Re: [Imap-uw] RE: Stale imapd process hanging around

2010-04-16 Thread Mark Crispin 

On Fri, 16 Apr 2010, Moray Henderson wrote:

Anyone?  Is there a build option I could try?  Should I have posted to
another list?  Was it a silly question?  Shall I give up trying to fix
it and just write a script to kill old imapds?


Have you tried to see what is specific to that user?

If you are using traditional UNIX mailbox format and the user has a
ridiculously large mailbox, perhaps you should look into a different
mailbox format.

You should also consider trying a different kernel (e.g., Linux) in just
this is a specific problem to CentOS.

-- Mark --

http://panda.com/mrc
Democracy is two wolves and a sheep deciding what to eat for lunch.
Liberty is a well-armed sheep contesting the vote.
___
Imap-uw mailing list
Imap-uw@u.washington.edu
http://mailman2.u.washington.edu/mailman/listinfo/imap-uw

[Imap-uw] RE: Stale imapd process hanging around

2010-04-16 Thread Moray Henderson 
I wrote:
>Hello all,
>
>Using uw-imap-2007e with squirrelmail-1.4.8 on CentOS 5.4, one user
>consistently leaves stale imapd processes.  Other users can use
>SquirrelMail, and their imapd processes terminate normally.  Discovered
>the problem when xinetd refused to start any more imapds.
>
>According to netstat, all his processes are in CLOSE_WAIT, but have
data
>in the Recv-Q:
>
>Active Internet connections (w/o servers)
>Proto Recv-Q Send-Q Local Address   Foreign Address
>State   PID/Program name
>tcp 2898  0 localhost.localdomain:imap
>localhost.localdomain:58161 CLOSE_WAIT  4818/imapd
>tcp 2898  0 localhost.localdomain:imap
>localhost.localdomain:58165 CLOSE_WAIT  4848/imapd
>tcp 2898  0 localhost.localdomain:imap
>localhost.localdomain:57142 CLOSE_WAIT  4877/imapd
>tcp1  0 localhost.localdomain:imap
>localhost.localdomain:52544 CLOSE_WAIT  11984/imapd
>tcp1  0 localhost.localdomain:imap
>localhost.localdomain:52545 CLOSE_WAIT  11996/imapd
>tcp 5321  0 localhost.localdomain:imap
>localhost.localdomain:40515 CLOSE_WAIT  15245/imapd
>tcp 2898  0 localhost.localdomain:imap
>localhost.localdomain:38256 CLOSE_WAIT  4916/imapd
>tcp 2898  0 localhost.localdomain:imap
>localhost.localdomain:38258 CLOSE_WAIT  4944/imapd
>tcp1  0 localhost.localdomain:imap
>localhost.localdomain:43402 CLOSE_WAIT  12002/imapd
>tcp1  0 localhost.localdomain:imap
>localhost.localdomain:43404 CLOSE_WAIT  12014/imapd
>
>The user has a 200K mailbox in /var/spool/mail and 52M of mail folders
in
>his home directory.  There do not appear to be any stray null
characters
>in the mailboxes.
>
>Any suggestions on how to get his imapds to close properly, or how to
>debug this further?


Anyone?  Is there a build option I could try?  Should I have posted to
another list?  Was it a silly question?  Shall I give up trying to fix
it and just write a script to kill old imapds?


Moray.
"To err is human.  To purr, feline"

___
Imap-uw mailing list
Imap-uw@u.washington.edu
http://mailman2.u.washington.edu/mailman/listinfo/imap-uw