QUE Failure

[phoenix]

Is it likely that you're running out of disk space? Try 'df -h' and see what's used/free.

[tbullock]

Disk space is plentiful

[zimbra@webmail libexec]$ df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/VolGroup00-LogVol00
459G 19G 418G 5% /
/dev/sda1 99M 12M 82M 13% /boot
none 252M 0 252M 0% /dev/shm
/dev/shm 150M 236K 150M 1% /opt/zimbra/amavisd-new-2.3.3/tmp
/dev/shm 150M 236K 150M 1% /opt/zimbra/amavisd-new-2.3.3/tmp

[phoenix]

I agree with the - I'm running out of ideas at the moment. I'll keep looking.

[tbullock]

Is ./zmfixperms supposed to be executed as user 'zimbra'?

I can only run it as 'root', perhaps that is part of the problem?

[jholder]

Originally Posted by tbullock

Is ./zmfixperms supposed to be executed as user 'zimbra'?

I can only run it as 'root', perhaps that is part of the problem?

Hi t,
Yup, it has to be run as root.
Also, I would advise a backup, and try to do a reinstall.

If you'd like, I'd be willing to SSH into the box and take a look around. I've done that quite a bit. (not on your box, ofcourse, but on a couple forum members.)

john

[tbullock]

That would be great. I have been nosing around my maillogs, and I have a few questions.

I will open a port on the m firewall and send you a PM on how to connect.

Maybe we could hook up on MSN as well if you have time.

At this point I am suspecting that the backup script running on my backup machine might be hogging the Interface at night, causing the Que to basically crash and pile up with traffic. I stuck around a bit later last night (till 2:00AM), and things started going wrong around 11:20PM.

The errors generated during my attempts to manage the QUEUE via GUI seem to be related to the number of messages in the QUEUE. Once they dropped below 500 or so, I was able to start putting things on HOLD, DELETING, REQUEUEING etc, with generating the SOAP errors.

Mail server is currently working. I babied it through the rough patch last night and it has not gone down since.

I have also cancelled the RSYNC backup to the Zimbra Box. I will have to come up with another solution for backups.

Here is some examples in the log. Basically, Zimbra could no longer communicate with 127.0.0.1.

(TIMINGING OUT WITH LOOPBACK)

Aug 29 23:20:21 webmail postfix/smtp[28000]: 3310F174C460: to=<hflieger@westernavionics.com>, relay=127.0.0.1[127.0.0.1], delay=1206, status=deferred (conversation with 127.0.0.1[127.0.0.1] timed out while sending end of data -- message may be sent more than once)


(QUE ACTIVE)

Aug 29 23:20:32 webmail postfix/qmgr[32052]: 2B75F174C46E: from=<samba-bounces+tbullock=avmax.ca@lists.samba.org>, size=3232, nrcpt=1 (queue active)
Aug 29 23:20:47 webmail postfix/qmgr[32052]: D9868174C46F: from=<werqytr@ATWONLINE.COM>, size=8003, nrcpt=1 (queue active)

(TIMING OUT WITH LOOPBACK)
Aug 29 23:23:06 webmail postfix/smtp[27373]: DB8F5174C45C: to=<alyxgoodnight@fcidms.com>, relay=127.0.0.1[127.0.0.1], delay=1206, status=deferred (conversation with 127.0.0.1[127.0.0.1] timed out while sending end of data -- message may be sent more than once)

Aug 29 23:26:16 webmail postfix/smtp[29220]: 51B0E174C454: to=<pwong@avmax.ca>, relay=127.0.0.1[127.0.0.1], delay=1214, status=deferred (conversation with 127.0.0.1[127.0.0.1] timed out while sending end of data -- message may be sent more than once)

Aug 29 23:26:44 webmail postfix/smtp[29428]: 13A64174C462: to=<dparkin@avmax.ca>, relay=127.0.0.1[127.0.0.1], delay=1213, status=deferred (conversation with 127.0.0.1[127.0.0.1] timed out while sending end of data -- message may be sent more than once)

Aug 29 23:27:04 webmail postfix/smtp[29432]: 8EFE8174C45D: to=<tbullock@avmax.ca>, relay=127.0.0.1[127.0.0.1], delay=1213, status=deferred (conversation with 127.0.0.1[127.0.0.1] timed out while sending end of data -- message may be sent more than once)

Aug 29 23:27:27 webmail postfix/smtp[29535]: 65425174C463: to=<tbullock@avmax.ca>, relay=127.0.0.1[127.0.0.1], delay=1207, status=deferred (conversation with 127.0.0.1[127.0.0.1] timed out while sending end of data -- message may be sent more than once)
Aug 29 23:27:29 webmail postfix/smtpd[5615]: initializing the server-side TLS engine
Aug 29 23:27:48 webmail postfix/smtpd[5610]: 75E51174C472: client=host252-41.pool8259.interbusiness.it[82.59.41.252]

Aug 29 23:27:51 webmail postfix/smtp[29538]: A5AA5174C464: to=<bdave@avmax.ca>, relay=127.0.0.1[127.0.0.1], delay=1209, status=deferred (conversation with 127.0.0.1[127.0.0.1] timed out while sending end of data -- message may be sent more than once)

Thanks for the input so far!

[McPringle]

I had the same problem on my Zimbra 4 Debian Sarge box. Disabling anti virus scanning and spam checking solved the problem - but I don't like a box without anti virus scanning and spam checking... :-(

McPringle

[claros]

Originally Posted by tbullock

Aug 29 23:27:51 webmail postfix/smtp[29538]: A5AA5174C464: to=<bdave@avmax.ca>, relay=127.0.0.1[127.0.0.1], delay=1209, status=deferred (conversation with 127.0.0.1[127.0.0.1] timed out while sending end of data -- message may be sent more than once)

As to me this could be an amavis problem...
Try to disable antivirus and spam check to see if the queue flush.
If the queue flush, try to telnet to the amavis port from localhost:
telnet localhost 10024
and send a test message manually (check amavis doc for an howto)
If you got errors, post it here.
Did you change or updated some perl modules via cpan?

Ciao

[McPringle]

Hello again,

I had Zimbra 4.0.0 running and after checking the SF.net page I saw that 4.0.1 was out. I upgraded to 4.0.1 and enabled anti virus and spam checking again and it seems all fine for a few hours now.

cu
McPringle

[tbullock]

It is related to Anti-virus/SpamAssassin. I am not sure what exactly the problem is, but when I disable he service mail is handled normally.

I am hoping the problem will go away once I upgrade to FC5 and 4.0.2, whenever that happens...lol..

I do seem to remember running a "yum" probably around the same time this happened. I have not changed any passwords.

Anyone have any thoughts?