Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

For a number of different reasons in the last year or two, I have had issues with servers that run processes which tend to cause very heavy load averages.

In these cases, I can often ping the server, but it becomes very unresponsive to any other connections, and I often have to power cycle the machine.

One reason for causing this huge load include a poorly configured Apache process on a virtual server that allowed too many concurrent connections for the hardware allocated to it, and I seem to have ran some programs on another server tonight that appears to have done something similar.

What I am curious about is linux has tools that detect a very large load average and interrupt those processes in some way, allowing the machine to recover from this?

My apologies if I have not worded this well, I appreciate it is pretty open-ended.

share|improve this question

2 Answers 2

up vote 5 down vote accepted

It seems to me that you've answered your own question.

One reason for causing this huge load include a poorly configured Apache process on a virtual server...

If you have a poorly configured Apache server, fix that configuration. You've already done the investigation, so now you should implement the proper fix. A script to interrupt/restart/kill processes that runaway is only a workaround.

All that being said - I don't know of any tools that would do what you're looking for. But I don't think you need a tool, you need a proper configuration for the system you're working on.

share|improve this answer

baumgart is correct, that you should solve the problem. However as a work around, you can have monit execute scripts if the load average goes too high. It's something of a hack though, so it's far from perfect. You could also write something to do this relatively trivially.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.