Description of problem: The "Reindex Data Tables Nightly" option on large tables has caused daily trouble, until I discovered how to turn it off. modules/enterprise/server/jar/src/main/java/org/rhq/enterprise/server/system/SystemManagerBean.java Runs: ALTER INDEX {0} REBUILD UNRECOVERABLE When being run on Oracle, this prevents inserts while the rebuild takes place, which may be several minutes. This has caused agents to time out and false alerts to be generated. Unrecoverable is not a good idea as it locks the table during the calculation: https://community.oracle.com/thread/1103546?tstart=0 The default should be false, as well UNRECOVERABLE is not a good idea either on Oracle because of the locking.
The primary trouble is with the traits table, although I suspect the availability table may also cause trouble during reindex.
Merged PR45 master commit 78541021d881c60dd97e3b21e0af1b0d59cc84ad Merge: ba0c4d3 537c51f Author: jshaughn <jshaughn> Date: Tue Jul 1 10:19:50 2014 -0400 Merge pull request #45 from genman/BZ1094424 BZ1094424 - do not reindex tables by default; don't use UNRECOVERABLE Some more follow up work coming...
master commit a68c656eca8f0574dd551a6faf51a0df829f0f39 Author: Jay Shaughnessy <jshaughn> Date: Tue Jul 1 13:05:39 2014 -0400 [1094424] "Reindex Data Tables Nightly" setting can cause agents to time This is follow-up work to PR45. Add setting the nightly re-index option to false in db-upgrade. Also, remove unused numeric data tables from being processed if in fact the option is set to true. This whole option should probably go away, see the discussion in the PR.
Bulk close of items fixed in RHQ 4.12 If you think this is not solved, then please open a *new* BZ and link to this one.