very high server load problem
 

hello every one.....

i am having Red Hat LINUX 5 Enterprise Server and facing problem regarding very very high server load (load average is going high up to 60-70)due to which server is getting hang.

Can any one suggest me the solution for this problem????

Hi
Use top, htop, atop, iostat or any similar software for dynamically monitoring system and processes.

Well, that's a high load number, and it is probably reasonable to say that, in general, a server with that high a load number will not be performing as snappily as a server with a more reasonable load number.

However, given the level of detail that is provided, it is only possible to provide very generic suggestions:

• find out what is wrong
• put it right

other options worth considering are:

• use your Red Hat support and get them to put it right (although, you'll actually find that this involves you answering a lot of their questions, and have them provoke you to put it right)
• hire someone who knows what they are doing, and get them to put it right
• increase you level of knowledge so that you can put it right yourself
• actually post enough detail so that someone could actually help you with the problem

and bear in mind that

• it could be that your load number is that high, because that high is how high the load number gets with a server of that spec with that workload

Looking forward to finding out more about the problem!

Hi,

Which praticluar process consumes more cpu usage.

dinakumar12
 

Process is mainly HTTPD

personally I find the various top/stat tools of minimal value. Sure, they tell you what's happening now, but how do you compare say the disk load to the top process or memory or cpu utilization? and by the time the screen refreshes you forgot what the name of the top process was or which disk was having issues.

As one might guess I always have collectl running, happily collecting hundreds of data points as well as process data. Then when I do encounter a problem I simply plot EVERYTHING over the time period in question and look at the patterns. If there is a high cpu load, what else is going on at the time? If this is an httpd process, is there a correcponding high i/o load? Might someone be hammering your webserver? Could be almost anything but until you look at the overall system health rather than focusing on one or two things, you may never know.

-mark

Hi,

Have you checked apache access logs and error logs.

High load averages can be due to a number of things:
1) CPU is overloaded and can't keep up with computational load (in my experience this is kind of unusual).
2) hard drive is overloaded and all the processes are waiting for the hard drive to respond. Sluggish hard drive might also mean pending drive failure which is BAD.
3) Not enough memory to handle all the work so redhat is using the hard disk swap. This is quite common and installing more memory can be quite helpful.
4) Network overload. Network card is overwhelmed and all your processes are waiting on it.

To find out which is really the problem, login via SSH and use the "top" command which lets you view and sort the processes by CPU load, memory consumption, etc.