Mail server seriously messed up for weeks

Question

Level 1

19 points

Mail server seriously messed up for weeks

A couple of months ago I had a serious crash that among other things affected Mail and some other services. With much help I determined that certain servermgr bundles were corrupted and was able to replace them with backups which appeared to fix the problems. However, I have had intermittent problems with my mail server ever since. I had used the command line tool by petrobyte to rebuild my mail database and it solves the problems for a while. However, after a while mail quits working (though it still is running) and I have to rebuild again. This is happening about once a week now. I have tried to determine if anything such as spam assassin or clamscan was to blame and have turned them off without improvement. At this point I am at a loss how to stop this from continuing to happen. Sometimes it causes my whole server to stop working and auto restart itself. Any ideas would be greatly appreciated!

dual 1.8g desktop, Mac OS X (10.4.2)

Posted on Oct 10, 2006 4:38 PM

Reply

Answer 1

UptimeJeff

Level 4

3,479 points

Oct 10, 2006 5:10 PM in response to elmgator

When a failure occurs, what do the mail.log and system.log show?
More information is needed to provide some guidance...

What other troubleshooting have you performed?

Have you tried to verify or repair the disk using Disk Utility?

Jeff

Reply

Answer 2

elmgator Author

Level 1

19 points

Oct 10, 2006 5:49 PM in response to UptimeJeff

Thanks much for the offer to help me figure this out!

I have used disk warrior and diskutility and they typically do not find problems when this happens. Here is an example of the moment it starts to fill my system logs with errors. It inevitably happens at a pop login but the accounts differ.

Oct 10 13:58:48 uno pop3[7801]: login: [192.168.1.121] com answers123tjmings APOP User logged in
Oct 10 13:59:03 uno ctl_cyrusdb[8173]: checkpointing cyrus databases
Oct 10 13:59:03 uno ctl_cyrusdb[8173]: DBERROR db4: DB_LOGC->get: log record LSN 1/1768765: checksum mismatch
Oct 10 13:59:03 uno ctl_cyrusdb[8173]: DBERROR db4: DB_LOGC->get: catastrophic recovery may be required
Oct 10 13:59:03 uno ctl_cyrusdb[8173]: DBERROR db4: PANIC: DB_RUNRECOVERY: Fatal error, run database recovery
Oct 10 13:59:03 uno ctl_cyrusdb[8173]: DBERROR: critical database situation
Oct 10 13:59:07 uno postfix/smtpd[8171]: warning: 192.41.84.18: hostname pub-192-41-84-18.center7.com verification failed: Host not found
Oct 10 13:59:09 uno lmtpunix[8180]: DBERROR db4: PANIC: fatal region error detected; run recovery
Oct 10 13:59:09 uno lmtpunix[8180]: DBERROR: critical database situation
Oct 10 13:59:09 uno master[91]: service lmtpunix pid 8180 in READY state: terminated abnormally
Oct 10 13:59:09 uno lmtpunix[8181]: DBERROR db4: PANIC: fatal region error detected; run recovery
Oct 10 13:59:09 uno lmtpunix[8181]: DBERROR: critical database situation
Oct 10 13:59:09 uno master[91]: service lmtpunix pid 8181 in READY state: terminated abnormally
Oct 10 13:59:09 uno lmtpunix[8182]: DBERROR db4: PANIC: fatal region error detected; run recovery
Oct 10 13:59:09 uno lmtpunix[8182]: DBERROR: critical database situation
Oct 10 13:59:09 uno master[91]: service lmtpunix pid 8182 in READY state: terminated abnormally
Oct 10 13:59:09 uno lmtpunix[8183]: DBERROR db4: PANIC: fatal region error detected; run recovery
Oct 10 13:59:09 uno lmtpunix[8183]: DBERROR: critical database situation
Oct 10 13:59:09 uno master[91]: service lmtpunix pid 8183 in READY state: terminated abnormally
Oct 10 13:59:09 uno lmtpunix[8184]: DBERROR db4: PANIC: fatal region error detected; run recovery
Oct 10 13:59:09 uno lmtpunix[8184]: DBERROR: critical database situation
Oct 10 13:59:09 uno master[91]: service lmtpunix pid 8184 in READY state: terminated abnormally
Oct 10 13:59:09 uno lmtpunix[8185]: DBERROR db4: PANIC: fatal region error detected; run recovery
Oct 10 13:59:09 uno lmtpunix[8185]: DBERROR: critical database situation

Reply