Hi,
I am trying to figure out why AutoScaler is not working. I have deployed my
Flink job on Yarn using AWS EMR.

My configurations are:

job.autoscaler.enabled: 'true'
job.autoscaler.decision.interval: 10m
job.autoscaler.stabilization.interval: 50m
job.autoscaler.vertex.min-parallelism: 35
jobmanager.scheduler: adaptive
jobmanager.adaptive-scheduler.combined-restart.enabled: true
jobmanager.adaptive-scheduler.combined-restart.window-interval: 10m

In the job manager log I see this:

2025-09-10 07:33:14,198 INFO org.apache.flink.runtime.scheduler.autoscaler.
Autoscaler [] - Scaling Metrics History for job
2705bd9bac1ad0498aa5983c3ca55b82
is empty
2025-09-10 07:33:14,198 INFO org.apache.flink.runtime.scheduler.autoscaler.
Autoscaler [] - Empty Metric History for job 2705bd9bac1ad0498aa5983c3ca55b82,
so not evaluating.

Is there something I should check ?

I remember while back the autoscaler was working when I was going with
default values for intervals, and I found it too aggressive.
So I relaxed these intervals and after that I don't see this working.

Is there any issue with my configs causing it to not work ?

Thanks
Sachin

Reply via email to