checkpoint for a description of the checkpoint properties.


26 oracle Guide

The Groups tab

This tab allows user to create multiple groups which can be associated with profiles.

To add a group, right click in the Group tab and select New option.

Add New Group dialog appears.



Enter a name for the group and click OK.

New Group dialog appears. Enter a description for the new group in the Description text box. From the Group Checkpoints section, select the checkboxes for the check points that you wish to enable for the new group.

Click the OK button.


28 oracle Guide

The new group can now be selected from the Group drop-downj in the Edit Profiles dialog.

Creating a custom checkpoint



Perform the following steps to create a custom checkpoint for the oracle probe.

In the oracle probe GUI, go to the Templates tab. In the grid view, right-click and select Create new from the context menu.

Add New User Template dialog appears. Enter a name for the custom checkpoint. Click OK.


30 oracle Guide

Edit template checkpoint dialog opens. Go to the Query tab.

The Connection drop-down list displays a list of pre-configured connections, created from the Connections tab. Select the connection you wish to use.

To edit the connection parameters, click the Edit/New button.

Edit Connection dialog opens. If required, modify the details as described below:

Field Description

Description This field is non-editable



User ID and Password

Enter the user ID and password for the Oracle server

Service name Specify the name of the TNS listener with which the connection is to be established

Retry attempts Specify the number of retry attempts to be made for the connection

Retry delay Specify Time delay between 2 rety attempts

Click the Test button to validate the connection settings. A confirmation message appears.

Click OK.

In the Query tab of Edit template checkpoint dialog, enter the required query statement in the Query field. Click the Test button to validate the query.

Query Result dialog appears, displaying the result of the query entered.


32 oracle Guide

Click Close.

In the Query tab of Edit template checkpoint dialog, enter the details as described below:

Field Description

Checked value Name of the columns for whioch reporting should be done

Condition Select one of the comparison operators (=, <,>, etc.)

Row identification Row number for which reporting

Message variables Click Edit button to choose one or more message variables

Query File Name of the file to store the query enterred in Query field



Click OK.

In the General tab of Edit template checkpoint dialog, enter the details as described below:

Field Description

Description Short description of the purpose of the checkpoint.

Active Check this option to activate the checkpoint.

Condition Information, describing how the threshold values are evaluated.

Check interval Interval value used for this checkpoint. Every checkpoint can have a different check interval value. Default is taken from the profile definition, if not defined than from the default checkpoint list.

Send Quality of Service Activates QoS values being send into the QoS database. If not available in a checkpoint, checkbox is disabled.


34 oracle Guide

Field Description

QoS List Clicking this button opens the QoS list, showing the current QoS definitions (default is one definition per checkpoint).

Right-clicking in the list lets you add new QoS definitions and copy, edit or delete an existing QoS definition.

definition

The "Edit QoS" dialog offers available metrics (numerical variables which could be reported as QoS) and available object variables (if any - to be added to the QoS source).

The name of the QoS has to start with the checkpoint name.

QoS can be activated/deactivated as usual.

Note that some of the checkpoints have no QoS possibilities - from these checkpoints the QoS dialog cannot be activated.

Samples The probe will save the number of samples specified here and calculate an average value. This average value will be compared to the alarm threshold specified (see threshold description below the table).

Setting "Samples = 1", no sampling is done.

Setting "Samples = 3", the average of the 3 last samples will be used.

Setting "Samples = 0" (in profile), number of samples will be taken from the template. If not set there, no sampling is done.

Initially after start-up, the probe calculates the average value from the number of samples available.

Example, Samples=3:

In the first interval the first sample value is used

In the second interval, the average of sample 1 and 2 will be used etc.

NOTE: Many checkpoints calculate an "interval value", therefore in the first interval there is no value at all (no threshold checking).



Field Description

Use excludes Checking this option gives you the possibility to add excludes to the "exclude list" to some of the checkpoints (as it does not make sense for all checkpoints).

Using excludes, you can define objects that you do NOT want to monitor on the checkpoint. The excludes patterns found if clicking the Excludes list button (see below) will be used for the checkpoint.

Excludes list Clicking this button opens the Excludes list. This list shows if excludes are defined for the checkpoint.

The excludes found in the list will be used for the checkpoint if the Use excludes option (see above) is checked.

Right-clicking in the list lets you add new excludes or edit, copy or delete existing excludes.

When adding (or editing) an exclude pattern, a "match expression" dialog is opened, letting you edit or define the exclude pattern.

Excludes are defined using regular expression patterns.

A test button lets you test the exclude pattern defined.

This test is possible only for running active profiles and checkpoints.

The test uses the status list (on the status tab) as input:

Note that if there already are active excludes, the excluded objects are excluded from the status list BEFORE the test.

When clicking the test button, an exclude test list pops up, showing the result of the test:

Red text lines show the objects which would be excluded using the tested pattern.

The "object thresholds" are functioning as an "include list" - it means, if there are special thresholds defined for a special object, this object will always stay in, even if the exclude pattern would eliminate it normally. This is considered also in the test function.


36 oracle Guide

Field Description

Scheduling This field lets you select how to use the schedules settings, if any (see description below the table).

Rules

Selecting Rules means to run according the rules described in the Schedules settings.

Exceptions

Selecting Exceptions means to run except the rules described in the Schedules settings.

Clear message Message name used for clear alarm message.

Clear severity Severity, used for message issued in normal state.

Thresholds/Schedules See description below the table.

Click the QoS List button . QoS list dialog appears. Right-click in the grid view and select New from the context menu.



Edit dialog appears, with the name of the checkpoint (Custom-checkpoint in this example). Enter the necessary details such as name, description, unit, metric, abbreviation and max value.

Click OK in the Edit dialog and then in the QoS list dialog.


38 oracle Guide

Thresholds

The list contains the predefined set of monitoring profiles that you may use in your profiles and that you can modify to your preferences. By default, most profiles are active with a reasonable default threshold value. The threshold values may be defined by modifying checkpoints in the respective profile. Every checkpoint has to have at least one threshold, but there can be additional thresholds defined.

The threshold identification consists of an object name (if applicable), like tablespace name, userid etc. and a threshold ID, numbered from 0. Threshold values have to be descending or ascending, depending on condition used in a checkpoint, starting with the highest severity threshold condition.

Field Description

Threshold object name Monitoring object name, if applicable or default. Some special checkpoints have a second threshold called ‘count" (e.g. "locked_users").

Threshold value Value used for threshold evaluation

Current value If invoked from the status report, it contains the last measured value.

Severity Alarm severity.

Message Name of message used for threshold alarm.

Message text Text of the message, containing variables, which will be replaced in run time. If the message text is changed from a profile list, you will be forced to create new message.



Field Description

Variables List of variables, available in the checkpoint.

Schedules

You can specify a schedule for the checkpoint to run.

Right-click in the Schedules section of General tab, and select New from the context menu.


40 oracle Guide

If the schedules list is empty, the checkpoint will be executed in interval matter, 24 hours a day. Additional to it there can be defined number of schedules per checkpoints, which can define additional ‘rules’ to the check interval or ‘exceptions’ of it. ‘Rules’ and ‘exceptions’ cannot be mixed in one

checkpoint.

In principle, a schedule is a definition of an execution period (or execution break if ‘exceptions’ used) with specified days, time from/to and date from/to values. Additionally to it, if only ‘date from’ and ‘time from’ is defined, first execution can be defined. ‘Run once’ will cause the checkpoint run only once a day in the defined period (unlike multiple times if ‘interval’ used).

Click OK. A confirmation message appears asking you to confirm the modification of the template. Click Yes button to continue. In addition, you can also select the Do not ask this question again checkbox to prevent this message from appearing the next time you modify a template.



On clicking the Yes button, the new checkpoint is added to the Templates tab of the oracle probe GUI, as shown below (Custom-checkpoint in this example).

Editing a checkpoint

42 oracle Guide


The checkpoint properties may be used in a profile either dynamic, using the template values, or they can be added to the profile and managed static in the profile.

■ Static

To edit the properties for a checkpoint locally for a profile, right-click the profile in the Checkpoints list in the profile dialog and change it to static. Then double-click the checkpoint to modify it. When modified, the new settings will be valid for this profile only.

■ Template


The properties for checkpoints are described below:

The upper part of the window contains general checkpoint settings. The lower part contains two lists with threshold and schedule settings.

Field Description

Description Short description of the purpose of the checkpoint.

Active Check this option to activate the checkpoint.

Condition Information, describing how the threshold values are evaluated.



Field Description

Check interval Interval value used for this checkpoint. Every checkpoint can have a different check interval value. Default is taken from the profile definition, if not defined than from the default checkpoint list.

Send Quality of Service Activates QoS values being send into the QoS database. If not available in a checkpoint, checkbox is disabled.

QoS List Clicking this button opens the QoS list, showing the current QoS definitions (default is one definition per checkpoint).

Right-clicking in the list lets you add new QoS definitions and copy, edit or delete an existing QoS definition.

definition

The "Edit QoS" dialog offers available metrics (numerical variables which could be reported as QoS) and available object variables (if any - to be added to the QoS source).

The name of the QoS has to start with the checkpoint name.

QoS can be activated/deactivated as usual.

Note that some of the checkpoints have no QoS possibilities - from these checkpoints the QoS dialog cannot be activated.


44 oracle Guide

Field Description

Samples The probe will save the number of samples specified here and calculate an average value. This average value will be compared to the alarm threshold specified (see threshold description below the table).

Setting "Samples = 1", no sampling is done.

Setting "Samples = 3", the average of the 3 last samples will be used.

Setting "Samples = 0" (in profile), number of samples will be taken from the template. If not set there, no sampling is done.

Initially after start-up, the probe calculates the average value from the number of samples available.

Example, Samples=3:

In the first interval the first sample value is used

In the second interval, the average of sample 1 and 2 will be used etc.

NOTE: Many checkpoints calculate an "interval value", therefore in the first interval there is no value at all (no threshold checking).

Use excludes Checking this option gives you the possibility to add excludes to the "exclude list" to some of the checkpoints (as it does not make sense for all checkpoints).

Using excludes, you can define objects that you do NOT want to monitor on the checkpoint. The excludes patterns found if clicking the Excludes list button (see below) will be used for the checkpoint.



Field Description

Excludes list Clicking this button opens the Excludes list. This list shows if excludes are defined for the checkpoint.

The excludes found in the list will be used for the checkpoint if the Use excludes option (see above) is checked.

Right-clicking in the list lets you add new excludes or edit, copy or delete existing excludes.

When adding (or editing) an exclude pattern, a "match expression" dialog is opened, letting you edit or define the exclude pattern. Excludes are defined using regular expression patterns.

A test button lets you test the exclude pattern defined.

This test is possible only for running active profiles and checkpoints.

The test uses the status list (on the status tab) as input:

Note that if there already are active excludes, the excluded objects are excluded from the status list BEFORE the test.

When clicking the test button, an exclude test list pops up, showing the result of the test:

Red text lines show the objects which would be excluded using the tested pattern.

The "object thresholds" are functioning as an "include list" - it means, if there are special thresholds defined for a special object, this object will always stay in, even if the exclude pattern would eliminate it normally. This is considered also in the test function.

Scheduling This field lets you select how to use the schedules settings, if any (see description below the table).

Rules

Selecting Rules means to run according the rules described in the Schedules settings.

Exceptions

Selecting Exceptions means to run except the rules described in the Schedules settings.

Clear message Message name used for clear alarm message.

Clear severity Severity, used for message issued in normal state.

Thresholds/Schedules See description below the table.


46 oracle Guide

Thresholds

The list contains the predefined set of monitoring profiles that you may use in your profiles and that you can modify to your preferences. By default, most profiles are active with a reasonable default threshold value. The threshold values may be defined by modifying checkpoints in the respective profile. Every checkpoint has to have at least one threshold, but there can be additional thresholds defined.

The threshold identification consists of an object name (if applicable), like tablespace name, userid etc. and a threshold ID, numbered from 0. Threshold values have to be descending or ascending, depending on condition used in a checkpoint, starting with the highest severity threshold condition.

Field Description

Threshold object name Monitoring object name, if applicable or default. Some special checkpoints have a second threshold called ‘count" (e.g. "locked_users").

Threshold value Value used for threshold evaluation

Current value If invoked from the status report, it contains the last measured value.

Severity Alarm severity.

Message Name of message used for threshold alarm.



Message text Text of the message, containing variables, which will be replaced in run time. If the message text is changed from a profile list, you will be forced to create new message.

Variables List of variables, available in the checkpoint.

Schedules

If the schedules list is empty, the checkpoint will be executed in interval matter, 24 hours a day. Additional to it there can be defined number of schedules per checkpoints, which can define additional ‘rules’ to the check interval or ‘exceptions’ of it. ‘Rules’ and ‘exceptions’ cannot be mixed in one checkpoint.

In principle, a schedule is a definition of an execution period (or execution break if ‘exceptions’ used) with specified days, time from/to and date from/to values. Additionally to it, if only ‘date from’ and ‘time from’ is defined, first execution can be defined. ‘Run once’ will cause the checkpoint run only once a day in the defined period (unlike multiple times if ‘interval’ used).

Checkpoints metrics

48 oracle Guide

Checkpoints metrics

There are five kind of metrics used:

Count

Absolute number of events in the interval. It is calculated as delta between count at the beginning of the interval and at the end. In the first interval, counts are not checked because their interval value cannot be calculated. If there is a "total" value in the message, it means "since the start of the instance".

Gauge

Absolute number, describing the actual state of the system.

Ratio

Calculated percentage, using interval counts. In the first interval, it is calculated from total counts (as the interval count cannot be calculated).

Average

Calculated using interval counts. In the starting interval, it is calculated from absolute counts.

Status

Absolute value like ‘ONLINE’ etc.

oracle Metrics

The following table describes the checkpoint metrics that can be configured using the oracle probe.

Monitor Name Units Description

Active_users Checks for currently active Oracle users

dataguard_gap Monitors Dataguard Instance Gap

dataguard_status Monitors Dataguard Instance Status

dataguard_timegap Monitors Dataguard Instance Timegap

flash_recovery_area_memory_free Monitors flash recovery area memory free

resource_util Monitors database utilization for resources (Processes, Sessions and Transactions)

Checkpoints metrics


tablespace_temp_free Monitors Temporary tablespaces for space left

active_connection_ratio ratio Calculated as: (active/total) * 100.

Monitors % of active connections to total allowed connections.

Buf_cachehit_ratio ratio Calculated as:

(logical-reads - physical-reads)*100 / physical-reads

Tells, how often data has been found in cache compared to how often it has to be read from the disk. It can be low at the start of a new application, before the cache is filled with data. If it stays low for many intervals you should consider increasing the size of the cache (DB_BLOCK_BUFFERS).

Buf_cachehit_ratio_users ratio Same as buf_cachehit_ratio, only per user. It helps to find a poor query or application.

Chained_rows gauge Measures number of chained rows in a table. As chaining causes performance degradation, reported tables should be reorganized as soon as possible.

Check_dbalive status This checkpoint tries to connect to an instance. In case the connection cannot be established, an alert is generated. This checkpoint cannot be deactivated!

In case of Alert check the Oracle connectivity, the instance or the server where the instance is running.

Database_size ratio Monitors free space in database and log files in %. Source: Query.

Datafile_status status Reports all files being in other status as 'ONLINE'.

Dbfile_io ratio Calculated as:

(Physical-reads / logical-reads) * 100

(Physical-writes / logical-writes) * 100

The higher of the above values is taken for checking. It helps to identify database files with higher I/O contention.

Segments with high I/O activity should be divided into more files and eventually spread over different devices.

Checkpoints metrics

50 oracle Guide

Dict_cachehit_ratio ratio Calculated as:

(hits - misses) * 100 / hits

Measures number of times where Oracle metadata are found in memory. It could be low at the start of a new application, before new data is placed into the cache.

In case the ratio is low over more intervals, you should consider increasing the cache size (SHARED_POOL_SIZE).

Enqueue_timeouts count Measures number of table and row locks (acquired and converted) that timed-out before they could complete.

Timeouts can cause serious performance degradation and should never occur. the number of enqueue resources should be increased. (ENQUEUE_RESOURCES or

Extents gauge Reports tables that have more extents as threshold.

Fragmentation could cause performance degradation; therefore the reported tables should be reorganized as soon as possible.

Note: Running this checkpoint is relatively resource intensive; therefore this checkpoint should not run in every interval (use interval multiplier.

Gc_av_lock_get_time average (Oracle 9i and above)

Measures average interval global lock get time.

High values indicate either contention on the connection channel between the involved instances or on the instance SGA (SHARED_POOL_SIZE).

Note: This checkpoint applies only to Real Application Cluster (RAC) configuration and is therefore per default inactive.

gc_blocks_corrupt count (Oracle 9i and above)

Measures number of blocks corrupted during transfer.

Corruption should never occur. Indicates networking problems.

gc_blocks_lost count (Oracle 9i and above)

Measures number of blocks lost during transfer.

It should never occur. Indicates networking problems.

Checkpoints metrics


gc_convert_timeouts count (Oracle 9i and above)

Measures number of times lock converts in global cache timed out.

Occurrence means contention on global cache – check SGA size or network connection.

gc_cr_timeouts count (Oracle 9i and above)

Measures number of times process request for a consistent read block timed out.

Occurrence means contention on global cache – check SGA size or network connection.

gc_fusion_wrt_rto ratio (Oracle 9i and above)

Calculated as:

Fusion writes / physical writes

Measures percentage of blocks written to disk that have been "global dirty".

High number can result from insufficiently sized caches (SGA) or low checkpoint frequency.

gc_service_util ratio (Oracle 9i and above)

Calculated as:

Global cache service requests / logical reads

Measures global cache utilization.

High number can result from insufficient cache size (SGA). Also, it can indicate inappropriate spread of data and applications in the RAC configuration.

index_status status Lists all invalid indexes.

Invalid index needs to be recompiled.

invalid_objects status Lists all invalid objects.

Invalid object needs to be rebuilt.

lib_cachehit_ratio ratio Calculated as:

(hits – misses)*100 / hits

Measures percentage of times object metadata have been found in memory.

This number can be lower at the start of new application. If it is low for several intervals, the SHARED_POOL_SIZE should be increased.

Checkpoints metrics

52 oracle Guide

locked_users count Measures number of locked users in an instance and displays their list.

Inappropriate locking can cause serious performance degradation. Check also the ‘user_locks’ checkpoint for more information about locking situation in the system.

This checkpoint has two threshold values – ‘default’ and ‘count’. If the ‘default’ threshold is set = 1, than the probe will issue for every locked user a message with details about the user locking situation. If set > 1, no message this types will be issued.

The ‘count’ threshold defines the number of locked users as a threshold for this checkpoint.

lock_waits count Measures number of time, user session was waiting for a lock in last interval.

Inappropriate locking can cause serious performance degradation. Check also the ‘user_locks’ and ‘locked_users’ checkpoints for more information about locking situation in the system.

login_count count Measures number of currently active user session except for user "SYS" and "SYSTEM".

long_queries count The checkpoint generates two kinds of alarms:

Alarm if number of long queries is higher than threshold (called "count") - this value is also used as QoS.

One alarm for every query running longer than set threshold. The threshold key is: "osuser.sid.operation" - this key can be also used for exclude list. (osuser stands for operating system user-id, compared to oracle user id).

memory_usage count Measures total amount of memory (in bytes) user sessions consume (PGA).

If the system has enough physical memory, increase parameter PGA_AGGREGATE_TARGET or change to automatic PGA management (9i). Otherwise look into V$PROCESS, V$PGASTAT, V$SQL_WORKAREA views to find out who is spending most of the memory.

Checkpoints metrics


mts_response count Monitors the average MTS server response time in ms.

If the response time is too high, consider increasing number of dispatchers (MTS_MAX_DISPATCHERS and MTS_DISPATCHERS).

mts_uga_memory gauge Monitors the MTS user memory consumption.

If the memory consumption is too high, consider limiting it by using parameter PRIVATE_SGA.

mts_wait count Monitors the MTS server wait time in ms.

If the wait time is too high, consider increasing number of dispatchers (MTS_MAX_DISPATCHERS and MTS_DISPATCHERS).

no_next_extents count Monitors objects, which are not able to allocate next extent. It considers only objects with AUTOEXTEND OFF.

Object size has to be increased.

redo_logs ratio Calculated as:

Misses * 100 / hits

Measures redo log contention.

If the ratio is too high, consider increasing the redo log buffer

(parameter LOG_BUFFER). Also, check LOG_SMALL_ENTRY_MAX_SIZE and LOG_SIMULTANEOUS_COPIES parameters.

remaining_extents count Measures remaining number of extents for object.

Increase number of max. extents.

rollback_seg_status status Lists rollback segments which are not ‘ONLINE’.

Make sure rollback segments are ‘ONLINE’ (e.g. with ALTER ROLLBACK SEGMENT).

rollback_segments count Measures number of waits for rollback segment.

Number of waits should be close to 0. Otherwise increase number of extents or add new rollback segment. Also check the parameter OPTIMAL and MAXEXTENTS.

Checkpoints metrics

54 oracle Guide

session_waits count Measures number of waits in interval.

Number of waits should be close to 0. There are different reasons for waits – locking, insufficient number of rollback segments, freelist contention etc. Check the view V$SESSION_WAIT or V$SESSION_EVENT for details.

sga_memory count Measures SGA memory usage in bytes by pool.

Check the SHARED_POOL_SIZE parameter.

sga_memory_free count Measures SGA free memory usage in bytes.

Check the SHARED_POOL_SIZE parameter.

sort_ratio ratio Calculated as:

Memory sorts * 100 / total sorts

Too many disk sorts can cause performance degradation. Check parameter SORT_AREA_SIZE and SORT_AREA_RETAINED_SIZE.

system_statistics information Displays available system statistics counters. These counters contain number of events since the instance start.

Use the statistics to analyze diverse system problems or setup meaningful checkpoint thresholds.

system_waits count Measures number of system waits by event.

These counters can be used to identify system bottlenecks. Use the table V$SYSTEM_EVENT to find the most "expensive" event.

tablespace_alloc_free ratio Measures percentage of free space in already allocated files (not considering the AUTOEXTEND parameter).

If the reported tablespace has AUTOEXTEND OFF you should allocate new space. Otherwise it solemnly indicates that Oracle will soon allocate new space.

Note: This checkpoint is set inactive per default.

tablespace_deficit count Lists tablespaces with AUTOEXTEND OFF, which cannot allocate new extent.

Increase tablespace size for listed tablespaces.

String matching methods


tablespace_free ratio Measures percentage of free space in a tablespace, considering the maximal possible tablespace size.

Increase tablespace size or set AUTOEXTEND ON for listed tablespaces.

Notice: Oracle Enterprise Manager Console doesn’t consider maximum possible tablespace size on most of its screens (status 9i)!

tablespace_status status Lists tablespaces that have status different as ‘ONLINE’ or ‘SYSTEM’.

Set the listed tablespace online.

tablespace_size count Measures the total tablespace size.

user_locks count Lists user holding a lock.

Helps to identify user, blocking other sessions (checkpoint locked_users).

This checkpoint has two threshold values – ‘default’ and ‘count’. If the ‘default’ threshold is set = 1, than the probe will issue for every locked user a message with details about the user locking situation. If set > 1, no message this types will be issued.

The ‘count’ threshold defines the number of locked users as a threshold for this checkpoint.

String matching methods

The probe supports two methods of string-matching. The simple method is called pattern matching, and is very similar to the one found in many popular shells (UNIX), and the more powerful one, called regular expressions.

Glossary 57

Glossary

Nimsoft Monitor oracle Guidedocs.nimsoft.com/prodhelp/en_US/Probes/Catalog/oracle/4.4/oracle.pdf ·...

Documents

Transcript of Nimsoft Monitor oracle Guidedocs.nimsoft.com/prodhelp/en_US/Probes/Catalog/oracle/4.4/oracle.pdf ·...