QUE Failure

[tbullock]

My mail que has crashed and will not deliver messages. I am wondering if it has anything to do with the unclean shutdown changing directory ownerships to root:root

What directories should be owned by zimbra:zimbra?

Can I execute a chown -R zimbra:zimbra on the /opt directory?

I will also attach a blurb from the /var/log/zimbra.log file that may shed some light.

[phoenix]

Try using zmfixperms in /opt/zimbra/libexec

[tbullock]

Ok I tried that, but things are not looking any better. I still have over 400 messages in my "Active" que that don't seem to be going anywhere. Obviously this is making things very slow. The following are some oddities I see in the /var/log/zimbra.log file:

Aug 25 12:43:59 webmail postfix/qmgr[18226]: warning: qmgr_active_corrupt: save corrupt file queue active id BFFC51748736: No such file or directory

Aug 25 12:43:53 webmail amavis[15381]: (15381-01-8) SA TIMED OUT, backtrace: at /opt/zimbra/zimbramon/lib/Mail/SpamAssassin/BayesStore/DBM.pm line 898\n\teval {...} called at /opt/zimbra/zimbramon/lib/Mail/SpamAssassin/BayesStore/DBM.pm line 898\n\tMail::SpamAssassin::BayesStore:BM::get_st orage_variables('Mail::SpamAssassin::BayesStore: BM=HASH(0xa157960)') called at /opt/zimbra/zimbramon/lib/Mail/SpamAssassin/BayesStore/DBM.pm line 288\n\tMail::SpamAssassin::BayesStore:BM::tie_db _writable('Mail::SpamAssassin::BayesStore:BM=HAS H(0xa157960)') called at /opt/zimbra/zimbramon/lib/Mail/SpamAssassin/Bayes.pm line 712\n\teval {...} called at /opt/zimbra/zimbramon/lib/Mail/SpamAssassin/Bayes.pm line 702\n\tMail::SpamAssassin::Bayes::learn('Mail::Spa mAssassin::Bayes=HASH(0xa0baa04)', 0, 'Mail::SpamAssassin::Message=HASH(0xa75b80c)', 'undef') called at /opt/zimbra/zimbramon/lib/Mail/SpamAssassin/PerMsgLearner.pm line 130\n\tMail::SpamAssassin::PerMs...


And a lot of these 'que actives':

Aug 25 12:43:49 webmail postfix/qmgr[18226]: 4A0DC17486E5: from=<bairobinj@0451.com>, size=16851, nrcpt=1 (queue active)

[phoenix]

Did you reboot after applying the zmfixperms? Iif not could you do that and see what happens.

[tbullock]

Thanks Bill,

I have rebooted the machine. The QUE currently sits at 120 Active messages so it is slowly winning the battle. I am still concerned though, obviously as it seems strange that it would take so long.

When I arrived this morning, 5 hours ago, there were over 600 messages in the QUE, so it is handling approximately 120 messages per hour...so 2 messages per minute out of the Active QUE, not counting the messages generated both incoming and outgoing during that period as well. Rough estimate.

[phoenix]

I must admit to not knowing much about the queue manager. I'm assuming that this is a reasonable spec machine and there's no more errors in the log? The only other ting I can think of is the number of threads that handle mail processing might be too low but I don't know how to change it (if it's possible).

[rsharpe]

You may want to try flushing the queue to speed up delievery of the messages.

[tbullock]

I wonder if it is someone who is sending attachments that are too large that is tying things up?

Is there a way to look at message sizes that are being process by the que and who is sending them?

[rsharpe]

You can use postfix's command to view the queue it shows the message size:

postqueue -p prints the queue

Run the command as the zimbra user.

[tbullock]

OK, once again I have arrived to work with over 600 messages in "Defered".

I was in last night around 9:00 or 9:30 and the Que was empty and processing messages. Now 10 hours later, there are over 600 deferred???

What is going on?

Here is a snipet of the postqueue -p command:

[zimbra@webmail root]$ postqueue -p
-Queue ID- --Size-- ----Arrival Time---- -Sender/Recipient-------
00C151748655 5610 Tue Aug 29 01:34:51 2386411.1132009@totallywire.com
(delivery temporarily suspended: connect to 127.0.0.1[127.0.0.1]: read timeout)
mesperidion@avmax.ca

0BD0017487F8 25268 Tue Aug 29 07:23:09 dinod@aviation-industry.com
(delivery temporarily suspended: connect to 127.0.0.1[127.0.0.1]: read timeout)
vscott@westernavionics.com

0E0A31748717 3487 Tue Aug 29 05:03:45 bodneryeimhi@filho.net
(delivery temporarily suspended: connect to 127.0.0.1[127.0.0.1]: read timeout)
psmart@avmax.ca

06A1F17487AF 1751 Tue Aug 29 06:54:10 techschop@polybuchhandlung.ch
(delivery temporarily suspended: connect to 127.0.0.1[127.0.0.1]: read timeout)
techresumes@avmax.ca

0048A17486F1 3449 Tue Aug 29 04:32:54 carvere@hardrockers.com
(delivery temporarily suspended: connect to 127.0.0.1[127.0.0.1]: read timeout)
jhorvat@avmax.ca

053C117486D9 1154 Tue Aug 29 04:02:52 daisy@bollywood2000.com
(delivery temporarily suspended: connect to 127.0.0.1[127.0.0.1]: read timeout)
ggraham@avmax.ca
rbinder@avmax.ca


Same message for all 600+