On Tue, May 11, 1999 at 02:57:19PM +0100, Kevin J Collins wrote:
> Sorry to be a bit off topic, but can anyone help me understand
> what may be going on here?
>
> The Symptoms: Exim-2.12 running on RedHat 5.2 on a Pent. II/450
> with 500M ram. The imapd in use is Washington Uni vers 12.250.
>
> Periodically, the system all but keels over with the following
> messages in the messages log:
>
> May 10 11:04:42 mail1 inetd[334]: imap/tcp server failing (looping), service terminated
> May 11 13:20:06 mail1 inetd[334]: imap/tcp server failing (looping), service terminated
> May 11 14:08:37 mail1 inetd[334]: imap/tcp server failing (looping), service terminated
> May 11 14:19:09 mail1 inetd[334]: imap/tcp server failing (looping), service terminated
> May 11 14:23:20 mail1 inetd[334]: imap/tcp server failing (looping), service terminated
> May 11 14:27:36 mail1 inetd[334]: imap/tcp server failing (looping), service terminated
This has nothing to do with exim. inetd has a "feature", whereby it turns
off a service temporarily (10 mins) if the service receives too many
connections in a short time. This feature is there to protect the machine
from DOS attacks, for example.
2 solutions:
1. Either change your inetd config to allow more connections per second
(read the manpage for inetd for details), or,
2. Use tcpserver from the ucspi-tcp package. Find ucspi-tcp at:
ftp://koobera.math.uic.edu/pub/software/ucspi-tcp-0.84.tar.gz
tcpserver does not shutdown services like inetd does, and can also do
excellent connecton limiting.
--
System Administrator
See complete headers for address, homepage and phone numbers
--
*** Exim information can be found at
http://www.exim.org/ ***
