12c / Cloud Control

Enterprise Manager 12.1.0.4 and Server Adaptive Threshold Metric

One of the big changes in 12.1.0.4 with the 12.1.0.6 database plugin is most server generated alerts are now evaluated by the agent.

The “Server Adaptive Threshold” is not one of the metrics which utilises this new behaviour but remains for backward compatibility. What this means is you cannot set thresholds for these metrics via the Enterprise Manager metrics page and you may still may get alerts coming through.

This is what happened to me and I was scratching my head trying to work out why certain alerts were coming through which looked similar to alerts in my monitoring templates.

There’s a metalink doc which describes the issue – How to Prevent Alerts from the Server Adaptive Threshold Metric (Server_Adaptive_Threshold_Metric) in Enterprise Manager 12.1.0.4 onwards (Doc ID 1929003.1)

You can clear this threshold by going into the adaptive thresholds page in the performance tab but if you have many targets this can be very time consuming.

Luckily you can use the DBMS_SERVER_ALERT package to remove the problematic thresholds. This was fairly straight forward to automate and run against all my database targets. I tried this via the EM job function against all targets but had trouble with authentication on all hosts. Looks like I hit the issues described in How to Configure the Enterprise Management Agent Host Credentials for PAM and LDAP (Doc ID 422073.1)

To get this applied to my targets I used a little bit of perl and dbi running the following pl/sql

Now we can query all the databases to see if the metrics are gone and in my case they are no longer there.

Leave a Reply

Your email address will not be published. Required fields are marked *

CAPTCHA * Time limit is exhausted. Please reload CAPTCHA.