Forum Discussion

Hare_Krshn's avatar
3 years ago

LM monitor VMWare guests

Hello.,

We are using cloud based logicmonitor, and we added both AWS as well as VMWare to logicmonitor. And we are getting what we want in terms of alerts and resource usage graphs.

Question mark - 

- In VMWare, we have 2500+ guest machines and we are trying to  identify the "idle" guest machines and reclaim if no one is using. 

- Is there any process in LogicMonitor (plugin sort of) in order to tell that guest machines were idle for a specific time (say 1 week) so that we can take action and claim in order to avoid causing exhausted capacity.

5 Replies

  • Anonymous's avatar
    Anonymous

    Ok, just to set the stage: you are using a Collector installed on a VM in AWS. You are using that collector to directly communicate with VMs. You want to know if there is a way to highlight the idle days of a VM.  If all that is right, then...

    There is a property source in the exchange called Idle_Days, locator TYTXCL. You'll need to add the api credentials allowing the Collector to log into the LM API so that it can fetch the deep data that gets translated into the number of days that a device has been idle.

  • 18 minutes ago, Stuart Weenig said:

    Ok, just to set the stage: you are using a Collector installed on a VM in AWS. You are using that collector to directly communicate with VMs. You want to know if there is a way to highlight the idle days of a VM.  If all that is right, then...

    There is a property source in the exchange called Idle_Days, locator TYTXCL. You'll need to add the api credentials allowing the Collector to log into the LM API so that it can fetch the deep data that gets translated into the number of days that a device has been idle.

    Is there any document about this "Idle_Days" and "property source in the Exchange" so that we can identify that a device has been idle. 

  • Anonymous's avatar
    Anonymous

    The Idle_Days property source is self documenting. As for how to grab it from the exchange, you can just go to the exchange in your portal and there's a "Locator search" option in the top right.

  • @Stuart Weenig I think what @Hare Krshn is looking for is more along the lines of VMs with idle CPUs, not necessarily VMs that are totally down.

    I'm not sure if a Report can do this. Top 10 graphs for VM CPU will show the least-idle machines, but I can't think of a way to easily see the idle machines.

  • Anonymous's avatar
    Anonymous

    Oh I see. Yeah, the definition of idle is important. 

    Probably top-N graph would show that.