Re: onbar error message

On Wed, 26 May 2004 10:00:51 -0400, Davorin Kremenjas wrote: It may be that during that time, there is a long transaction running which is not permitting the backed up logical logs to be released for reuse. If you can determine that this is the case (run onstat -l perhaps from a cron periodically or using the -r option to see if this is happening) then you need to do two things: - Increase the size or number of logical logs to accomodate these large transactions, and/or - Recode the tasks that are performing these huge transactions to break them into smaller logical pieces to reduce the dependency on log space and reduce recovery time in case of a crash. Art S. Kagel > Hello! > I have an Informix backup configured as an onbar + Legato Networker > combination with IDS 9.40 UC3 on Solaris 9 machine. We take dbspace backups > every night and continously backup logical logs using log_full.sh script. > All works well except one thing: if logical logs start to fill very quick > (around 50 logs in 2 minutes) error messages start to appear in bar_act.log: > > "The logical logs are full. ON-Bar may encounter errors writing to the > sysutils database." > > Weird thing is that logs do back up, which can also be seen from > bar_act.log: > > ... > 2004-05-26 04:15:17 29921 29919 Begin backup logical log 9234. 2004-05-26 > 04:15:18 29976 29974 /usr/informix/bin/onbar_d -b -l 2004-05-26 04:15:18 > 29921 29919 Completed backup logical log 9234 (Storage Manager copy ID: > 1085537717 1085537720). 2004-05-26 04:15:18 29921 29919 Begin backup > logical log 9235. 2004-05-26 04:15:18 29976 29974 A log backup is already > running. Can't start another. > 2004-05-26 04:15:19 29921 29919 The logical logs are full. ON-Bar may > encounter errors writing to the sysutils database. 2004-05-26 04:15:19 29921 > 29919 Completed backup logical log 9235 (Storage Manager copy ID: > 1085537718 1085537722). 2004-05-26 04:15:19 29921 29919 Begin backup > logical log 9236. ...etc... > > > I thought of increasing the size of logs so they don't fill up so quickly > but I'm not sure whether it will solve the problem. At this moment onconfig > looks like this: > > # Logical Log Configuration > LOGFILES 256 # Number of logical log files LOGSIZE > 512 # Logical log size (Kbytes)> > The reason why the logs start to fill so quickly is that we have update > statistics for tables and procedures run every morning at 04:15. > > Any ideas? > Thanks everyone.