For anyone thats updated to Patch 2 to fix the problem with performance data not showing, how long did it take before data started to appear?
I installed the patch yesterday around lunch time and this morning I still do not see any data. The row count on vpx_hist_stat before I left work yesterday was 100,000,000 rows, and now I'm at 99,000,000. Honestly I don't know if its even doing anything, or maybe something went wrong with the upgrade. We have 7 ESX Servers hosting about 55 VMs.
Does anyone know if there are logs for the rollups, or if there's any other way to check to verify that my performance stats are starting to roll up? We are running VC on a Oracle 10G RAC cluster.
Make sure that the date, time, and timezone are set properly on the VC server, your management PC, as well as the ESX servers themselves.
I've seen perf data refuse to come up because my work PC (running the VI client) didn't have the date/time set close enough to the VC server.
Paul
I have an open SR for
-the database issue with tempdb (vpx_hist_stat) growing ... growing ... growing and
-slow response when looking at perf data in virtual center.
Before patch 2, there was a manual process to fix the issue which involved truncating the vpx_hist_stat table and running the stored procedure vpx_stats_rollup (but I am not very sure about this process since most of the forum articles mention the procedure but do not spell it out, recommending that you open an SR). Here is the skinny on the communities pain with this issue going back to last year: http://www.vmware.com/community/thread.jspa?threadID=61856
http://www.vmware.com/community/thread.jspa?threadID=65255&tstart=0
All this is no more and fixed with patch 2 for VirtualCenter 2.0.1.
http://www.vmware.com/support/vi3/doc/vc-201-200702-patch.html
Support recommended installing this patch to correct the performance and tempdb issue.
So, after I apply the patch, is vpx_hist_stat going to be reduced to a much smaller size, or is there a manual process to truncate the table? ? ? ?
FYI, anyone planning to upgrade to patch 2 should check out this thread by Hicksj (specifically if you are running HA or have problems with the host agent upgrades):
http://www.vmware.com/community/thread.jspa?threadID=74479&tstart=0
100m rows? damn! I have 7 hosts, 110 VM's. I was not able to run the workaround jobs for a few days this weekend and when I came in today I had 14m rows. I ran the patch and after about an hour I had 5m rows.
you can do a 'sp_who active' to see how long and how many i/o's the internal patch 2 jobs are doing.
What statistics level are you running? I'm running level 2.
Ben