We use Event hooks with filtering to send various things into workflows and than do something, mainly around user gets added to one of these groups do X. We are running into an obnoxious issue where the event hook will fire two events that are exactly the same even down to the event ID.
This isn’t a major issue but does cause impact some downstream services, as we have been able to build some logic into some flows where it does affect those services. But this leaves me thinking I might as well have just used a user added to group trigger and than done all the filtering in the flow, surely this isn’t the intended outcome of event hook filtering ?
Nope I only see one generated event in the system logs that matches up with the time stamp. Though I do get the following error:
[nested exception is io.netty.handler.timeout.ReadTimeoutException. Exceeded maximum retry limit of 1]
Which is weird as a bit of google fooing is saying that error should only be generated when it fails to send and receive a response within a certain timeframe, since I’m sending internally I would have thought that would be a 0 issue.
Yea annoyingly I have done all that, will reach out to Okta support, I did create a ticket ages ago (as we where very keen to use the feature) that’s how we have the workaround. But I would think if it’s been fixed and not intended, these new hooks/flows shouldn’t be acting this way.
Seems like you must respond within the 3 second window and send a close. This will stop it from sending the event again. I feel like Okta should update this so if it’s an internal request it should be handled gracefully insidie the API Endpoint card.