Re: (ITS#5439) back-bdb race condition

[quanah@zimbra.com]

--On Monday, March 31, 2008 6:27 PM +0000 rein@basefarm.no wrote:

> We did see segmentation faults both Friday and Monday (not very much
> going on during the weekend..), but so far those appear to be related to
> ITS#5445. A slapd and db_checkpoint (run out of cron) deadlock which I
> hope is due to the BerkeleyDB bug mentioned in ITS#5391 has also
> occurred, so I'll upgrade from 4.6.18 to 4.6.21.1 to see if that problem
> goes away.


Why do you run db_checkpoint out of cron?  Slapd will already take care of 
checkpointing the database for you using the "checkpoint" directive...

--Quanah


--

Quanah Gibson-Mount
Principal Software Engineer
Zimbra, Inc
--------------------
Zimbra ::  the leader in open source messaging and collaboration