If Apache fails, and will not successfully start again, check the error log. If you see an error similar to the following, it could indicate that your server has run out of semaphores.
semget: No space left on device
To see how many semaphores are being used, SSH to your server as root and run the following:
ipcs -s
In order to get Apache started again, we must clear the semaphores. Run this for-loop to flush them:
for whatever in `ipcs -s | awk ‘{print $2}’`; do ipcrm -s $whatever; done
On older servers that command may not work. In these cases, you may need to do the following:
/sbin/service httpd stop
ipcs -s | grep nobody | gawk ‘{ print $2 }’ | xargs -n 1 ipcrm sem
/sbin/service httpd start
If this is a common problem for you, you may want to increase the semaphore limits on your server. You can do that by adding the following to the /etc/sysctl.conf file:
# Increases the semaphore limits & extend Apache’s uptime.
kernel.msgmni = 512
kernel.sem = 250 128000 32 512
Then load the new settings into the kernel:
sysctl -p
Note: This post assumes you are running Apache on a Linux server, are familiar with the command line, and have root access to the server.
Recent Comments