Am 01.07.2011 um 16:33 schrieb Reuti: > Hi, > > Am 01.07.2011 um 14:27 schrieb Fabio Martinelli: > >> I've enabled reporting in SGE 6.2u5 and I've installed and configured SGE >> dbwriter to dump that file inside a MySQL DB, that works fine but the >> reporting file is deleted like a consequence of this dump step; please how >> to preserve the reporting file ? >> >> just an example of why I want to preserve it, maybe in the future I could >> replace MySQL with PostgreSQL and import all my cluster history again; also >> sometimes I run 'grep' against the reporting file. >> >> REF. >> http://wikis.sun.com/display/GridEngine/ARCo+Configuration+Files+and+Scripts > > you could change the source to write a second persistent file. > > I don't know, how the dbwriter is invoked. Maybe a wrapper could do which > will call the real dbwriter after you first append the new lines to another > file. > > A less reliable solution in your situation could be syslog-ng. Since version > 3.1 it can follow a file by inode or name (i.e. even rotating ones) and send > the new lines to somewhere else via the syslog protocol. I use this to > collect the logfiles from IBM's Storage Manager at a central place from > certain machines, and it's also working on Windows this way. But I'm not > sure, whether it misses a lines here and there. As I'm mostly interested > whether the backup was successful, it doesn't matter in my case.
I completely forgot, before syslog-ng I used: tail --follow=name -n 0 /var/spool/tsm/dsmerror.log 2>/dev/null | logger ... -- Reuti > -- Reuti > _______________________________________________ > users mailing list > [email protected] > https://gridengine.org/mailman/listinfo/users _______________________________________________ users mailing list [email protected] https://gridengine.org/mailman/listinfo/users
