Shane Chrisp wrote:
On Fri, 2005-07-01 at 19:49 +0300, Anton Butsyk wrote:
For now only 11 connections to qmail:
97 processes: 11 running, 86 sleeping
CPU states: 95.7% user, 0.0% nice, 2.0% system, 2.3% interrupt, 0.0%
idle
Mem: 53M Active, 53M Inact, 46M Wired, 59M Buf, 329M Free
Swap: 1024M Total, 1024M Free
PID USERNAME PRI NICE SIZE RES STATE TIME WCPU CPU COMMAND
818 root 124 0 3244K 2332K RUN 0:44 9.08% 9.08%
qmail-smtpd
827 root 123 0 3244K 2332K RUN 0:19 9.08% 9.08%
qmail-smtpd
820 root 124 0 3244K 2332K RUN 0:29 9.03% 9.03%
qmail-smtpd
825 root 123 0 3244K 2332K RUN 0:24 9.03% 9.03%
qmail-smtpd
831 root 123 0 3244K 2332K RUN 0:17 9.04% 9.03%
qmail-smtpd
833 root 123 0 3244K 2332K RUN 0:16 9.04% 9.03%
qmail-smtpd
828 root 123 0 3244K 2332K RUN 0:18 8.99% 8.98%
qmail-smtpd
849 root 123 0 3244K 2332K RUN 0:09 8.88% 8.64%
qmail-smtpd
851 root 123 0 3244K 2332K RUN 0:06 9.46% 8.64%
qmail-smtpd
I definately agree that there is something wrong here. You seem to have
heaps of memory free, but what speed is the cpu in the system? What
happens if you stop and restart the qmail-smtpd service? Does it help if
you only allow 5 concurrent connections at a time?
Shane
Thanks you all very match.
It was just rewrite qmail-smtpd script:
#!/bin/sh
QMAIL="/var/qmail"
ME=$(head -1 $QMAIL/control/me)
PATH="$QMAIL/bin:$PATH"
USER=${USER:="qmaild"}
exec /usr/local/bin/softlimit -m 5000000 \
/usr/local/bin/tcpserver -v -DURl $ME -x
$QMAIL/control/qmail-smtpd.cdb -u "qmaild" 0 smtp \
/usr/local/bin/rblsmtpd -t 10 -b \
-r bl.spamcop.net \
-r list.dsbl.org \
-r cbl.abuseat.org \
$QMAIL/bin/qmail-smtpd 2>&1
and everything begin works fine.
Also add -b option to rblsmtpd to use a 553 error code for IP addresses
listed in the RBL
because early they just standed in the queue.
Regards,
Anton