Larry Rosenman writes: On Monday 18 April 2005 02:10 am, Jim Tittsler wrote:
On Apr 15, 2005, at 12:22, Larry Rosenman wrote:
I have 178MEG of bounce-events-* files in my /data directory.
> > >
Should these files be hanging around?
No. It sounds like either:
- your bounce queue runner isn't running
It had been crashing, due to the Python 2.4.* issues. Now fixed.
My bounce queue runner dies so often that I had to change MAX_RESTARTS
from 10 to 100. I also have a script that restarts mailman every day.
I've had the bounce runner die as often as 41 times in one day.
Was the issue causing the bounce runner to crash fixed in a newer
version of python? If so, what version was the fix introduced in?
Are the extra files a result of the bounce runner crashing, or a
result of me restarting Mailman every day?
I'd be inclined to just delete the old bounces (since some of them may
be stale by now) and restart Mailman.
What determines if they are stale? I don't understand the details of
the way the bounce events are handled. I think that every 15 minutes
(by default) Mailman processes the pending bounce events, but what
determines what events are pending? Does mailman keep a list of those
somewhere? Or does Mailman just look for files named