ISPConfig Crash due to unsuccessful Apache graceful restart

#1
Hi!

First of all I wanted to say that I highly appreciate the work of the people here in the forum. I always find highly professional answers and articles about topics which already helped me a lot.

I am runnin ISPConfig 3.0.1.6 on Ubuntu 9.10 w. Apache 2.2.12.

I figured, that after i change certain values in the ISPConfig GUI, it stops working.

After checking the Apache logs, i found out, that the apache error log shows:


Code:
[Mon Mar 22 08:30:02 2010] [notice] Graceful restart requested, doing restart
[Mon Mar 22 08:30:02 2010] [notice] seg fault or similar nasty error detected in the parent process
[Mon Mar 22 08:30:02 2010] [error] (9)Bad file descriptor: apr_socket_accept: (client socket)
So I tried around starting and stopping apache witht he init.d command and it seems, that after issuing the stop command, there are always still apache processes running, which obviously causes the graceful restart to crash after downing the apache.
netstat -nlp gives:

Code:
tcp        0      0 0.0.0.0:25              0.0.0.0:*               LISTEN      1823/master
tcp        0      0 127.0.0.1:10024         0.0.0.0:*               LISTEN      960/amavisd (master
tcp        0      0 127.0.0.1:10025         0.0.0.0:*               LISTEN      1823/master
tcp        0      0 0.0.0.0:3306            0.0.0.0:*               LISTEN      1102/mysqld
tcp        0      0 127.0.0.1:783           0.0.0.0:*               LISTEN      1166/spamd.pid
tcp        0      0 0.0.0.0:21              0.0.0.0:*               LISTEN      1836/pure-ftpd (SER
tcp        0      0 78.46.99.253:53         0.0.0.0:*               LISTEN      1735/mydns
tcp        0      0 127.0.0.1:53            0.0.0.0:*               LISTEN      1735/mydns
tcp        0      0 0.0.0.0:22              0.0.0.0:*               LISTEN      943/sshd
tcp6       0      0 :::25                   :::*                    LISTEN      1823/master
tcp6       0      0 :::443                  :::*                    LISTEN      2118/apache2
tcp6       0      0 :::993                  :::*                    LISTEN      1695/couriertcpd
tcp6       0      0 :::995                  :::*                    LISTEN      1731/couriertcpd
tcp6       0      0 :::110                  :::*                    LISTEN      1710/couriertcpd
tcp6       0      0 :::143                  :::*                    LISTEN      1671/couriertcpd
tcp6       0      0 :::8080                 :::*                    LISTEN      2118/apache2
tcp6       0      0 :::80                   :::*                    LISTEN      2118/apache2
tcp6       0      0 :::21                   :::*                    LISTEN      1836/pure-ftpd (SER

This is AFTER issuing the
Code:
apache2 stop
command. You can see that there is still an apache process running for protocol tcp6, although there is none anymore for protocol tcp.

I can fix this issue by killing hte processes with e.g.
fuser -k -n tcp 80,
but this keeps happening at times when i use ispconfig and change values that require a graceful restart.

I checked the confs, it seems clean, no double Listeners or such...

I already resinstalled the system once, then it was ok, but now it reappeared, so i would like to find the reason for it...

Any ideas?
 

Werbung