[ngw] GWDVA crashing?
gene at genehoman.com
Thu Oct 6 01:14:14 UTC 2016
The below is from running 188.8.131.52 build 124595 ?
>>> "Joe Brugaletta" <JBrugaletta at braytonlaw.com> 10/05/16 3:57 PM >>>
for last day or so multiple gwdva processes on the server are eating quite a bit of CPU. Not sure if its playing catch up or what..and users are reporting searches are very slow.
(this paste from TOP might not come through sorted properly..)
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
15504 root 25 5 353m 193m 11m R 96 1.9 3:29.30 gwdva
15344 root 25 5 249m 82m 10m S 50 0.8 1:06.60 gwdva
15688 root 25 5 162m 22m 9316 S 50 0.2 0:22.35 gwdva
6039 root 20 0 521m 269m 9904 S 3 2.7 356:41.91 gwpoa
>>> "Gene Homan" <gene at genehoman.com> 10/4/2016 8:59 AM >>>
Make sure you have more than one DVA assigned to your POA. I assign up to three per POA and stagger the load among a pool of three DVAs for all my POAs and WebAccess services in the same network location. The WebAccess service puts a heavier load on the DVAs since its asking the DVA to convert the file to HTML format while the POA just asks for "word lists" to be sent back.
Make sure you are running build 124209 or newer of the DVA where Micro Focus improved the behavior of the DVA. So 184.108.40.206 would have the DVA fixes in it.
Check the DVA log files (hope it was in verbose mode). The latest DVA will auto cleanup the TMP files left behind by the Oracle Stellent subroutine.
>>> "Joe Brugaletta" <JBrugaletta at braytonlaw.com> 10/04/16 10:30 AM >>>
Every night at 8pm I get an email saying "QuickFinder Indexing will not start until the configured DVA is active"
If I check "rcgrpwise status" the gwdva is "dead". Anyway to figure out whats happening? or any temp files that need to be cleared?
More information about the ngw