If there isn't enough memory allocated on a server tasks may hang on the server. Find step by step tips to solve it.

Find excellent tutorial on restart on task hang. Know how to setup the server so it restarts if this occurs automatically.

Sometimes on a Linux server, particularly if there isn't enough memory allocated on a server or the tasks being queued cannot be done with the available resources over a period of time, tasks may hang on the server. This can be fatal and cause the server to become unresponsive.

To setup the server so it restarts if this occurs automatically you need:

Login to SSH

To make these settings live without rebooting type:
sysctl kernel.panic=5
sysctl kernel.hung_task_panic=1
sysctl vm.panic_on_oom=1

Edit the configuration file so the changes will also occur after a reboot type:
pico /etc/sysctl.conf

Add the following to the bottom of the file

# Reboot 5 seconds after panic
kernel.panic = 5

# Panic if a hung task was found
kernel.hung_task_panic = 1

# Panic if OOM-Killer is initiated
vm.panic_on_oom = 1

To save Press: CTRL O
To exit Press: CTRL X