Exim receiving problem - error 550

Mr. Brightside · 15 Oct 2006 at 16:58

I've finally got around to setting up a mail server on my Debian Etch box today, using this as my guide: http://koivi.com/exim4-config/

I can connect through Outlook via IMAP fine to check for new mail, so Courier is working fine. However I've got one "minor" problem in that Exim isn't allowing any incoming mail! I sent a test message from my G-Mail account and got this delivery failure message:

PERM_FAILURE: SMTP Error (state 9): 550 relay not permitted

Any ideas how I can fix this?

Thanks, null

Garp · 15 Oct 2006 at 18:27

Troubleshooting time.. using telnet.
Its oh so fun.

telnet ip-of-server 25
HELO
mail from:<your g-mail address>
rcpt to:<your destination address>

Note, the '<' and '>' are deliberate and necessary

Does it come back with the 550 error?

Type the command QUIT to leave the SMTP session at any stage.

Mr. Brightside · 15 Oct 2006 at 19:08

Nope, 250 accepted. I think I might have unknowingly fixed the original problem at some point earlier before starting this thread, but mail is still not getting through; presumably it's accepted by Exim but then getting lost somewhere in the system.

220 server.mydomain.tld ESMTP Exim 4.63 Sun, 15 Oct 2006 18:47:12 +0100
mail from:<[email protected]>
250 OK
rcpt to:<[email protected]>
250 Accepted

I also tried this on the server:

exim4 -v [email protected]
From: [email protected]
To: [email protected]
Subject: Testing exim

Testing exim
.

And the above message appeared in my inbox, checked from Outlook via IMAP. Not sure what use that is but thought I'd mention it.

Thanks for the help, null

Garp · 15 Oct 2006 at 20:32

What do your logs say? (should be /var/log/exim or /var/log/exim4 by default)

Mr. Brightside · 15 Oct 2006 at 20:58

2006-10-15 20:54:59 1GZC4Z-0005Ox-4L malware acl condition: clamd: unable to connect to UNIX socket /var/run/clamav/clamd.ctl (Connection refused)
2006-10-15 20:54:59 1GZC4Z-0005Ox-4L H=nf-out-0910.google.com [64.233.182.191] F=<[email protected]> temporarily rejected after DATA
2006-10-15 20:55:11 1GZC4l-0005T4-Mv malware acl condition: clamd: unable to connect to UNIX socket /var/run/clamav/clamd.ctl (Connection refused)
2006-10-15 20:55:11 1GZC4l-0005T4-Mv H=nf-out-0910.google.com [64.233.182.184] F=<[email protected]> temporarily rejected after DATA

Thanks again, null

Garp · 15 Oct 2006 at 22:01

null said:
Thanks again, null

Ahh.. you've not got Clam Anti-virus running which its looking for. Do you actually have it installed?

Mr. Brightside · 15 Oct 2006 at 22:28

Yes it's installed, and all set up as described by the guide, however, it seems it doesn't want to run properly - calling /etc/init.d/clamav-daemon restart returns this:

Stopping ClamAV daemon: clamd/etc/init.d/clamav-daemon: line 75: kill: (-29261) - No such process
failed!
Starting ClamAV daemon: clamd.

If I then try restarting it again, immediately after, it returns the same.

Thanks for the help, null

Garp · 15 Oct 2006 at 22:46

null said:
Yes it's installed, and all set up as described by the guide, however, it seems it doesn't want to run properly - calling /etc/init.d/clamav-daemon restart returns this:

If I then try restarting it again, immediately after, it returns the same.

Thanks for the help, null

if you run the "ps -aux | less" command, do you see any hint of clamav running?

Clam should be creating logs somewhere that would hint at why its not initialising succesfully. Check inside /var/log again, if you can't see anything specific, try doing the restart and then running

tail -20 /var/log/messages

and seeing if there is anything of interest there.

Might be worth trying to trigger clam manually instead of using the service. If you don't know the path to the executable, you could do
"less /etc/init.d/clamav-daemon" and it should be clear where it is in there.

Mr. Brightside · 15 Oct 2006 at 22:52

Got this which looks like something to do with Clam:

clamav 29100 0.0 0.5 5420 1356 ? Ss 14:08 0:00 /usr/bin/freshclam -p /var/run/clamav/freshclam.pid -d --quiet

That right or not?

Thanks, null

edit: Looked at /var/log/clamav/clamav.log and these are the last few lines:

Sun Oct 15 22:27:55 2006 -> +++ Started at Sun Oct 15 22:27:55 2006
Sun Oct 15 22:27:55 2006 -> clamd daemon 0.88.4 (OS: linux-gnu, ARCH: i386, CPU: i486)
Sun Oct 15 22:27:55 2006 -> Log file size limit disabled.
Sun Oct 15 22:27:55 2006 -> Running as user clamav (UID 106, GID 107)
Sun Oct 15 22:27:55 2006 -> Reading databases from /var/lib/clamav
Sun Oct 15 22:27:58 2006 -> Protecting against 73019 viruses.
Sun Oct 15 22:27:58 2006 -> WARNING: Socket file /var/run/clamav/clamd.ctl exists. Unclean shutdown? Removing...
Sun Oct 15 22:27:58 2006 -> ERROR: Socket file /var/run/clamav/clamd.ctl could not be removed: Permission denied
Sun Oct 15 22:28:24 2006 -> +++ Started at Sun Oct 15 22:28:24 2006
Sun Oct 15 22:28:24 2006 -> clamd daemon 0.88.4 (OS: linux-gnu, ARCH: i386, CPU: i486)
Sun Oct 15 22:28:24 2006 -> Log file size limit disabled.
Sun Oct 15 22:28:24 2006 -> Running as user clamav (UID 106, GID 107)
Sun Oct 15 22:28:24 2006 -> Reading databases from /var/lib/clamav
Sun Oct 15 22:28:26 2006 -> Protecting against 73019 viruses.
Sun Oct 15 22:28:26 2006 -> WARNING: Socket file /var/run/clamav/clamd.ctl exists. Unclean shutdown? Removing...
Sun Oct 15 22:28:26 2006 -> ERROR: Socket file /var/run/clamav/clamd.ctl could not be removed: Permission denied

Garp · 15 Oct 2006 at 22:56

null said:
Got this which looks like something to do with Clam:
That right or not?

Not used clam myself yet, will be looking at it at some stage this week, but I'd reckon so. I'm a little baffled where the daemon service is getting its PID for when its killing clam. It must be digging that up from somewhere!

edit: NVM

Sun Oct 15 22:28:26 2006 -> WARNING: Socket file /var/run/clamav/clamd.ctl exists. Unclean shutdown? Removing...
Sun Oct 15 22:28:26 2006 -> ERROR: Socket file /var/run/clamav/clamd.ctl could not be removed: Permission denied

That does suggest that clam has been triggered by something and has left a lock file, presumably that ties in with what you saw from the ps -aux command.

Mr. Brightside · 15 Oct 2006 at 22:57

Garp said:
Anything in your logs?

See edit

Garp · 16 Oct 2006 at 07:55

null said:
See edit

Spotted it

http://www.exim.org/mail-archives/exim-users/Week-of-Mon-20040809/msg00113.html

Looks interesting, maybe give that a shot?
Your error message suggests it keeps hitting up against this lockfile business, and I'm not sure why. Exim shouldn't / wouldn't have started a new clamav session each time I wouldn't have thought, there would be no need to do so.

It may be that you'll need to use "kill -8" or if that doesn't work "kill -9" on the clam session and manually remove that file.

Mr. Brightside · 21 Oct 2006 at 11:28

I tried that but it didn't work so I just nuked the entire setup (i.e. removed config files aswell as removing the packages) and started again from scratch; it seems to be working now - the only other change is that my MX records now point to venus.piggott.me.uk instead of piggott.me.uk.

Thanks for all the help Garp

null