Monitoring suggestions??

[agentguerry]

What are your suggestions on what you monitor/collect on the linux side of things as far as capacity planning?

thanks!

[nelsonab]

Depends on what specifically you're looking for, however Disk, CPU and Memory are all good places to start. You want to also know how much swap space is used along with the amount of cache memory used. Free memory tends to be sucked up by cache, however if you find your system running with little or no memory dedicated to cache you might be cutting it thin on memory.

Also you might also want to look at your IO wait state (cpu details), that will tell you if you've got an IO issue.

Disk is a pretty obvious one for most folks. However the fist time you can look at a graph and trend when you will run out of disk space completely will have you sold pretty much forever.

You might also want to look at how your network cards are doing, are you maxing things out in that regard?

Also looking at how various applications are performing might be useful such as some of the various mysql monitoring threads.

If you can get temperature monitoring working I would recommend that.

Hope that helps get an idea of where to start.

[agentguerry]

great comments.
we're trying to do capacity planning.
I think i'll start here, and go forward.

thanks.