Our production dovecot/postfix server has been stable for a number of years. In
the last month or so, we are seeing increasing errors such as these:
Dec 6 15:51:20 mail dovecot: imap(redacted at lilythicket.com): Warning:
Transaction log file /home/vmail/lilythicket.com/diana/Maildir/dovecot.index.log
was locked for 322 seconds
Dec 6 15:50:54 mail dovecot: imap(redacted at theormans.com): Warning: Maildir
/home/vmail/theormans.com/connieorman/Maildir/.Junk: Synchronization took 66
seconds (1 new msgs, 0 flag change attempts, 0 expunge attempts)
Dec 6 15:51:43 mail dovecot: master: Error: service(pop3-login): Initial status
notification not received in 30 seconds, killing the process
Dec 6 15:51:43 mail dovecot: master: Error: service(pop3-login): command
startup failed, throttling
Dec 6 15:51:43 mail dovecot: master: Error: service(imap-login): child 5868
killed with signal 9
Dec 6 15:51:43 mail dovecot: master: Error: service(imap-login): command
startup failed, throttling
Dec 6 15:55:31 mail dovecot: imap-login: Fatal: Corrupted SSL
ssl-parameters.dat in state_dir: Truncated file
Dec 6 15:55:32 mail dovecot: pop3-login: Fatal: Error reading configuration:
Timeout reading config from /var/run/dovecot/config
And so forth. Seems to be all over the place. The server slows down to a
crawl. Restarting dovecot or postfix has no effect on the problem. Only a
server reboot solves it, temporarily. Sometimes for weeks, sometimes for hours.
The hard drive SMART status reads okay.
During this time, of course, users cannot connect to check their email.
Thoughts on where to go to troubleshoot this and why it?s happening?
Ethon
