Fix Apache frequent reboot problem-caught Sigterm, shutting down

There is a host in the host cluster is always a reboot phenomenon, very frequent, almost 2 days a time. At first Cosbeta did not notice it, because the time of reboot is very short, may also cause the Web server to interrupt 10-30 seconds.

However, the time for today's restart occurs during the day in China, just a user on the QQ to me said that the site can not open, and so I see the message to test, but also open, it seems that the Apache server restart speed is very fast, although the impact is not easy to solve the inevitable a little uncomfortable.

According to Cosbeta's experience, the problem is mostly Apache file configuration A little problem, after all, the Apache service can be forced to start, so first to check the error file, get the wrong hint as follows:

(show/hide) plain text [Sat 12:58:48 2009] [ERROR] [client 94.26.67.199] File does not exist:/VAR/WWW/HTML/WHM [Sat May 12:58:48 2009] [ERROR] [client 94.26.67.199] File does not exist:/var/www/html/favicon.ico [Sat 30 14:23:43 2 009] [notice] caught sigterm, shutting down

From this error log can not completely see what the problem is, so or test the Apache file configuration is correct. SSH Log on server:

Root@serverx [~]# apachectl configtest
[Tue June 09 02:03:45 2009] [ERROR] VirtualHost 66.40.x.x:0-mixing * Ports and non-* ports with a namevirtualhost address are not supported, proceeding with Undefined results

Sure enough, the original is a VirtualHost configuration error, IP after the addition of wildcard * as the port number, it seems that the Apache do not eat this set, the solution is either remove the wildcard * to the wildcard, or specify a specific port.

After correcting, again test Apache configuration, all OK, problem solved.