Red Hat Bugzilla – Bug 11878
Short repeated IMAP sessions causes inetd to switch it off
Last modified: 2008-05-01 11:37:55 EDT
If your mailer does remote sync (eg netscrape) which involved going into a
large number of IMAP folders saying "Anthing new?" and exiting if the
answer is no, then inetd sees this as looping and issues:
inetd: imap/tcp server failing (looping), service terminated
It then bounces all further connections to port 143.
Can anything be done to alleviate this. This is either a documentation
problem or a mismatch between the 3 components netscape, inetd & imapd.
The are all doing what they are supposed to do, but the overall function
of syncing mail across a number of frequently un-updated mail folders
This is a feature, not a bug. Add a maximum rate in /etc/inetd.conf like
imap stream tcp nowait.<MAX> root /usr/sbin/tcpd imapd
where <MAX> is the number of permitted spawns per scond (default per man
page is 40)
Oops ... didn't spot this in the man pages .. mea culpa.
The description of the optional .max parameter was tucked away as the last
sentence in a paragraph mostly about UDP packet management.
Maybe a para break in the man pages ??