Home / server / Questions / 544345
Accepted
malat
Asked: 2013-10-08 07:21:06 +0800 CST

Monitoring average load

  • 772

I am trying to monitor a system load average using uptime. On my debian box, I used the following:

$ for ((i=0;i<=1000000000;i++)); do uptime; done

However looking at top, 18% of CPU was allocated to bash ! This means that my uptime checking is way too agressive.

How do I gently check the load average ?

linux

3 Answers

  1. Best Answer

    Have you tried watch?

    watch uptime
    • 5

  2. Isn't this a good case for running a lightweight monitoring daemon like Monit? You can set a threshold and have the system email you under certain conditions...

    Also, is top an option here?

    • 5

  3. Your sample code is querying the uptime as fast as the system can issue serial uptime commands, which is not what you want.

    You need to utilize a tool that will query at an interval (polling).

    Your existing code could be "fixed" by adding a sleep statement. This would cause the polling granularity to be lessened, but your overhead caused by your monitoring would be reduced. This sample would check every minute.

    $ for ((i=0;i<=1000000000;i++)); do sleep 1m; uptime; done

    That said, you really want to use a tool that was built for monitoring rather than writing your own.

    • 4