Alarms auto acknowledge issues

book

Article ID: 207455

calendar_today

Updated On:

Products

DX Infrastructure Management

Issue/Introduction

We have lot of alarms in the UMP console, e.g., oracle and logmon alarms, which are not getting auto cleared even though issue has been resolved. Please let us know how this can be fixed as it creates lot of confusion with new alarms getting triggered in console.

Cause

- probe configuration

Environment

Release : 9.2.0

Component : UIM NAS

Resolution

oracle alarms

The oracle probe being used should be version 5.45 GA. See:

https://techdocs.broadcom.com/us/en/ca-enterprise-software/it-operations-management/ca-unified-infrastructure-management-probes/GA/alphabetical-probe-articles/oracle-oracle-database-monitoring/oracle-im-configuration.html

Regarding the customer's oracle probe and autoextend alarms,

If in Oracle, autoextend  is enabled, then in the oracle probe, consider that the alarm issued here is really only for Informational purposes. The alarm should clear again when the database server has allocated more size for the datafile, and is only meant as an informational message that the datafile will autoextend.

tablespace_free is the correct checkpoint to use since it considers the maximum possible allocation. tablespace_free measures the percentage of free space in a tablespace, considering the maximum possible tablespace size. It will only give you an alarm when you are coming close to not having more free space at all for the datafile. Say you have 100 K allocated and autoextend = yes up to 1000 K you will not get an alarm before nearing the 1000 K mark using this checkpoint.

Note also that you can exclude certain tablespaces.

The oracle probe provides the ability to exclude tablespaces that match a pattern string from monitoring, e.g., like SYS, SYSTEM, ROLLBACK, UNDO or TEMP.

  • Use excludes:

allows you to define objects to be excluded from monitoring. The probe uses regular expressions to search patterns that are excluded from the checkpoint.

  • Excludes list:

allows you to specify the patterns to be excluded in the checkpoint. Click the Excludes list  (  )

icon to specify existing patterns or define pattern in the Exclude list dialog. You can define objects that you do not want to monitor in the template checkpoint hence stop the alarms in that manner.
  • (From version 5.10 or higher: for the tablespace_free checkpoint) 

'Enable Threshold on Space'

- enables you to define another/additional threshold value for the tablespace_free checkpoint. If you select this option, the probe evaluates the values specified in both the Threshold value and Threshold Value for Space (In GB) fields before generating an alarm for threshold breach.

In my opinion and experiences in the past with Oracle monitoring when autoextend is enabled for tablespaces, I personally wouldn’t even bother monitoring any tablespaces that have AUTOEXTEND set to yes, because the oracle feature is robust and automatically extends the tablespace. Otherwise it creates a lot of alarm traffic with little to no value as a result. Its more of an  info type of alarm since the tablespace is automatically handled/extended. Of course in the end, it’s the customer’s choice.

logmon alarms

A good place to check for info like this is in the UIM community.

Here are two examples/descriptions of how to clear logmon alarms for configured Watchers when a given error/issue/problem is actually resolved/cleared, but the alarm remains.

https://community.broadcom.com/communities/community-home/digestviewer/viewthread?MID=746852

https://community.broadcom.com/communities/community-home/digestviewer/viewthread?MID=773316

For probes in general, here is a relevant KB Article: 
 
Why do some alarms not clear automatically and how can I clear them?