Unable to clear alerts

[bjohn]

Running a trial of vrealize with the horizon view (5.1.3) adapter

Version 6.0.1.2523163 Build 2523163

I have a number of "virtual machine has chronic high memory workload leading to memory stress" alerts that I'm unable to clear. Every time I clear them, they just come back. I know these machines have no memory issues. Anything I can do to clear them permanently?

[funksoul]

Hi, bjohn.

In vR Ops, alerts are always fired if conditions are met.

That means you cannot just clear alerts without resolving the alert conditions.

In your case, you can review the alert definition and change the symptom threshold or disable the alert for group of objects from the policy.

Regards,

Ho-Sung.

[bjohn]

Thanks for replying. So, how do I clear this alert - virtual machine has chronic high memory workload leading to memory stress. This alert may have been triggered when I recomposed the view desktops, but no longer do the VM's have a high memory workload. I don't want to disable the alert because I would like to know when this condition is true again. Thanks!

[funksoul]

Hi, bjohn

There are many ways to clear that alert.. for example,

- Make the recompose task routine so that vR Ops can learn the behavior and create a new alert definition that adds a dynamic threshold criteria to the original alert.

- Put the VMs into maintenance mode in vR Ops while the recompose task is in progress. (metrics are not collected and no alert fires when an object is in maintenance mode)

- Push the snooze button to make the alert go into sleep 9999999999 minutes. (just a joke but it will do the work anyway)

Regards,

Ho-Sung.

[bjohn]

I'll have to remember 1 and 2 when I recompose next time, but how do I get rid of the alerts for now? I guess I don't really understand vrealize. If the situation that was raised thru an alert no longer exists, shouldn't I be able to cancel the alert and not have it pop-up until the situation actually happens again?

[funksoul]

Hi, bjohn

If the situation that was raised thru an alert(symptoms in vR Ops) no longer exists, the alert get cancelled automatically.

Though there are alerts that can be cancelled by pressing the cancel alert button,

performance metric based alerts cannot be cancelled by this way because the metrics are refreshed on each collection interval.

They only get cancelled by the system if the symptoms are resolved.

So if you want to control an alert:

1. Review the alert definition with related symptoms and

2. [Modify the alert definition] or [adjust the symptoms thresold in the policy]

Regards,

Ho-Sung.