Oracle instance CI duplication on discovery

cyked · ‎10-15-2015

We're seeing issues with multiple CIs representing a single oracle instance. Normally you see duplicate CI creation when something like the params on the running process classified change. Using process handlers addresses this.

However, what I see on the ora_pmon oracle instance running process records, there are NO command or paramater information recorded. Currently I can't compare the current actively discovered/updated oracle instance to the "pervious" one because the difference in creation between the two is > 30 days and the running process record of the "previous" CI has already been nuked after being "absent" for over 30 days.

Does anyone else have this issue? I'm not finding what discovery identification is finding different in the existing CI record/running process record to create a new CI.

cyked · ‎10-27-2015

Ticket was updated today. Tech chalked it up to potentially being system upgrades/patches as the issue. Recommended removing the duplicates and moving on.

View solution in original post

Valor1 · ‎10-15-2015

What Version SN are you running?

I believe that (at least in Fuji) this is handled by specific Oracle sensors.

cyked · ‎10-16-2015

Fuji patch 6 in Prod. lower environments are patch 9, but we don't actively run disco schedules in the lower environments.

Valor1 · ‎10-16-2015

Hmm.. and to confirm, you're using the OOB Oracle Sensors?

I don't have an Oracle test environment but Steve Bell (Cloud Sherpas) or Mark Amann (Cloud Sherpas) may have an idea.

cyked · ‎10-16-2015

Yep. the sensor for oracle instance detail is 11-5-2013, ben.yukoch