As part of OAT, we have defined which monitoring metrics and thresholds we want to monitor and deployed these to the OAT servers.
Now we need to document that the metrics we defined are on the server.
I figured that would be east
So i got to Enterprise -> Monitoring Templates, Choose a template and click Compare
Then i select the targets where that database instance template is deployed (also visible in the past operations tab)
I run the compare and get back ... Differences on every host.
So i think, ok something may have changed - so lets apply the template, using the replace all metrics on target (i am assuming here this only means metrics for the target type in the template) and again hit apply.
Then rerun the compare expecting that now all metrics with thresholds will match the target ... but no.
So i think maybe the template is from an older version of OEM and those metrics dont apply to db instances.
So create a new template, deploy it, run the comparison and again there are metrics which do are not applied.
I am assuming if the metric is disabled, it would not be deployed - makes sense but would be good if the comparison screen gave a way to omit those.
However there are metrics with thresholds which have a collection schedule which also do not get applied.
Anyone suggest why this is the case ?
Examples of non disabled metrics with thresholds not being deployed include
- dbjob_status:Broken job count - i know dbms_job was replaced by dbms_scheduler
- Deferred Transaction Count Warning > 100
- Data Guard Status if error contains "warning" or "error"
- Global Cache Average Get Time Warn > 0.3, Critical > 0.6 (we are deploying to RAC instances) ?
Maybe some of these metrics are there for older versions of Oracle that might be monitored by this OEM version ?
have some of these been superseeded by other metrics which might be deployed ?
If it has not deployed metrics because there is nothing of the version to monitor, why does the compare still show them as differences and why doesnt it give any reasons for why.
Do i really have to go through every metric to work out why.
I can remove the failed ones from a custom template and redeploy but would be good to understand what is going on first.
The aim was to provide the auditors with proof we are monitoring the metrics we said we would.
Is there anywhere OEM shows this because the comparison report doesnt seem to help ?
Thanks in advance.