Incoming and outgoing stopped - manually deleted messages from Queue

[prservices]

Hi,

I am running plesk 7.5.2 on a linux.
Am having trouble with spammer relaying messages through Qmail. My server provider allows up to 3000 per day and there were over 5000 in queue. So I manually deleted them using the rm commang in SSH. After doing this now the outgoing and incoming emails have stopped working.

Any help appreciated!

 

[BIS]

honestly - my best advice is to contact plesk as this is not a vanilla qmail install. You probably have a messed up / corrupt queue though.

 

[thedaver]

If you manually deleted files from your queue, you HAVE broken the queue enough to confuse qmail, irrespective of plesk.

You need to google for "qmail queue repair".

D.E.R. Management - IT Project Management Consulting

http://www.dermanagement.com/

 

[prservices]

Thanks for the help, will look into queue repair.. in the meantime when type /etc/init.d/courier-imap status

it return this message:
courier-imap dead but subsys locked

Is this related to my manually screwing up the queue?

Thanks!

 

[prservices]

I found a qmail queue program at:

http://pyropus.ca/software/queue-repair/

Installed it following the instructions.

Then:

Stop Qmail:
/sbin/service qmail stop

Run program to repair, create, split and specify big todo:
/usr/local/lib/queue-repair/queue_repair.py -r -c -s23 -b

Start Qmail:

/sbin/service qmail start

Sending mail works again, but receiving email still doesn't work.. thanks very much for the help!

 

[thedaver]

Your courier-related authentication is messed up... probably a hung process.

1) Issue '/etc/init.d/courier-imap stop' at least twice and then try to restart it...

2) You should also have an accompanying 'courier-authlib' script ... '/etc/init.d/courier-authlib'... try to stop that twice and then restart it.

your authentication should work...

The reference you noted in your "subsys locked" is because there's either 1) a zombie daemon still running from the courier family, or 2) the daemons are shutdown correctly, but one of them left their *.lock or *.pid file somewhere...

The location of the .lock or .pid file is available in the scripts above. However, it's been my experience that asking the scripts to clean up (i.e. 'stop' ) at least twice will normally clear up problems with daemons, locks, pids, etc.

Report back, interested to hear how it goes.

D.E.R. Management - IT Project Management Consulting

http://www.dermanagement.com/