Related to Scheduled tasks should avoid overloading the streamprocessor · Issue #13870 · camunda/zeebe · GitHub we have experienced high processing queue. After update to Zeebe 8.2.18 the internal events cache was introduced which improved the behavior but yesterday the problem reoccurred with low fni ~50. It might be related to open issue Scheduled commands cache uses the wrong intent for caching message expiry · Issue #15574 · camunda/zeebe · GitHub as we have a lot of publish message commands with ttl but there is no way to confirm it from our side.
Hi @marek.hlavacek, welcome to the forums! @Niall asked me to take a look at this, and I admit, this is a bit over my head! I’m going to share this with our engineers and get their thoughts.
Hi, thank you for that
Hi @marek.hlavacek - the Zeebe team has been looking at the information you provided. The second GitHub issue you mention (#15574) is probably not the culprit, as that is related to something introduced in 8.3 and you are on 8.2.
They recommended trying to update to 8.4, because 8.4 comes with additional metrics (such as % of written commands grouped by value type/intent) that will further help identify what is happening.
I know you mentioned you had issues with 8.3.3 also; can you share some more information about that? 8.4 was an out-of-band release for us, intended to further stabilize the 8.3 release. It’s possible the issue(s) you experience with 8.3 have already been resolved with 8.4.
Lastly, have you been in touch with our sales team? From my understanding, you are working on a proof of concept. Depending on what you’re working on, we might be able to get some engineers to help troubleshoot deeper.
Hi, thank you for the update. I will check with the colleagues and let you know.
For the last, yes we are in the contact with sales discussing the partnership and contract.
Discussion with the sales team will continue after our internal discussion about the future steps with the platform it self. For the issue i have reported we are now deploying 8.4 to our test environments.
@marek.hlavacek - thanks for the update, let me know if I can help in any way. Have you tested with 8.4 yet, or deploying it now to see if the issues are resolved?