Problem with metric event

[radek_jasinski]

Hi,

In one of my environments, I need to send an alert when metrics are not sent from Azure to Dynatrace. I have a basic integration with Azure configured and am collecting information about the VMs (CPU, memory, disk). Unfortunately, no metrics are available to inform me, for example, that a VM has been shut down or restarted. We have prepared a metrics event that has the "Alert in missing data" option enabled.

builtin:cloud.azure.vm.cpuUsage:filter(and(not(in("dt.entity.azure_vms,entitySelector("type(azure_vm),tag(~"(Azure)V:XXXX~")"))))):splitBy("dt.entity.azure_vm")))

Of course, the threshold is 101% CPU utilisation so that it does not report other deviations.

The preview shows correct operation. Unfortunately when running in a production environment the alerts are reported randomly and close a minute later. This happens for several vm's at the same time. The time windows are set as shown in the screenshot:

Radek

[Mizső]

Hi @radek_jasinski,

Have you tried this metric for host restart? builtin:host.uptime

I hope it helps.

Best regards,

Mizső

[radek_jasinski]

Hi @Mizső 

I haven't tried it. I understand that this metric works for VM's in Azure?

[radek_jasinski]

Unfortunately, this metric is not available in my configuration. I don't have OA installed on my hosts.

[Mizső]

Hi @radek_jasinski,

Sorry I missed this information.

Best regards,

Mizső

[radek_jasinski]

I'm reporting the problem to support - I'll let you know if I find out anything.

[radek_jasinski · Accepted solution]

On the recommendation of support we have increased the value for violating samples to 24 and for the moment this has helped:)