OTish - Mail stuck in mqueue

Howard Robinson howard at harper-adams.ac.uk
Wed Apr 14 14:55:31 IST 2004


Hello list members

We run MailScanner on a gateway that forwards internal email to a novell
system. Two unusual machine errors have happened, one during the week
when the Mailscanner Machine crashed due to a UPS fault and the second over
weekend the Novell servers bombed.
As soon as I restarted the MailScanner gateway and corrected a few faults on
/var email started flowing. Same again after the weekend when the Novell
servers were restarted, email flowed again.
However there are 4000 emails 'stuck in /var/spool/mqueue.
I have had a look and there are about 40 starting with Q the rest are df or qfs (in
pairs ~4000 each) plus 3 tf files.
Looking through the archives I came across this snipped bit --
---
If the mail is in qf* and df* files, then the next time the queue is run by
sendmail it will all get sent. If they have been renamed to Qf* then you will
need to rename them back again. -- Julian Field
--- end of snipped bit
So they should be on there way as sendmail (via MailScanner) is running.
Mail is going in and out but these 4000 are not moving.
I noticed that whilst all are owned by root some are in the root group and some
in smmsp group (not sure why). Also I had a look at a random selection of the
q files and they all contained connection refused or deferred messages - many
but not all from when the Novell was down.
It's now a good 36 hours since the novell system was restarted.
What can I do to send these on their way & will the deferred ones eventually go?





Regards

Howard Robinson
(Senior Technical Development Officer)
Harper Adams University College
Edgmond
Newport
Shropshire
TF10 8NB UK

E-mail: hrobinson at harper-adams.ac.uk
Tel.  : +44(0)1952 820280 Via switchboard
      : +44(0)1952 815253 Direct line
Fax.  : +44(0)1952 814783
College Web site http://www.harper-adams.ac.uk



More information about the MailScanner mailing list