Why my alert is not creating a problem?

[davidBermu]

We have an issue with our AWS Backup alert in Dynatrace. The metric only sends a value of 1 when a backup job fails and no data when there are no failures.

The alert simulation detects the condition, but no corresponding event or Problem appears in Dynatrace.

We suspect that the absence of data prevents the alert from being evaluated as expected. We would like to confirm whether DQL supports returning 0 when no data is available, and whether this approach is sufficient to ensure reliable alert evaluation and Problem creation. The sliding window was set to 5 and the violating samplet o 1. So, I don't understand why the alert is not generating a problem.

Could you please advise on the recommended configuration for this scenario?

[rastislav_danis]

which event.type have you used in alert definition?

[davidBermu]

I have defined a CUSTOM_ALERT as event.type in details alert window. 
I have more alerts created and the events works, but i don't understand why this kinf of event doesn't create an event or a problem... When the events was checked in the fetch events, I've not seen anything.

[rastislav_danis]

Try to add ", interval:1m" into timeseries command (before ", by:").

Can you post your dql from alert definition as text, it's not very readable from picture.

[davidBermu]

Hi, 

The dql with the modification was:
timeseries failed_jobs = sum(
`cloud.aws.backup.NumberOfBackupJobsFailed.By.BackupVaultName.ResourceType`, default: 0),
by: {
`aws.account.id`,
BackupVaultName
},
interval: 1m

The alert was updated with the change... since tomorron I will not know if this works or not... I will update tomorow with an update on that.