Fwd: Kolab not excepting new mail.

[John McMonagle]

On Monday 06 June 2011 10:08:00 am Skip Morse wrote:
> 2011/6/5 John McMonagle <mcmonagle101 at gmail.com>:
> > To give a brief  history
> > Running kolab 2.2.4. and it has been fine for a long time.
> > It's running under xen. Using debian lenny.
> > Xen is running on raid1 drives..
> > A day ago the power company did some work. Power was out a while but
> > suspect there may been more happening :-(
> > The only issue remaining that I know of it this issue.
> >
> > I had a lot to fix but will stick to related parts.
> > Air conditioner died and room was 88 F. Fixed...
> > One of the disks died. Replaced it.
> > Did a fsck on all the partitions and saw no problems.
> > There is a ups but suspect it did not shutdown properly.
> >
> > I get errors like this in postfix log for each incoming mail:
> > Jun 05 19:53:10 mail.advocap.org <info> postfix/error[20404]:
> > 6D19521A6AB: to=<mikeb at advocap.org>, relay=none, delay=16906,
> > delays=16906/0.36/0/0.23, dsn=4.4.1, status=deferred (delivery
> > temporarily suspended: connect to 127.0.0.1[127.0.0.1]: Connection
> > refused)
> >
> > My top suspect is that amavisd is not working.
> >
> > When I look in rc.amavisd  it looks like there should be a processes
> > amavisd and amavis-milter but I see neither one.
> > There  are not new entries in /kolab/var/amavisd  since it died.
> >
> > starting amavisd gives no errors.
> >
> > It's partitioned like this:
> > df -h
> > Filesystem            Size  Used Avail Use% Mounted on
> > /dev/xvda2            4.0G  1.3G  2.5G  35% /
> > tmpfs                 1.3G     0  1.3G   0% /lib/init/rw
> > udev                   10M  460K  9.6M   5% /dev
> > tmpfs                 1.3G  4.0K  1.3G   1% /dev/shm
> > /dev/xvda3            508M  4.7M  503M   1% /tmp
> > /dev/xvda4            2.0G  858M  1.2G  43% /kolab
> > /dev/xvda5            4.0G  2.1G  2.0G  52% /kolab/var
> > /dev/xvda6             50G   29G   22G  57% /kolab/var/imapd/spool
> >
> > I do have backups.
> > I'm thinking of doing a fresh install on a vm to see how it's supposed to
> > run.
> >
> > Any suggestions?
> >
> > John
>
> what does netstat tell you?  I've seen a similar error when I've
> broken the smtp chain by messing with my master.cf config file... The
> only input I have is just to verify that you have smtpd listening on
> ports 10025 and 10026, that was my issue.  But your log message
> doesn't really mention a port which I think is weird.
>
> Thanks,
> -Skip

Figured it out a couple hours after posting question.
Now on my normal work email address.

I did a new test installation.
Was missing amavisd processes.
In /kolab/var/amavisd in the test installation I noticed that amavisd.pid and 
amavisd.sock went away when  amavisd was stopped.
 
On the production server  had  amavisd.pid and amavisd.sock.
I removed them and restarted and all is well.

From the list I suspect others have had this problem.
Maybe the rc scripts could remove them.
It would take a  test to make the a process of the pid did not really exist. 
Or as an alternative blow them away on  boot startup.

John