2015-01-13_13-28-52.png

2797341

RachelB-Support-Oracle

Hiya

This is a repository metric (ie. a metric which is calculated in the repository by combining data from the cluster metric mgmt_rac_services and  the  rac_database metric, service_performance).    It seems like this metric has been coded only to allow the editable threshold for 'database service status'.

I haven't tried this, but  if you also want to receive alerts based on some criteria relating to the 'Preferred Instances list' you might be able to create your own repository based metric extension to do this.

You can see the data from the repository by doing something like:-

select collection_timestamp, value, key_value from sysman.mgmt$metric_details

where target_name='rac11204'

and target_type='rac_database'

and metric_name='db_service_status'

and metric_column='prefInstList'

Then you can try creating your own repository side metric extension from

Enterprise/Monitoring/Metric Extension

Create/Repository Side Metric Extension

The SQL would need to be something like:

SELECT target_guid,

        value AS PrefInstance

FROM   mgmt$metric_current

WHERE  target_type = 'rac_database'

        AND metric_name = 'db_service_status'

        AND metric_column = 'prefInstList'

I stopped here, because i'm not sure what kind of criteria you want to apply to this preferred instances list information.  But there is a really good example here on Courtney's blog (which is a similar requirement to yours (to collect something which is not currently thresholdable).  See:

EM12c Release 4: Using Repository Side Metric Extensions to Alert on Collected Metrics - Courtney Llamas

Thanks

Rachel