[Mailman-Users] Problem subscribing to lists via web]

[Mark T. Valites]

Jon, thanks for such a quick response,

The mailman dir is located on a sun T3 storage array which is directly 
attached to the machine running mailman in this case, a sunfire 280R 
running SunOS 5.8  Mailman's home dir is located in a directory which is 
indeed nfs exported to several machines, however none of them use this 
mailman dir...  The quota for user mailman is no where near being 
filled, and the slice it resides on has plenty of free room.

What confuses me further is that dumpdb appears to to get all the info 
for the list.  I was hoping not to need to ask, but should I need to (oh 
and I feel the pain already - 626 lists is going to hurt), what are 
recommendations for rebuilding these lists?  There are backups, but I'm 
not sure how much that is going to help - I work at an academic 
institution, where activity from mailman pretty much ceases during the 
summer, so I'm not even sure when this broke.  What magic can I concoct 
from the existing mailman utils and information from dumpdb?  I've also 
read that there are differences between mm 2.0 (run on this machine) and 
2.06, which I may be interested in upgrading to if I need go there.

Jon Carnes wrote:

>>Still no luck with the problems below.  I'm not sure what each of the
>>functions listed in the cron output and the error logs do, but I began
>>to think that my config.db's all got fubared somehow.  I have since
>>added additional lists, thereby creating new databases, however, I still
>>get the same errors.  I'm beginning to doubt the dbs are messed up.  I'm
>>not certain it's a specific script that's puking either, since several
>>are listed in the errors below.  Perhaps they all have a common built-in
>>function that's croaking?  I'm learning python  on the fly - am I
>>missing something obvious?  Can any one give me more info about the
>>functions in the error codes, or point me down a (narrower) road to
>>investigate?  I have 626 lists that no one can subscribe themselves to
>>right now.
>>
>Sounds like a hard-drive or an NFS mounting problem.  Are the directories
>mounted via NFS?  Is there currently a problem with NFS on the machines?
>Has one of the partitions filled up with data so there is no more room on
>it?
>
>If any of the above is true you would get those errors and then your config
>files really could easily become corrupt as users attempt to add and delete
>themselves to your lists.
>
>Jon Carnes
>
>
>------------------------------------------------------
>Mailman-Users maillist  -  Mailman-Users at python.org
>http://mail.python.org/mailman/listinfo/mailman-users
>