Weird Exim issue .. stuck queue but can deliver manually?

[rpmws]

Weird Exim issue .. stuck queue but can deliver manually? Here is what the headers look like if I view the message in the WHM mail queue. Note that these are a list of about 500 emails. They ramp up my queue to about 600 and it stays there until they are automatically deleted or deleted by me. I can hit "deliver message now" and it goes out no problem in like 1/2 second. Almost all of them listed in there that are in queue will deliver, they just won't go out automatically. Header below...thanks in advance.

1BzNK4-00044G-T7-H
root 0 0
<[email protected]>
1093300192 0
-helo_name hisdomain.com
-host_address 200.78.83.122.62151
-host_name dsl-200-78-83-122.prod-infinitum.com.mx
-host_auth fixed_login
-interface_address my.server.ip.is.here.25
-received_protocol asmtp
-body_linecount 202
-auth_id [email protected]
-deliver_firsttime
XX
1
[email protected]

214P Received: from dsl-200-78-83-122.prod-infinitum.com.mx ([200.78.83.122] helo=hisdomain.com)
by my.fullservername.com with asmtp (Exim 4.41)
id 1BzNK4-00044G-T7
for [email protected]; Mon, 23 Aug 2004 17:29:54 -0500
042 Mail-System-Version: [email protected] HTML 4.5.1
024 X-Sender: [email protected]
038 Date: Mon, 23 Aug 2004 17:23:14 -0500
098F From: "Deanna and Chris Clark - Life Giving Outreach Ministries, Reynosa, Mexico" <[email protected]>
053T To: "Alice and Nicholas Bentin" <[email protected]>
067 Subject: God at work through the hearts and hands of his faithful.
024R Reply-To: [email protected]
018 MIME-Version: 1.0
024 Content-Type: text/html

Can somone tell why these just won't go out? ..oh ..i checked the loads and they are low.

 

[chirpy]

That is odd. If you run the following, exactly do you get?

ps axf | grep -v grep | grep exim

 

[rpmws]

That is odd. If you run the following, exactly do you get?

ps axf | grep -v grep | grep exim

been running edge builds for 2 years, this has been an issue since day one, about a year ago for this user. All other mail delivers fine. Also upgraded to latest exim and same deal

23173 ? S 0:01 /usr/bin/perl /usr/local/cpanel/bin/eximstats
25664 ? S 0:01 /usr/sbin/exim -bd -q20m
1865 ? S 0:00 \_ /usr/sbin/exim -q
31239 ? S 0:00 | \_ /usr/sbin/exim -q
31242 ? S 0:00 | \_ /usr/sbin/exim -q
9300 ? S 0:00 \_ /usr/sbin/exim -q
32330 ? S 0:00 | \_ /usr/sbin/exim -q
32331 ? S 0:00 | \_ /usr/sbin/exim -q
15730 ? S 0:00 \_ /usr/sbin/exim -q
349 ? S 0:00 | \_ /usr/sbin/exim -q
350 ? S 0:00 | \_ /usr/sbin/exim -q
31232 ? S 0:00 \_ /usr/sbin/exim -bd -q20m
32339 ? S 0:00 \_ /usr/sbin/exim -bd -q20m
395 ? Z 0:00 | \_ [exim <defunct>]
32559 ? S 0:00 \_ /usr/sbin/exim -bd -q20m
416 ? S 0:00 \_ /usr/sbin/exim -bd -q20m
432 ? S 0:00 \_ /usr/sbin/exim -bd -q20m
433 ? S 0:00 \_ /usr/sbin/exim -bd -q20m
473 ? S 0:00 \_ /usr/sbin/exim -bd -q20m
474 ? Z 0:00 \_ [exim <defunct>]
25671 ? S 0:00 /usr/sbin/exim -bd -oX 26
25677 ? S 0:00 /usr/sbin/exim -tls-on-connect -bd -oX 465
31386 ? S 0:00 /usr/sbin/exim -Mc 1BzcNH-00088f-Ag
31394 ? S 0:00 \_ /usr/sbin/exim -Mc 1BzcNH-00088f-Ag
32595 ? S 0:00 /usr/sbin/exim -Mc 1BzcQi-0008Pd-13
32601 ? S 0:00 \_ /usr/sbin/exim -Mc 1BzcQi-0008Pd-13
32609 ? S 0:00 /usr/sbin/exim -Mc 1BzcQm-0008Tq-Ta
32616 ? S 0:00 \_ /usr/sbin/exim -Mc 1BzcQm-0008Tq-Ta

 

[chirpy]

Well, that looks OK to me. That seems to show that you're running your queue process every 20 minutes (the -q20m ones) and it's trying to dequeue emails (the -q ones).

The best thing I can suggest would be to (on a Linux server) setup a monitor window with:

tail -f /var/log/exim_mainlog

In another window, stop and restart exim:

/etc/init.d/exim restart

Exim should run a queue on startup and you ought to see any errors in the other window as to why it is not delivering.

A couple more things:

1. What do you have at the very end of your /etc/exim.conf?

It ought to look something like:

*                      *           F,2h,15m; G,16h,1h,1.5; F,4d,8h

2. Have you had MailScanner installed in the past and removed it? If so, what do the first 10 lines or so at the top of your /etc/exim.conf look like?

 

[rpmws]

Well, that looks OK to me. That seems to show that you're running your queue process every 20 minutes (the -q20m ones) and it's trying to dequeue emails (the -q ones).

The best thing I can suggest would be to (on a Linux server) setup a monitor window with:

tail -f /var/log/exim_mainlog

In another window, stop and restart exim:

/etc/init.d/exim restart

Exim should run a queue on startup and you ought to see any errors in the other window as to why it is not delivering.

A couple more things:

1. What do you have at the very end of your /etc/exim.conf?

It ought to look something like:

*                      *           F,2h,15m; G,16h,1h,1.5; F,4d,8h

2. Have you had MailScanner installed in the past and removed it? If so, what do the first 10 lines or so at the top of your /etc/exim.conf look like?

I run tail on my exim_mainlog all teh time on a machine I use for MRTGs and logs. When I restart exim it's like it skips over these emails ...but YES I did have mailscanner installed and infact when I restart exim I see errors. I removed mailscanner on this bax some time ago becuase of the loads it was causing... what should I look for a change to make sure I don't have any old mailscanner inteferace ?

I am showing this at end of exim.conf

*                      *           F,2h,15m; G,16h,1h,1.5; F,4d,8h

 

[rpmws]

I run tail on my exim_mainlog all teh time on a machine I use for MRTGs and logs. When I restart exim it's like it skips over these emails ...but YES I did have mailscanner installed and infact when I restart exim I see errors. I removed mailscanner on this bax some time ago becuase of the loads it was causing... what should I look for a change to make sure I don't have any old mailscanner inteferace ?

I am showing this at end of exim.conf

*                      *           F,2h,15m; G,16h,1h,1.5; F,4d,8h

also when I restart exim I get this.

Shutting down clamd: [FAILED]
Shutting down exim: [ OK ]
Shutting down antirelayd: [ OK ]
Shutting down spamd: [ OK ]
Starting clamd: ERROR: Can't open config file /etc/clamav.conf !
ERROR: Can't parse the config file /etc/clamav.conf
[FAILED]
Starting exim: [ OK ]
Starting exim-26: [ OK ]
Starting exim-smtps: [ OK ]
Starting antirelayd: [ OK ]
Starting spamd: [ OK ]


Don't have mailscanner running.

 

[chirpy]

If you're not using clamav you can clear up those errors with:

rm -Rfv /usr/bin/clam*
rm -Rfv /usr/sbin/clam*
rm -Rfv /usr/lib/libclam*
rm -Rfv /usr/share/clam*
rm -Rfv /usr/include/clam*
rm -Rfv /usr/bin/freshclam*
rm -Rfv /usr/etc/clamav*

I suspect you've got some MailScanner stuff hanging around. I would suggest doing this first:

/scripts/buildeximconf

Then, check the top of exim.conf and make sure that you do not have any of the following lines at the top:

spool_directory = /var/spool/exim_incoming
queue_only = true
queue_only_override = false

Then restart exim:

/etc/init.d/exim restart

 

[rpmws]

If you're not using clamav you can clear up those errors with:

rm -Rfv /usr/bin/clam*
rm -Rfv /usr/sbin/clam*
rm -Rfv /usr/lib/libclam*
rm -Rfv /usr/share/clam*
rm -Rfv /usr/include/clam*
rm -Rfv /usr/bin/freshclam*
rm -Rfv /usr/etc/clamav*

I suspect you've got some MailScanner stuff hanging around. I would suggest doing this first:

/scripts/buildeximconf

Then, check the top of exim.conf and make sure that you do not have any of the following lines at the top:

spool_directory = /var/spool/exim_incoming
queue_only = true
queue_only_override = false

Then restart exim:

/etc/init.d/exim restart

Thanks so much for all your help chirpy. That got rid of all the clam stuff and I **think** I just saw a hand full of those queued email leave!!!!

 

[chirpy]

That's great It may take a while for them to all become unforzen, but if you go into WHM > Manage Mail Queue > and click Attempt to Deliver all messages in Queue you gave give it a hurry up.

 

[jols]

I know this is an old thread, but Chirpy, we've got MailScanner running on all of our servers, and also have very heavy problems with excessive queues, i.e. usually in the range of 6k to 22k messages in queue. Some have been there as long as 110 days.

I have added this at the top of the exim.conf file:

auto_thaw = 6d
ignore_bounce_errors_after = 7d
timeout_frozen_after = 8d


But I also have this at the top of the file as well:

spool_directory = /var/spool/exim_incoming
queue_only = true

Should I comment out both of these lines (above)? Would that mess up MailScanner if I did?

Thanks much.

P.S. Also, is there any automatic way of purging these thousands of messages that are older than 8 days?

 

[chirpy]

You should never edit exim.conf directly, simply use the Exim Configuration Editor. If you remove those lines from exim.conf you will break MailScanner and email delivery.

If you're getting a large queue, why are you keeping those emails in the queue for such a long time. I'd suggest not using auto_thaw and changing the settings to:

ignore_bounce_errors_after = 1d
timeout_frozen_after = 4d

There's no reason to keep emails in the queue after 4 days as that's as long as is normally expected.

 

[jols]

Okay, did that using the WHM exim config editor, HOWEVER, I see that further down the page, we have the following which is in an un-editable area:

# primary_hostname =
deliver_queue_load_max = 3
auto_thaw = 6d
ignore_bounce_errors_after = 7d
timeout_frozen_after = 8d

I am puzzled as to why these settings had no effect if they were already on the exim.conf file.

ALSO, I would love to change the deliver_queue_load_max setting to 10 or 14 because this particular server is under a continuual high load, and I think the queing sessions are not running because of this.

But apparently, the only way to change the deliver_queue_load_max setting is by editing the exim.conf directly? Which you say I could never do?

 

[jols]

I am still puzzeled why the other settings had no effect (i.e. if email was supposed to be deleted after 8d why we were seeing email still in queue after 110d.

But please know that I found the answer to the other quesitons in this post - deliver_queue_load_max = 6

Thanks much for your help and for any response to the remaining quesiton.