Am 06.05.2011 um 18:56 schrieb Stuart Barkley:
> On Fri, 6 May 2011 at 09:33 -0000, Dave Love wrote:
>
>> E.g. http://arc.liv.ac.uk/SGE/howto/rotatelogs.html#logrotate
>
> On systems where logrotate available (CentOS for us) this seems much
> simpler than the SGE logchecker script.
Whichever way you use: in case you configured local spool directories on the
nodes it must be implemented there too.
> I'm not sure why I couldn't find this information (looking back, I may
> have just not read to the bottom of the page where the useful
> information was). I had been manually rotating the accounting logs
> for a while, but was vaguely aware of logchecker.sh.
>
> Last week, I finally put the following into /etc/logrotate.d/sge_$CELL
>
> # rotate $CELL logs automatically
>
> /opt/sge_root/$CELL/common/accounting /opt/sge_root/$CELL/common/reporting
> /var/spool/$CELL/spooldb/messages
In `spooldb' shouldn't be any messages file. Do you mean the
qmaster/node[0-9][0-9] directories?
-- Reuti
> {
> compress
> nocreate
> dateext
> delaycompress
> ifempty
> # keep logs essentually "forever"
> rotate 5000
> weekly
> }
>
> SGE appears to handle log file rolling pretty well and just creates
> the log if it is missing. I used the "delaycompress" option "just in
> case" things where still writing to the files, but will probably
> remove it soon.
>
> I date stamp the files and want to keep everything for later analysis.
>
> Stuart
> --
> I've never been lost; I was once bewildered for three days, but never lost!
> -- Daniel Boone
> _______________________________________________
> users mailing list
> [email protected]
> https://gridengine.org/mailman/listinfo/users
_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users
