Discussions
Categories
- 17.9K All Categories
- 3.4K Industry Applications
- 3.3K Intelligent Advisor
- 63 Insurance
- 536.4K On-Premises Infrastructure
- 138.3K Analytics Software
- 38.6K Application Development Software
- 5.8K Cloud Platform
- 109.5K Database Software
- 17.5K Enterprise Manager
- 8.8K Hardware
- 71.1K Infrastructure Software
- 105.3K Integration
- 41.6K Security Software
EPM 11.2.6 HFM Performance Bug
Answers
-
If it's only relevant for multi app server environments then we're safe (we've only got one big app server to avoid the sync time discussions).
So thanks again Steve!
CAWD forces a calc yes, but not all of our planned consolidations are CAWD and we would like to see correct figures ;)
[update]
Preliminary results (as I didn't restart EPM services yet, of which I'm not sure is required for the setting), but setting usecalcstatuscache to 1 seems to save an additional 10%. So total saving now at 27%.
-
For those waiting for 11.2.8, tentative release date 5:th of April, patch 33899947. Also planned for 11.2.5, but no ETA.
-
Hi Steve, if I remember correctly, Oracle changed the status mechanism in 11.1.2.4.208 because of that issue but the performance impacts all the systems. Not just the ones with multiple servers.
Cheers,
Thanos
-
@Thanos A :thanks for the clarification. Then I will need to discuss internally if they want to take the risk.
-
Well this is strange. I got a call from the 11.2.6 client that HFM Consolidation performance had worsened for the test POV since the work 2 weeks ago.
I therefore ran the test consolidations (via Maestro!) in UAT and Prod and - yes - Consolidation was 100% slower than my tests outlined above. Oh dear.
Don't despair though! After restarting the application across all App Servers (via Insights) and then running the POVs again I can get the hugely improved Consolidation timing from my posts above. UAT had been up for 4 days solid and Prod for 2 days at this point.
Conclusions:
- Is there a latent code issue (e.g. memory leak?) that affects timings if XFMDataSource is up (and idle?) for any length of time (24-48 hours)?
- In this case is DSSStartUpOption >0 actually working against me here, since if I don't let XFMDataSource stay up, then it should shut down and restart as needed - effectively automating this flush. I will therefore set DSSStartUpOption to 0 and see if the performance issues recur.
- I realise that in and active (i.e. Live) HFM system a regular restart of HFM Java is going to be tricky to arrange BUT DSSStartUpOption may prove to be the saviour here.
Stay tuned....
Steve
-
11.2.8.1 patch is now released. You still can't find it without the patch number tho: 33899947
-
Hi Steve,
That's not good news. I think that 11.1.2.4.106 had the same issue.
Your comment re DSSStartUpOption = 0 it will be an improvement for any application that is not 24/7 up and running.
Cheers,
Thanos
-
@deeziel - thanks for the update and patch reference
Steve
-
What bugs does this patch address? Tried searching on site but not much info on the specifics. i doubt it fixes our multi consolidations running at once degradation in performance because Oracle has yet to hear from Microsoft on resolving.
-
The 11.2.8.1 patch did not help improve anything on HFM consolidation performance issue, looks like next step is with Microsoft :)
