Alerts via Azure Ressource Health have no details and dont aggregate

[Niklas Klein]

Hey everyone,

I am currently trying to achieve the following: We have a customer who runs JIRA datacenter cluster via the Azure Template (https://azuremarketplace.microsoft.com/en-us/marketplace/apps/atlassian.jira-data-center?tab=overview). We have an integration into our own JIRA Service Desk in that way, that if an Azure Service Health or an Azure Ressource Health Alert is triggered, a new Ticket is created with all the information.

However, as Azure Alerts are very chatty, our Service Desk gets kinda flodded with tickets if something fails. Therefore I am currently  evaluating if it would be a good Idea to put in Opsgenie inbetween to aggregate the Azure Alerts e.g. if some Alert is flapping and triggered every 5 minutes we would not end up with 12 JIRA-Issues after an hour but instead only 1 Issue as Opsgenie would just increase the count of the alert to 12.

However, my Opsgenie Alerts do not have any additional data, nor do they get aggregated.

What did I do: First I deployed the jira software datacenter template in a stagging environment to experiment with and setup an Opsgenie Trial. I configured a team and the azure integration in Opsgenie. Then I created an Azure Resource Health Alert that triggers on a flapping Network usage as seen below

As you can see in the first screenshot I get alerts and I also get Mails, both directly from Azure and Opsgenie. However, the alerts are not aggregated and more or less "empty".

I tried to switch between the traditional and new alert scheme in Azure. No effect. I tried to find any additional configuration in Opsgenie but it does not look promising. 

If I have a look in the logs of Opsgenie I get a warning that he "Skipps incoming data as there are no mathcing actions found". However, I have no Idea what that means ;)

I am glad for any ideas and tipps. And if I made an rookie mistage, just point me to the correct documentation, I would be fine with that :)

Greetings, 

Niklas

[Samir]

Hi @Niklas Klein !

Good idea to have the alerts go from Azure to Opsgenie first, and then let Opsgenie create the jira issue so you can leverage alert de-duplication in Opsgenie, so you only have 1 Opsgenie alert, and only 1 Jira issue gets created!

It looks like the issue is that the payload we are receiving is not the payload that the AzureResourceHealth integration is "expecting". Did you setup the alerts on the Azure side according to this section of our doc? https://support.atlassian.com/opsgenie/docs/integrate-opsgenie-with-microsoft-azure-resource-health/#Configuration-in-

Also - the logs should show another log for the raw payload we received. The log entry will have title "Received Integration Request". The one you sent is the "processed" log entry, so it shows that the integration processed all the fields as null since the payload it received is not what the integration is expecting.

Often times, the solution is to just use the regular Azure integration, depending on the format of the payload we are receiving. Please feel free to create a ticket with us via support.atlassian.com, so that we can take a look at your account to see the logs and the payload that is being received!

Thanks,

Samir

[Niklas Klein]

Thanks for the answer @Samir . I just had no time to check on this issue but will post any progress as soon as I am able. At least I could quickly check the logs and yep, there is data incoming.

[Samir]

Sounds good @Niklas Klein !

That payload looks like the one our regular Azure integration is expecting.

e.g.

So you may want to use the regular Azure integration in Opsgenie, instead of the AzureResourceHealth integration that you are currently using.