Re: regarding interconnect latency

Posted by ErmanArslansOracleBlog on Sep 12, 2018; 3:49pm
URL: http://erman-arslan-s-oracle-forum.124.s1.nabble.com/regarding-interconnect-latency-tp6698p6709.html

Ok. At first, I didn't see the sqlid and session info..
now it is ok.
Well. it is interesting..

1)what is the cost of that query?
2)Is that still executed in the system? (so is it now executed with reasonable durations)
3)there are resource manager specific bugs, did you check them?
   3a)actually, your OS is fully loaded, I mean your Cpus are fully loaded.. It is not the resource manager that makes these processes wait.. They already allocated the cpu and they were spinning on that. So by knowing this, we can suspect from the wait -> event GC CR BLOCK2 WAY ..  Again there are bugs for this GC wait. Did you check them? (I say check the bugs, because this problematic query seems to be an ordinary one...)
4)Are you sure about that sql? Because these seems to be coming from a customization, and maybe you were just seeing the last sql executed, but not the one , which was currently executed by those cpu intensive session? Are you sure it is the correct sql?
5)Did you speak with the developers, who developed this customization -> XXSAS_AP_VERF_DESK ? (maybe they can give you a clue.)