At 12:48 AM -0400 10/19/02, Morgan Miskell wrote:
Suddenly got a new development on one of my qpopper servers, have 5 machines, all running RH 7.0 with Qpopper 4.03, 4 of them are still fine, the 5th machine now has a "hanging" problem.
Symptoms appear to be that ever 3-4 hours the machine gets an error in the logs that looks like the following:
Oct 18 08:02:16 qpop5 popper[27438]: I/O error flushing output to client eagleloc at xxx.xxx.xxx.xxx [xxx.xxx.xxx.xxx]: Operation not permitted (1) [pop_send.c:689]
Then multiple entries in the log that are like:
Oct 18 09:52:37 ns5 popper[13502]: (null) at xxx.xxx.xxx.xxx (xxx.xxx.xxx.xxx): -ERR POP EOF or I/O Error [popper.c:820]
Once this starts, the only way to get qpopper back to running to restart xinetd. If you telnet to port 110 during this period you successfully connect but never get the initial ok message.
I have recompiled and re-installed the binary, rebooted the server, checked the file systems, checked memory and disk space, anyone have any ideas?
Yesterday I upgraded the Daemon to 4.04, but that didn't have any effect, I am running the Daemon in server mode via xinetd. Thanks in advance!
/etc/xinetd.d/pop3 entry as follows:
service pop3
{
socket_type = stream
protocol = tcp
wait = no
user = root
instances = UNLIMITED
server = /usr/local/sbin/popper
server_args = popper -S -s -T 30 -c -R -C
port = 110
}
What does strace(1) on the hung process show?
