Hello, we recently installed the early adopter Oracle Database Insights feature on a problematic database in our environment. The database itself runs on Oracle 12.
We are currently seeing abnormal peaks in disk reads and user calls, that definitely look out of scale in respect to the scope of this database; thus, before going forward with our analysis, we would like to better understand how to read certain data. Sure, some of it may look obvious to read, but we want to rule out everything since it is our first time using this feature.
The DBA team is just baffled as we are, because such readings are definitely too high for this database.
Specifically, we are wondering:
Thanks in advance for your help.
Solved! Go to Solution.
How is the DB endpoint configured: via SID or serviceName?
What type of Oracle deployment is it: single instance, multitenant (CDB/PDB), or RAC cluster?
Not all combinations are currently fully supported by DB Insights, so please also check the prerequisites listed in docs: https://www.dynatrace.com/support/help/shortlink/database-insights#prerequisites.
Thank you, @dglugla.
Eventually, we came to this same conclusion yesterday: the issue was in the endpoint being configured via serviceName as opposed to SID. We are now seeing correct metrics.
(On a side note: strangely enough, we did a parallel test on a similarly configured database, but in that case we could see proper metrics even when configured via serviceName. Therefore, at first we didn't think this was the issue.
At this point, I wonder what could be a relevant difference in this regard between the two databases; in any case, such issue seems related to the DB itself and not to how the Insights feature operates. If anyone experiences a similar issue, my advice would be to try different endpoints.)