Auto close or do not alert on CloudWatch Alarms

[Paul Donnelly]

Hi,

I have set up an AWS CloudWatch integration with OpsGenie.

I am receiving alerts that I don't want to receive. When a CloudWatch alarms goes from INSUFFICENT to OK I get an alert in OpsGenie. Is there a way to stop these alerts?

I currently have close alert configured like this but it doesn't seem to work

The description basically contains part of the alarm description I receive into OpsGenie

Is there anything else I should try?

Thanks

[Paul Donnelly · Accepted solution]

In the end I set up an ignore rule for any alarms that have a status of OK and this stopped the recovery alerts coming in 

[Nick H]

Hi @Paul Donnelly ,

If these CloudWatch requests are creating alerts instead of closing them - then you will want to review the Create Alert action / which condition(s) is matching.

All integration actions are processed in an order of operations from top >> down. Opsgenie will use the first matching action - so in this case it sounds like Opsgenie is being told to create an alert before it has the chance to review the close action / filter.

For more on that: https://support.atlassian.com/opsgenie/docs/use-advanced-integration-settings/

You should also be able to simplify and consolidate that Close action's filter to something like this using regex:

[Paul Donnelly]

Thanks Nick

Does the alarm description have to match exactly? The reason I ask is that these alarms start with the same description but then can vary each time they come as they have different environment names or machine names

For example one alarm could come in as ScheduleUpgrade-dev-12345

Next time it could be SchedlueUpgrade-prod-56789

Is it OK to set the alarm description as ScheduleUpgrade

EDIT: I tried the alarm description as  ScheduleUpgrade but I still got an alert. 

Is there a wildcard option I can use? Something like  ScheduleUpgrade-*

[Nick H]

You should be able to use regex in that case.