Why doesn't Exim ever clean out /var/spool/exim/input?
Here's a question that has always been bugging me.
Ever since we moved from smail to exim many years ago at my isp, exim
never seems to discard messages in the input queue.
Even though the single retry rule is the stock one (which retrys for
something like 4 days), we end up with stuff that is weeks... months
old. Periodically, it would get pretty full and we'd notice that there
were about 10 queue runners going and so I'd go in and do a "find" and
remove anything older than 14 days or so. I *had* to do a find, because
doing an "ls" would just sit there an churn for about a half-hour.
Anyhow, as our customer base has grown and as their e-mail usage has
grown, the problem has reached an all-time high. With this SCO DDoS
virus going around, I had occasion to go clean out the input queue
again.....
The directory was using 17 megs....
I'm not talking about the FILES in the directory... I'm talking about
the directory ENTRIES (filename, inode number, etc.). I was forced to
just say "screw it!" and I mv'd the input and msglog folders to other
names and then created new, empty ones so that our mail server wouldn't
buckle under the load.
But anyway, like the subject line says, my real question is: why doesn't
Exim ever clean this stuff out itself?
- Joe
Reply to: