We seem to have intermittent connectivity issues client to server. Progress bars will just show connecting and just not do anything, sometimes it will eventually throw up and error that "the server can't be reached at this time".
I can contact the server through ssh and remote desktop with no problem. Memory seems not to be an issue, nor processors; they seem to be at a reasonable state. The logs are either no help, or I am not looking for the right things.
Can anyone give me some direction on where to look and what to look for?
Version 4.5.6_GA_1044.MACOSXx86 Jul 6, 2007
OS X 10.4.10
2 x 2 GHZ Dual Core Intel Xeon, 6 GB RAM
-zimbra related logfiles: /docs/ne/latest/administration_guide/9_Monitoring.12.1.html#1075561
feel free to attach your logs if you'd like (/opt/zimbra/log/mailbox.log & /var/log/zimbra.log being the biggies)
-your system logfiles
-you also might want to analyze some tcpdumps
-as you're an NE customer you could also open a support request if you'd like someone to take a look
(of course there's also 4.5.9 you can upgrade too, but that's an asside)
Thank you for your quick reply.
If we don't see something from these logs, I will open a service request. I did not see mailbox.log in /opt/zimbra/, there is a mail.log? in /var/log/ is that the same? The logs are both well beyond the file size attachment suggestions, will that be a problem? The zimbra.log is 22 MB and the mail.log is 65 MB
meant to say /log and not /logs
so it's /opt/zimbra/log/mailbox.log
Certainly crop portions and attach them in a zip as needed.
Of course you can host them somewhere on your own or another free location. (the yousendit's etc)
There's also zimbra private pastebin - collaborative debugging tool if needed.
-If you end up contacting support for it there will be ways of getting them the logs.
But you can certainly ask us, aka: 'the free forum support' as well!
Ah, I see where I went wrong with the log location.
We did find a warn in the log. Is it relevant?
mailbox.log.2007-11-08:2007-11-08 13:59:59,107 WARN [LmtpServer-16]
[name=emailuser@ourdomain,org;] dbconn - Connection pool is 75%
utilized. 12 connections out of a maximum of 12 in use. Turn on
debug logging for zimbra.dbconn to see stack traces of connections not
returned to the pool.
could you attach your output of:
zmprov gs server.domain.com (zmprov gacf is a little less accurate if you've changed your settings in the individual server vs globally)
Looks like it is going to have to be 3 posts as the output is too long. Thanks for your help.
zmail:~ zimbra$ postconf
zmail:~ zimbra$ zmlocalconfig
zmail:~ zimbra$ zmprov gacf
My best guess is that your database connections can't keep up with the load or something.
Originally Posted by dketchum
12 out of 12 is not 75% -But I think it's just that the warning crops up at 75%
So if you wanted to track it down-
and add: log4j.logger.zimbra.dbconn=DEBUG
But why don't you make a good backup and upgrade to 4.5.9 because there have been a lot of performance improvements.
As you on NE:
-First use zmbackup (the backups your currently used to)
-Though I know you can restore from the backups, if you can afford the extra few min of downtime it often helps to have the structure easily accessible:
Obviously backup any other mounted locations that you use, the below is just assuming everything's stored in /opt/zimbra/
exit/switch back to root
ps aux | grep zimbra (kill -9 pid on any remaining)
rsync -avHK /opt/zimbra/ /backup/zimbra
When that finishes, if you want you can tar it up as well for 'easy' movement to another location:
tar -zcvf /backuptar/zimbra.456_1044.200711##.backup.gz -C /backup/zimbra .