[Mailman-Users] List has been running for a year and suddenly last night it stopped.

[Angus Fox]

On Monday, November 15, 2004, at 08:52AM, Brad Knowles <brad at stop.mail-abuse.org> wrote:

>At 11:24 PM -0800 2004-11-14, Noah wrote:
>
>>  possibly you have soft-updates in your /etc/fstab on your /var partition.
>>  therefore your filesystem is claiming space for unwriten log files, etc.  so
>>  you will not see them physically on the drive.
>
>	You could have problems with log files that have been deleted but 
>which are held open by some process (and therefore the disk space has 
>not yet been recovered), regardless of what type of filesystem is 
>being used.
>
>	Moreover, the filesystem could have run out of inodes but have 
>plenty of disk space available -- a "df -i" should show whether or 
>not that was the case, looking in the "ifree" and "%iused" columns.
>
>
>	Excepting these two nits, the advice is good.
FIXED!

My friend from Red Hat had a look and fixed this for me (thank you Mark). It turned out to be some kind of shunt error he spotted in mailman 2.12 logs. He upgraded through the versions as per instructions in the various readmes. Python, perl versions etc.. all remain the same as before. 

specific upgrade options were: 
./configure --with-mail-gid=mail --with-cgi-gid=httpd 
make install 

During upgrade it croaked on updating the qfiles and I didn't save the 
error. They are still there in 

/usr/local/mailman/qfiles/shunt 

and I believe it to be a problem with file format changes and dir 
structure changes between releases. 

Now all is well and its running perfectly again on mailman 2.15.

Thanks for your pointers, all were checked out but in the end moving through the upgrades to 2.15 seems to have solved the problem.

Angus