Open tomkerkhove opened 1 year ago
This would be perfect for us. For some specific queues, we leverage the visibility of messages in the queue, to postpone processing them. The current implementation means that KEDA will keep scaling up, because there are invisible messages in the queue, but the jobs it creates have nothing to do, since there are no visible messages in the queue.
So, being able to specify that for this particular trigger, we only want to consider the visible messages, would allow this to work for us without continuously creating useless jobs in the cluster.
Hi @tomkerkhove et al, is there any known solution/workaround here or a path forward? The issue still persists and blocks from using Azure Container App Jobs as advertised.
The PR is open as you can see above your comment.
Azure Container App Jobs as advertised
Mind elaborating on this? =)
Proposal
I believe we should introduce a new configuration option that will influence this:
Use-Case
We've been going back and forward if Azure Storage Queue scaler should consider invisible during evaluation or not, but this is not ideal for everyone. Example https://github.com/kedacore/keda/issues/4432.
Is this a feature you are interested in implementing yourself?
No
Anything else?
No response