The four processes were old and definitely not doing anything but chewing up
processor. Wish
I had saved some output. I'm sure that they had been sitting around longer than the
hard-timeout value allowed for a user to be logged-in.
I only have a few of my users testing sqwebmail, at the moment. At the time of night
I
noticed the processes, it was statistically impossible for the server to overloaded
from
sqwebmail users ;) And, I hope that a few users can't kill a PIII 733 with 512 MB RAM
and 18
GB 10,000RPM mirrored hard disks...
<rant>
BTW, 10,000 messages? Man, you need to start using maildrop to filter some of those
messages
into other directories ;) I thought my users were bad. I haven't enforced any
quotas, yet.
Some of my users have over 100 Mb of email. I'm on a ton of mailing lists, get lots of
attachments from friends, etc., and I only have about 13Mb of email. But, "you never
know
when you might need an old email in case someone tries to pull a fast one on you about
something they emailed you three years ago." CYA, I guess. I can't seem to get it
through to
them that email is insecure and, according to the RFCs, not even guaranteed to be
delivered.
And, don't get me started on all of these hoaxes, chain letters, spam, and virus
attachments... hehe
</rant>
Later,
Clint
Scott Ramshaw wrote:
> You guys might want to check your webserver's access log when sqwebmail
> goes nuts. I had over 10000 messages in my Inbox and whenever I would use
> sqwebmail it would use very high CPU resources while parsing through my
> Maildir with thousands of files. What you see as sqwebmail going nuts may
> just be a user with tons of email using sqwebmail.
--
S. Clint Bullock
Network Administrator
University of Georgia
Office of the Vice President for Research
626 Boyd GSRC
Athens, GA 30602-7411
(706) 542-5936
(706) 542-3837 FAX
