Hello!
I'm looking to get some feedback on this topic. Atlassian created a Bug Report for this issue and I'm tracking it but have low confidence it will be fixed anytime soon. After reflecting on the issue further I think things are more "working as designed."
Here's the bug report if you'd like to review the specifics of the wall we are hitting.
https://jira.atlassian.com/browse/JSDCLOUD-14424
Long Story short, we are ingesting alerts from a monitoring tool that is sending updates to...
https://api.atlassian.com/jsm/ops/integration/v2/alerts
We are sending an alertState: triggered to open and alertState: cleared to close. Both are going to the URL above. We are utilizing an Alias to correlate these alerts together. Alert correlation works, without a close rule JSM will dedup these alerts so I know the correlation is possible. JSM only looks for clears to be sent to the below...
https://api.atlassian.com/jsm/ops/integration/v2/alerts/{id}/close
This specific monitoring tool doesn't have the ability to store the alertID that JSM responds with in order to correlate the alert.
We technically can accomplish what's necessary with an Email integration although that will add a ton of needless work to properly format alert messages and then dissect the message once it is received by JSM.
Does anyone have any recommendations here? I've tried to think through possible workarounds with JSM but haven't been able to come up with something useful.
If you find the above Bug report to be important it would be appreciated if you could comment. If it affects your team please "vote" for the bug.
Thanks everyone!
Josh Rossman
I don't have any good reason to think this would work, but there could be that 0.1% chance we get lucky.
Looking at the old documentation for OpsGenie, they had an identifierType parameter that let you tell it "that thing in the URL is an alias" as opposed to an ID.
Maybe, just maybe, that parameter was ported over to the Atlassian API but wasn't documented.
Jim,
For some reason I wasn't notified of your response and just noticed it. I reviewed the old documentation you attached, it does seem like that would do the trick. I'm grasping at straws to get this working properly. If there's not going to be a workaround available within the next few weeks we may need to drastically change our approach to the topic.
Thanks,
Josh R.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.