Hi all, in <the docs> it is mentioned that Current...
# troubleshooting
a
Hi all, in the docs it is mentioned that > Currently,
Flink Autoscaler Standalone
only supports a single Flink cluster. It can be any type of Flink cluster, includes: > • Flink Standalone Cluster > • MiniCluster > • Flink yarn session cluster > • Flink yarn application cluster > • Flink kubernetes session cluster > • Flink kubernetes application cluster > • etc But in the source code of k8s operator, it is clear that SessionCluster SessionReconciler has a noop autoscaler. Am I missing something?
Copy code
public SessionReconciler(
            EventRecorder eventRecorder,
            StatusRecorderFlinkDeployment, FlinkDeploymentStatus statusRecorder) {
        super(eventRecorder, statusRecorder, new NoopJobAutoscaler<>());
    }
g
The autoscaler scales jobs not clusters
Session cluster is just the cluster , the autoscaler works for applications or session jobs on Kubernetes
a
I’m sorry I don’t think I get it. SessionReconciler is responsible for reconciling session jobs on Kubernetes and it uses
NoopJobAutoscaler
. So this means that SessionJobs cannot be scaled by operator right?
g
I think you are confusing the
SessionReconciler
with the
SessionJobReconciler
a
Both use NoopAutoscaler 😞 src
m
You are looking at an older release, this was updated in release 1.8
a
Yeah that’s intentional. I pulled up the doc and tag of 1.7.0.
g
ok then pull the latest 😉
a
Haha, we use 1.7.0. I’ll see if we 1.8.0 suits our use case and if I have a go-ahead for the upgrade.
g
1.9.0 is the latest version, I recommend using that
a
Hi Gyula, I upgraded one of our nonprod clusters to 1.9.0 and I am running a Flink 1.19.1 application. I can manually scale it but can’t seem to autoscale it. I tried putting load on it but it does not scale. Is there something I should know about the internals to see autoscaling in action? What metrics should I specifically target to trigger an autoscale?