Nimsoft Monitor. oracle Guide. v4.5 series

Transcription

1 Nimsoft Monitor oracle Guide v4.5 series

2 Legal Notices This online help system (the "System") is for your informational purposes only and is subject to change or withdrawal by CA at any time. This System may not be copied, transferred, reproduced, disclosed, modified or duplicated, in whole or in part, without the prior written consent of CA. This System is confidential and proprietary information of CA and protected by the copyright laws of the United States and international treaties. This System may not be disclosed by you or used for any purpose other than as may be permitted in a separate agreement between Notwithstanding the foregoing, if you are a licensed user of the CA Software you may make one copy of the System for internal use by you and your employees, provided that all CA copyright notices and legends are affixed to the reproduced copy. The right to make a copy of the System is limited to the period during which the license for the CA Software remains in full force and effect. Should the license terminate for any reason, it shall be your responsibility to certify in writing to CA that all copies and partial copies of the System have been destroyed. WITHOUT WARRANTY OF ANY KIND, INCLUDING WITHOUT LIMITATION, ANY IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE OR NONINFRINGEMENT. IN NO EVENT WILL CA BE LIABLE TO YOU OR ANY THIRD PARTY FOR ANY LOSS OR DAMAGE, DIRECT OR INDIRECT, FROM THE USE OF THIS SYSTEM, INCLUDING WITHOUT LIMITATION, LOST PROFITS, LOST INVESTMENT, BUSINESS INTERRUPTION, GOODWILL, OR LOST DATA, EVEN IF CA IS EXPRESSLY ADVISED IN ADVANCE OF THE POSSIBILITY OF SUCH LOSS OR DAMAGE. The manufacturer of this System is CA. Government is subject to the restrictions set forth in FAR Sections , , and (c)(1) - (2) and DFARS Section (b)(3), as applicable, or their successors. Copyright 2014 CA. All rights reserved. All trademarks, trade names, service marks, and logos referenced herein belong to their respective companies. Legal information on third-party and public domain software used in the Nimsoft Monitor solution is documented in Nimsoft Monitor Third-Party Licenses and Terms of Use (

3 Contact CA Contact CA Support For your convenience, CA Technologies provides one site where you can access the information that you need for your Home Office, Small Business, and Enterprise CA Technologies products. At you can access the following resources: Online and telephone contact information for technical assistance and customer services Information about user communities and forums Product and documentation downloads CA Support policies and guidelines Other helpful resources appropriate for your product Providing Feedback About Product Documentation Send comments or questions about CA Technologies Nimsoft product documentation to To provide feedback about general CA Technologies product documentation, complete our short customer survey which is available on the CA Support website at

4

5 Contents Chapter 1: oracle oracle Overview... 7 Installation notes Known issues Chapter 2: oracle Probe Deployment 13 Supported Platforms System Requirements Software Requirements Monitoring System Requirements Probe Deployment Information Supported Oracle Versions and Clients Specific Notes Chapter 3: Upgrades and Migrations 17 Upgrade from Oracle Probe v3.9x to v4.01 or Later Chapter 4: oracle Configuration 19 Probe Defaults Probe Configuration Interface Installation Probe Configuration The Setup tab The Connections Tab The Profiles Tab The Templates Tab The Status Tab The Groups Tab String matching methods Chapter 5: QoS Threshold Metrics 51 Checkpoint Metrics oracle QoS Metrics oracle Alert Metrics Default Settings Contents 5

6

7 Chapter 1: oracle 4.5 This description applies to oracle probe version 4.5. This section contains the following topics: oracle Overview (see page 7) Documentation Changes (see page 10) Installation notes (see page 10) Known issues (see page 11) oracle Overview The Oracle monitoring probe monitors local or remote Oracle database instances. This probe periodically scans through a configurable set of monitoring profile and run selected Oracle statements for extracting vital information about Oracle servers. The information is presented to the database administrator as alarms and report. Important! 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 Cache hit Ratio User Memory Consumption Number of rollback segments Objects that are unable to allocate next extent Invalid objects in database Chained rows in database Datafile Availability Concurrent Users Chapter 1: oracle 4.5 7

8 oracle Overview Enqueue Timeouts Redo Log contention MTS Response MTS Wait SGA Memory Usage Free SGA Memory Sort Ratio Index status Tablespace Availability 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 8 oracle Guide

9 oracle Overview Database Size Resource Utilization % Dataguard Status Dataguard Gap Dataguard Timegap Tablespace Temp Free Active Users Flash recovery area memory free Active Connection Ratio Chapter 1: oracle 4.5 9

10 Installation notes Documentation Changes This table describes the version history for this document. Version Date What's New? 4.5 February 2013 Added a checkbox - Skip Privilege Check on the New Connection dialog box of Connections tab. Added a field - Connection Failure Message on the New Profile dialog box of Profiles tab. The message selected in this field will be visible in the Message Pool list and you can edit it from there. Added Probe Defaults topic. 4.4 August 2011 Added support for oracle client 11.2 on Solaris Sparc. Related Documentation Documentation for other versions of the oracle probe The Release Notes for the oracle probe Monitor Metrics Reference Information for CA Nimsoft Probes ( Installation notes Add the Oracle libraries in the system PATH environment variable and must configure the Oracle client appropriately before installing the Oracle monitoring probe. Follow these steps for deploying 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 at /etc/init.d. Restart the robot after updating the nimbus script. 10 oracle Guide

11 Known issues 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. Important: The Current Enforcing Mode must be Permissive on Linux Platform. Follow these steps for setting 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. 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 ). Error ORA 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. Documentation Changes 11

12

13 Chapter 2: oracle Probe Deployment This section contains system requirements and deployment information for the oracle probe. This section contains the following topics: Supported Platforms (see page 13) System Requirements (see page 13) Software Requirements (see page 13) Monitoring System Requirements (see page 14) Probe Deployment Information (see page 14) Supported Oracle Versions and Clients (see page 14) Supported Platforms The oracle probe is supported on the same set of operating systems and databases as the Nimsoft Monitor Server solution. Please refer to the: Nimsoft Compatibility Support Matrix for the latest information on supported platforms. Support Matrix for Nimsoft Probes for additional information on the oracle probe. System Requirements The oracle probe should be installed on systems with the following minimum resources: Memory: 2-4 GB of RAM. Probe's OOB configuration requires 256 MB of RAM. CPU: 3 GHz dual-core processor, 32-bit or 64-bit. Software Requirements The oracle probe requires the following software environment: Nimsoft Monitor Server or later Nimsoft robot version 5.23 or later Java Virtual Machine version 1.6 or later (typically installed with NMS 5.0 and above) Note: For SOC functionality, NM Server 5.6 or later and UMP or later is required. Chapter 2: oracle Probe Deployment 13

14 Monitoring System Requirements Monitoring System Requirements The oracle probe monitors local or remote Oracle database instances, periodically scans through a configurable set of monitoring profiles, and apply checks to the local or remote Oracle instances. Probe Deployment Information There are three ways to distribute archive packages. You can distribute the package within the web-based Admin Console (for supported probes), from within Infrastructure Manager, or use the standalone Nimsoft Distribution application.see Probe Deployment for more information on deploying probes. Supported Oracle Versions and Clients The following matrix summarizes client and server combinations that are supported. The probe is built using libraries. Server Version Client Version # # #3 # Key Yes No Supported Not Supported 14 oracle Guide

15 Supported Oracle Versions and Clients Specific Notes Some notes of the New Probe probe are as follows: #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 the connections between 10.2/11.1/11.2 (or higher) and 9.2, the 9.2 end MUST be at or higher. Connections between 10.2 (or higher) and , or are not supported. Chapter 2: oracle Probe Deployment 15

16

17 Chapter 3: Upgrades and Migrations This section contains the following topics: Upgrade from Oracle Probe v3.9x to v4.01 or Later (see page 17) 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 Chapter 3: Upgrades and Migrations 17

18

19 Chapter 4: oracle Configuration This section describes the configuration concepts and procedures for setting up the oracle probe. This section contains the following topics: Probe Defaults (see page 19) Probe Configuration Interface Installation (see page 19) Probe Configuration (see page 19) String matching methods (see page 49) Probe Defaults At the time of deploying a probe for the first time on robot, some default configuration will be deployed. For this you have to drag and drop the test template to the profile. These probe defaults could be Alarms, QoS, Profiles and so on which save time to configure the default settings. These probe defaults will be seen on a fresh install, that is no instance of that probe is already available on that robot in activated or deactivated state. Probe Configuration Interface Installation The probe configuration interface is automatically downloaded and installed by the Nimsoft Infrastructure Manager when the probe is deployed on a robot. Probe Configuration 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. Notes: All windows containing lists are resizable. Click Apply to activate any configuration modifications done. Chapter 4: oracle Configuration 19

20 Probe Configuration The Setup tab The Setup tab contains two sub tabs - General (see page 20) and Message Pool (see page 22). By default, the General sub tab is selected. The General Tab This tab allows you to perform the required settings, for example clear alarms on restart of probe, size, log level, and so on. 20 oracle Guide

21 Probe Configuration This property-sheet will set the general run-time parameters regarding the oracle probe. The fields in the above dialog are explained below: Generate status only Instructs the probe to only generate status and not to issue an alarm when a threshold is breached. Select the Status (see page 46) tab to see the status for the different checkpoints. Clear alarms on restart Allows you to clear alarms on restart. By default, this check box is selected. 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, 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 select the option as major, then only the messages with severity level as major and upward are considered as alarms. Status Auto-Update Allows you to 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 be updated every x seconds. The checkpoints of the selected profile will be displayed until selecting another profile. Note: This parameter is a dialog value - means it is not saved in the configuration file, but in the machine running the dialog (same as for example windows size). Log Size 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. Chapter 4: oracle Configuration 21

22 Probe Configuration The Message Pool Tab The Message pool tab contains a list of all available alarm messages. You select messages from this list when editing the properties for a checkpoint. Right-clicking in the list allows you add, edit, copy, or delete messages. Follow these steps to add a new message: 1. Right in the message list and select the New option. The New message dialog is displayed. 2. Specify a message name in the Name field and click OK. Note: Use the checkpoint name for which you create the alarm message as Name. This helps to find the alarm message, when it is selected in the Properties dialog for the checkpoint. 22 oracle Guide

23 Probe Configuration The Edit message screen appears. 3. Select the checkpoint for which you create the alarm message in the Checkpoint drop-down list. All the variables available for that checkpoint will be listed in the Variables section. 4. Type the message in the Message text box, select the variable(s) you require, and click the button. The selected variables get populated in the Message text box. Note: You can also delete a wrongly-selected variable and then select the required one. 5. Click OK. The new message now appears in the message pool. Similarly, you can copy, modify, or delete a message. Chapter 4: oracle Configuration 23

24 Probe Configuration 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. The list contains one predefined connection that you may modify to your preferences. You may add, edit, delete, and copy connections. Follow these steps to add a new connection: 1. Right-click in the Connections window and select the New option. The Add New Connection dialog displays. 24 oracle Guide

25 Probe Configuration 2. Enter the connection name in the Name field and click OK. The New Connection dialog displays. The fields in the above dialog are explained below: Description Specifies the short description of the connection. User ID Specifies the user id with authorization to read v$...views. Password Specifies the valid password. Chapter 4: oracle Configuration 25

26 Probe Configuration Service name Specifies service name, as defined in the tnsnames.ora file. Retry attempts Specifies the 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 Specifies the time for which the probe will wait between two connection attempts. Skip Privilege Check Check this box for those users who do not have DBA and Select Catalog Role privileges so that custom checkpoints can be created. When you check this box, the two fields - DB Version and Instance become mandatory. DB Version Instance Defines the version of the database used for the connection. Specifies the instance name to which the connection is to be established. Note: It is assumed that you should provide valid instance information and database version. Selecting Skip Privilege Check check box, may affect the execution of normal checkpoints which require Select Catalog Role privileges. A scenario is stated here to explain the purpose of this check box. Scenario: Suppose User A does not have the Select Catalog Role privileges and want to execute some of the custom checkpoints in the oracle probe. For this, User A creates a connection with correct username and password. To execute those checkpoints, the Skip Privilege Check check box is selected and the correct DB Version and Instance information is provided. Now, the user can create custom checkpoint(s) and execute the checkpoint(s). 3. Enter the required fields as explained above and click OK. The new connection record gets added. Similarly, you can copy the details of an existing connection to a new connection, modify, or delete a connection. 26 oracle Guide

27 Probe Configuration 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 options in the profile properties dialog allows stopping / starting profile monitoring dynamic, without deactivating /activating the probe). Black icon shows the profile is inactive. Right-clicking in the Profiles dialog, you can add, edit, delete, copy, suspend, and resume profiles. Follow these steps to add a new profile: 1. Right-click in the Profiles dialog and select the New option. The Add New Profile dialog appears. Chapter 4: oracle Configuration 27

28 Probe Configuration 2. Enter the profile name and click OK. The New Profile dialog appears. The upper part of the window shows general profile properties and defaults. At the bottom, you will find a list of available checkpoints. The fields in the above dialog are explained after Step Enter / Select the required fields as explained below and click OK. The new profile gets added in the list. 28 oracle Guide

29 Probe Configuration Description Specifies the 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. Connection Defines the connection used in this profile. It has to be defined in Connections dialog before creating a profile. Check Interval Specifies the 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 Specifies the 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 Defines the 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 Defines the message name used for profile timeout alarm. Timeout severity Group Defines the severity for timeout messages. Allows you to select the defined groups from the Groups tab. Alarm source Possibility to change the source for issued alarms. If not used, default is assumed (robot IP). Chapter 4: oracle Configuration 29

30 Probe Configuration Notes: Connection Failure Message When there are some connection related issues, such as DB not connected, then the message selected in this field is displayed. Note: This message can also be edited in the Message pool 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. 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). While defining a profile, you can use two different strategies 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. 30 oracle Guide

31 Probe Configuration Now you can double-click the checkpoint to modify the properties, and the settings will be valid for this profile only. If attempting to modify a template checkpoint in the Profile dialog without changing it to static as described above, you will get a warning as shown below: Chapter 4: oracle Configuration 31

32 Probe Configuration 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. Therefore, the conclusion is: 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 Profiles list or Templates tab. When modified, the new settings will be valid for all profiles, unless overruled by static settings in the profile. The Templates Tab The list contains the predefined set of checkpoints that you may use in your profiles. These checkpoints can be modified as per 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 Profiles 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. 32 oracle Guide

33 Probe Configuration Template Creating a Custom Checkpoint 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. This section will familiarize you how to create a custom checkpoint for the oracle probe. Follow these steps: 1. In the oracle probe GUI, go to the Templates tab. In the grid view, right-click and select Create new from the context menu. The Add New User Template dialog appears. 2. Enter a name for the custom checkpoint and click OK. The Edit template checkpoint dialog opens. 3. By default, the General tab opens. 4. Click the Query tab. Chapter 4: oracle Configuration 33

34 Probe Configuration 5. Select the required connection from the Connection drop-down list, which displays a list of pre-configured connections, created from the Connections tab. Note: To edit the connection parameters, click Edit/New which displays the Edit Connection dialog (see Connections tab (see page 24)). 6. Click Test to validate the connection settings. A confirmation message appears. 34 oracle Guide

35 Probe Configuration 7. Click OK. 8. In the Query tab of Edit template checkpoint dialog, enter the required query statement in the Query field and click Test to validate the query. The Query Result dialog appears, displaying the result of the query entered. 9. Click Close. 10. Again in the Query tab of Edit template checkpoint dialog, enter the details as shown and described below: Chapter 4: oracle Configuration 35

36 Probe Configuration The fields in the above dialog are explained below: Checked value Name of the columns for which reporting should be done. Condition Select one of the comparison operators (=, <,>, etc.). Row identification Row number for which reporting is done. Message variables Click Edit to choose one or more message variables. Query File 11. Click OK. Name of the file to store the query entered in Query field. 12. Now, click the General tab of Edit template checkpoint dialog and enter the details as shown below: 36 oracle Guide

37 Probe Configuration The fields in the above dialog are explained below: Description Active Indicates the short description of the purpose of the checkpoint. Activates the checkpoint, if checked. Condition Identifies the 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 then from the default checkpoint list. Send Quality of Service Activates QoS values being send into the QoS database. If not available in a checkpoint, check box 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. 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: Some of the checkpoints have no QoS possibilities - from these checkpoints the QoS dialog cannot be activated. Samples Defines the number of samples which the probe will save to calculate the average value. This average value is compared to the specified alarm threshold value. (See Thresholds (see page 40)). 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 the start-up, the probe calculates the average value from the number of available samples. In the first interval the first sample value is used. In the second interval, the average of sample 1 and 2 will be used, and so on. Note: Many checkpoints calculate an "interval value", therefore in the first interval there is no value at all (no threshold checking). Use excludes Chapter 4: oracle Configuration 37

38 Probe Configuration If selected, 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, allowing you to edit or define the exclude pattern. Excludes are defined using regular expression patterns. A test button allows 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: If there are already 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 allows you to select how to use the schedules settings, if any (see Schedules (see page 42)). rules Selecting this option means to run according the rules described in the Schedules settings. exceptions Clear message Selecting this option means to run except the rules described in the Schedules settings. Specifies the message name used for clear alarm message. Clear severity Specifies the severity used for message issued in normal state. Thresholds/Schedules For descriptions, see Thresholds (see page 40) and Schedules (see page 42). 38 oracle Guide

39 Probe Configuration 13. After you have filled the above fields, click QoS List. The QoS list dialog appears. 14. Right-click in the grid view and select New from the context menu. The Edit dialog appears, with the name of the checkpoint (Custom-checkpoint in this example). 15. Enter the necessary details such as Name, Description, Unit, Metric, Abbreviation and Max value. 16. Click OK in the Edit dialog and then in the QoS list dialog. Now, you need to add the thresholds and scheduling information. For this refer Thresholds (see page 40) and Schedules (see page 42). Chapter 4: oracle Configuration 39

40 Probe Configuration Thresholds The thresholds 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, and so on 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. Follow these steps to add a threshold object: 1. In the Thresholds section of the Edit template checkpoint dialog, right-click and select the New option. The Edit Threshold dialog appears. 40 oracle Guide

41 Probe Configuration 2. Enter and select the information in the required fields as shown below. The fields in the above dialog are explained below: Threshold object name Specifies the monitoring object name, if applicable or default. Some special locked_users). Threshold value Defines the value used for threshold evaluation. Current interval value Severity If invoked from the status report, it contains the last measured value. Defines the alarm severity. Message Defines the name of message used for threshold alarm. Message text Defines the 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 Specifies the list of variables available in the checkpoint. Chapter 4: oracle Configuration 41

42 Probe Configuration 3. After you have entered the information in the Edit threshold dialog, click OK to add the threshold object. Schedules You can specify a schedule for the checkpoint to run. A schedule is a definition of an from/to and date from/to values. 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 check interval or exceptions of it. Rules and exceptions cannot be mixed in one checkpoint. Follow these steps to specify a schedule for the checkpoint: 1. Right-click in the Schedules section of General tab and select New from the context menu. 42 oracle Guide

43 Probe Configuration The Schedule ID dialog appears. Notes: If only Date from and Time from are defined, first execution can be defined. Run once will cause the checkpoint to run only once a day in the defined period 2. Click OK. A confirmation message appears asking you to confirm the modification of the template. 3. Click Yes to continue. Note: 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. 4. On clicking Yes, the new checkpoint is added to the Templates tab of the oracle probe GUI. Chapter 4: oracle Configuration 43

44 Probe Configuration Edit Checkpoint 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 as static. To edit the properties for a checkpoint, locally for a profile, you need to change the profile, in the checkpoints list in the Profiles dialog, to static. After modification, the new settings will be valid for this profile only. Similarly, to edit the properties for a checkpoint template, you need to double-click the checkpoint in the Profile list or Templates tab. After modification, the new settings will be valid for all profiles, unless overruled by static settings in the profile. Follow these steps to edit a checkpoint: 1. On the Templates tab, double-click the checkpoint to modify. The Edit template checkpoint dialog appears. 44 oracle Guide

45 Probe Configuration 2. Perform the required modifications in the upper part of the dialog. 3. Right-click on the threshold object, in the Thresholds section, that you want to modify and select the Edit option. The Edit threshold dialog displays. 4. Perform the required modifications and click OK. The OK button is disabled till you do not modify any parameter. The threshold object gets modified. 5. Similarly, you can modify the schedule for the checkpoint in the Scheduling section of the Edit template checkpoint dialog. 6. After the modifications are done, click OK on the Edit template checkpoint dialog. The Modify template confirmation dialog appears. 7. Click Yes to save the modifications, else click No. Chapter 4: oracle Configuration 45

46 Probe Configuration The Status Tab The status is presented in a hierarchical fashion, with 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. Through the Status dialog, the properties for an individual checkpoint object can also be modified. Follow these steps: 1. Select a profile and a monitored checkpoint in the left pane. 2. 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. 3. Refer section Editing a checkpoint (see page 44) for further information. 46 oracle Guide

47 Probe Configuration The Groups Tab This tab lets you create multiple groups which can be associated with profiles. You can add, copy, modify, or delete a group. Follow these steps to add a group: 1. Right click in the Group tab and select the New option. The Add New Group dialog appears. 2. Enter a name for the group and click OK. The New Group dialog appears. 3. Enter a description for the new group in the Description text box. Chapter 4: oracle Configuration 47

48 Probe Configuration 4. From the Group checkpoints section, select the check box(es) for the check points that you wish to enable for the new group. 48 oracle Guide

49 String matching methods 5. Click OK. The new group is now added in the Groups list. 6. The new group can now be selected from the Group drop-down in the Edit Profiles dialog. Similarly you can copy, modify, or delete a group. 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. Chapter 4: oracle Configuration 49

50

51 Chapter 5: QoS Threshold Metrics Many CA Nimsoft Monitor probes ship with the default QoS threshold values set. The default threshold values give an idea of the type of values to enter in the metrics fields. To help in tuning thresholds and reducing false-positive alarms, this section describes the QoS metrics and provides the default QoS thresholds. This section contains the following topics: Checkpoint Metrics (see page 51) oracle Alert Metrics Default Settings (see page 56) Checkpoint 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. Percentage 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 Chapter 5: QoS Threshold Metrics 51

52 Checkpoint Metrics oracle QoS Metrics This section contains the QoS Metrics for the oracle probe. QoS Name Units Description QOS_ORACLE_active_connection_ratio pct Calculated as: (active/total) * 100. Monitors % of active connections to total allowed connections. QOS_ORACLE_active_users count Checks for currently active Oracle users QOS_ORACLE_buf_cachehit_ratio pct 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. QOS_ORACLE_buf_cachehit_ratio_users pct Same as buf_cachehit_ratio, only per user. It helps to find a poor query or application. QOS_ORACLE_check_dbalive State This checkpoint tries to connect to an instance. In case the connection cannot be established, an alert is generated. Note: This checkpoint cannot be deactivated. QOS_ORACLE_database_size MB Monitors free space in database and log files in %. Source: Query. QOS_ORACLE_datafile_status State Reports all files being in other status as 'ONLINE'. QOS_ORACLE_dataguard_gap Archivefiles Monitors Dataguard Instance Gap QOS_ORACLE_dataguard_status State Monitors Dataguard Instance Status QOS_ORACLE_dataguard_timegap Seconds Monitors Dataguard Instance Timegap QOS_ORACLE_dbfile_io Percent 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. 52 oracle Guide

53 Checkpoint Metrics QoS Name Units Description QOS_ORACLE_dict_cachehit_ratio Percent 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. QOS_ORACLE_enqueue_timeouts Timeouts Measures number of table and row locks (acquired and converted) that timed-out before they could complete. QOS_ORACLE_extents Extents Reports tables that have more extents as threshold. QOS_ORACLE_flash_recovery_area_memory _free Bytes Monitors flash recovery area memory free QOS_ORACLE_Gc_av_lock_get_time Milliseconds Measures average interval global lock get time. QOS_ORACLE_gc_blocks_corrupt Blocks Measures number of blocks corrupted during transfer. QOS_ORACLE_gc_blocks_lost Blocks Measures number of blocks lost during transfer. QOS_ORACLE_gc_convert_timeouts Timeouts Measures number of times lock converts in global cache timed out. QOS_ORACLE_gc_cr_timeouts Timeouts Measures number of times process request for a consistent read block timed out. QOS_ORACLE_gc_fusion_wrt_rto Percent Calculated as: QOS_ORACLE_gc_service_util Percent 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. 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. Chapter 5: QoS Threshold Metrics 53

54 Checkpoint Metrics QoS Name Units Description QOS_ORACLE_lib_cachehit_ratio Percent 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. QOS_ORACLE_lock_waits Seconds Measures number of time, user session was waiting for a lock in last interval. QOS_ORACLE_locked_users Count Measures the number of locked users in an instance and displays their list. QOS_ORACLE_login_count Users Measures the number of currently active user session except for user "SYS" and "SYSTEM". QOS_ORACLE_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. QOS_ORACLE_memory_usage Bytes Measures total amount of memory (in bytes) user sessions consume (PGA). QOS_ORACLE_mts_response Milliseconds Monitors the average MTS server response time in ms. QOS_ORACLE_mts_uga_memory Bytes Monitors the MTS user memory consumption. QOS_ORACLE_mts_wait Milliseconds Monitors the MTS server wait time in ms. QOS_ORACLE_resource_utilization_count count Monitors database utilization for resources (Processes, Sessions and Transactions) QQOS_ORACLE_resource_utilization_percen t Pct Oracle resource utilization percentage QOS_ORACLE_rollback_seg_status Availability Lists rollback segments which are not QOS_ORACLE_session_waits Waits Measures number of waits in interval. QOS_ORACLE_sga_memory Bytes Measures SGA memory usage in bytes by pool. QOS_ORACLE_sga_memory_free Bytes Measures SGA free memory usage in bytes. 54 oracle Guide

55 Checkpoint Metrics QoS Name Units Description QOS_ORACLE_sort_ratio Percent 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. QOS_ORACLE_system_statistics Count Displays available system statistics counters. These counters contain number of events since the instance start. QOS_ORACLE_system_statistics_delta Count System statistics delta QOS_ORACLE_system_waits Waits 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. QOS_ORACLE_tablespace_alloc_free Pct Measures percentage of free space in already allocated files (not considering the AUTOEXTEND parameter). Note: This checkpoint is set inactive per default. QOS_ORACLE_tablespace_free Percent Measures percentage of free space in a tablespace, considering the maximal possible tablespace size. Note: Oracle Enterprise Manager Console does not consider maximum possible tablespace size on most of its screens (status 9i)! QOS_ORACLE_tablespace_size MB Measures the total tablespace size. QOS_ORACLE_tablespace_status Availability Lists tablespaces that have status different as QOS_ORACLE_tablespace_temp_free Bytes Monitors temporary tablespaces for space left QOS_ORACLE_user_locks Locks Lists user holding a lock. Helps to identify user, blocking other sessions (checkpoint locked_users). Chapter 5: QoS Threshold Metrics 55

56 oracle Alert Metrics Default Settings oracle Alert Metrics Default Settings This section contains the alert metrics default settings for the oracle probe. Alarm Metric Warning Threshold Warning Severity Error Threshold Error Severity Active connection ratio Major Active users Information Description Monitors the number of users having an active transaction at the moment of snapshot. Monitors the ratio of active connections to the total allowed connections. Buff cache hit ratio 95 Warning 75 Major Monitors the buffer cache hit ratio Buff cache hit ratio users 95 Warning 75 Major Monitors the user buffer cache hit ratio Chained rows Major Check for chained rows in database Check db alive Major Database size Information Monitors connectivity to the database instance Monitors space size (in bytes) for each database, logs, and data files together. Datafile status - - ONLINE Major Monitors the status of the datafiles Dataguard gap - 10 Major Monitors dataguard instance gap Dataguard status Major Monitors dataguard instance status Dataguard timegap Major DB file IO Major Dict cachehit ratio 90 Warning 75 Major Enqueue timeouts Major Extents Major Flash recovery area Major Avg global lock get time Major Monitors dataguard instance timegap Monitors the database file I/O operations Monitors the dictionary cache-hit ratio Monitors number of table and row locks that timed out Check for data fragmented segments Monitors flash recovery area memory free Monitors the average global lock get time (ms) 56 oracle Guide