Hi Bill, just noticing that in the first instance it states "started by 1001" and in the 2nd it reports "started by autodeploy". The user id of autodeploy is in fact 1001. ?
Karen On Monday 29 October 2007 15:07:55 Karen Loughran wrote: > Hi Bill, yes, the commit.log shows the following when snapinstaller fails > from the cronjob: > > 2007/10/29 15:03:03 started by 1001 > 2007/10/29 15:03:03 command: /opt/solr/bin/commit > 2007/10/29 15:03:03 failed to connect to Solr server at > http://localhost:8080/solr/update > 2007/10/29 15:03:03 commit failed > 2007/10/29 15:03:03 failed (elapsed time: 0 sec) > > solr is running on port 8080 on the localhost. > > > When I subsequently run "commit -V" the log shows: > > 2007/10/29 15:06:35 started by autodeploy > 2007/10/29 15:06:35 command: /opt/solr/bin/commit -V > 2007/10/29 15:06:35 ended (elapsed time: 0 sec) > > Karen > > On Monday 29 October 2007 14:52:42 Bill Au wrote: > > Snapinstaller uses commit to notify Solr to open a new Searcher. Is > > there anything in the commit log which shows why the commit failed? > > > > Bill > > > > On 10/29/07, Karen Loughran <[EMAIL PROTECTED]> wrote: > > > Thanks Yonik and Otis, > > > > > > I noticed from the log (snapinstaller.log) the following when the slave > > > attempts to perform snapinstaller: > > > > > > 2007/10/29 10:30:02 started by 1001 > > > 2007/10/29 10:30:02 command: /opt/solr/bin/snapinstaller > > > 2007/10/29 10:30:02 installing snapshot > > > /opt/solr/data/snapshot.20071029102014 > > > 2007/10/29 10:30:02 notifing Solr to open a new Searcher > > > 2007/10/29 10:30:02 failed to connect to Solr server > > > 2007/10/29 10:30:02 snapshot installed but Solr server has not open a > > > new Searcher > > > 2007/10/29 10:30:02 failed (elapsed time: 0 sec) > > > > > > Running bin/commit -V works - the deleted record is then reflected in a > > > search > > > (without restarting tomcat). > > > > > > I then (testing delete of a different record) performed the snappuller > > > and snappinstaller by hand on the slave and it worked, ie, no error in > > > the log file and change reflected immediately ! > > > > > > So it seems that my cron job command is not working as expected: > > > > > > 0,15,30,45 * * * * /opt/solr/bin/snappuller && > > > /opt/solr/bin/snapinstaller > > > > > > ";" separation for multuple commands (as per example given in solr > > > document "CollectionDistribution") did not work for me so I used && > > > according > > > to crontab documentation. > > > > > > I also tried the following crontab job: > > > > > > 0,15,30,45 * * * * /opt/solr/bin/snappuller > > > 1,16,31,46 * * * * /opt/solr/bin/snapinstaller > > > > > > Whereby snapinstaller would run a minute after snappuller. But I still > > > get > > > the following: > > > > > > 2007/10/29 11:16:01 started by 1001 > > > 2007/10/29 11:16:01 command: /opt/solr/bin/snapinstaller > > > 2007/10/29 11:16:01 installing snapshot > > > /opt/solr/data/snapshot.20071029111236 > > > 2007/10/29 11:16:01 notifing Solr to open a new Searcher > > > 2007/10/29 11:16:01 failed to connect to Solr server > > > 2007/10/29 11:16:01 snapshot installed but Solr server has not open a > > > new Searcher > > > 2007/10/29 11:16:01 failed (elapsed time: 0 sec) > > > > > > I have also confirmed with a "whoami" in crontab that the cron job is > > > running > > > as the same user as if running scripts by hand and as the user > > > specified in > > > scripts.conf. > > > > > > Do you know whey snapinstaller will not work within crontab but will > > > work by > > > hand ? > > > > > > Thanks > > > Karen > > > > > > On Friday 26 October 2007 16:36:22 Yonik Seeley wrote: > > > > On 10/26/07, Karen Loughran <[EMAIL PROTECTED]> wrote: > > > > > But after distribution of this latest snapshop to the slave the > > > > > collection does not show the update (with solr admin query url or > > > > > via java query client) UNLESS I restart tomcat ? > > > > > > > > Sounds like a config issue with the scripts... pulling the snapshot > > > > is obviously working, but snapinstaller (calling commit) is broken. > > > > > > > > try running bin/commit -V by hand on the slave > > > > > > > > -Yonik
