Help wanted with AWX / AKS problem

I hope someone can help me.

I upgraded the node pool for our AKS cluster from 2 Standard_D4s_v3 servers to 2 Standard D4as_v7 servers.
The nodes are almost picking their nose because they have very little to do, but since the upgrade I get the following error a few times now:
Task was marked as running but was not present in the job queue, so it has been marked as failed.
I can’t imagine that the change from D4s to D4as has that much impact on the performance of AWX, and CPU/Memory of the nodes never reach very high levels (highest memory usage of the node I’ve seen was around 60%).

I found Task was marked as running in Tower but was not present in the job queue, so it has been marked as failed · Issue #9594 · ansible/awx · GitHub but that issue is still open.

Anyone has a clue why a D4s_v3 wouldn’t show this behaviour and a D4as_v7 would?