I want to run some ad hoc reports against the Resource history stored in DPA to try and correlate various counters against times of day when we experience connection failures.
We have a set of mirrored databases that stay running on the primary instance, but 2 or 3 times a day at regular times a number of applications wrongly think that the databases have failed over and try connecting to the mirror instead. This causes an error for which I get an alert. There is a consistent pattern to the times of day that these events occur and I want to find out if it is a resource condition on the primary server (SQL Server / Windows / ESX) that is the root cause. Since all this data is in one place inside DPA's database it seems a good place to start looking.
In the DPA reports section there are no reports against the Resources, only SQL queries.
I found a KB article about creating custom resource graphs but this was for NPM not DPA.
In any case, I need to able to filter by specific time ranges across several days and see if there any patterns in the data that match up with my alerts.
Are there any queries I can run direct against the DPA database that would help me?
Have you checked this script out? You might need to tweak it a bit for your exact purposes:
Also, we summarize resource metrics, but if you want them within the past 30 days, you should be fine.
You didn't specify your repo version, so I just included the one for SQL Server...