[Mailman-Users] database error message when upgraded to 2.1.2

NOW Website Coordinator webperson at now.org
Fri May 2 00:15:12 CEST 2003


At 09:59 PM 04/29/2003 -0400, Jon Carnes wrote:
>Hmmm, we might be able to find out quickly exactly where the problem is
>if you run qrunner manually and use strace.
>
>First lets run it without using strace and see if that fails as
>expected:
>   su mailman
>   cd ~mailman
>   bin/qrunner -ovr All
>
>Now lets do it with strace:
>   strace bin/qrunner -ovr All



Unfortunately, I didn't get errors with either.  Although I did get the 
error message:

Warning: Database (2.5.12) != program (2.6.2) version.

again today.

For some reason, both on our old Solaris site and now on our new Linux 
site, the Outgoing Runner takes up 70- 90% of our CPU.

On our old Solaris site (where mail and the web sites are now shut down), 
mailman is still running and it's *still* taking up over 60% of the CPU 
with the outgoing queue (although of course it's doing nothing at all, it 
can't send mail and no one can send it new mail to process.):

20332 ?        R 662:43 qrunner /usr/local/mailman/bin/qrunner 
--runner=OutgoingRunner:0:1 -s

It's using version 2.1.1 on our old server, 2.1.2 on our new one.

Does Mailman just use up as much CPU as it can, even when it has no tasks 
to fulfill?  Maybe that's okay, in Unix, for not to be idle, but I don't 
like the way all of the other jobs are pushed down on the CPU list.  But 
even so, the load average on the Solaris -- which is doing nothing -- is 
high: .88 .  And it's all mailman when you do a "top."

One of the big reasons we switched servers was because Mailman had bogged 
down our machine so much, using up all of the memory and CPU.  I thought it 
we went to Linux and thus a faster server it wouldn't be such a problem.

Anyone know what's going on?

Thanks.






More information about the Mailman-Users mailing list