Hi! 2 things here:
1. You are mixing the main(unreleased) documentation with the stable docs. 2. Unless you are playing around with the experimental autoscaling logic (not part of 1.0.1), you should never need to set taskmanager replicas. Based on your spec this is configured automatically for you. I suspect in your case you should simply not set it. Cheers, Gyula On Wed, Jul 13, 2022 at 3:16 PM Moras, Clen <clen.mo...@solarwinds.com.invalid> wrote: > I have tested with the latest release : 1.0.1 > > Repo: ghcr.io/apache/flink-kubernetes-operator< > http://ghcr.io/apache/flink-kubernetes-operator> > > Image: c2148b1< > https://github.com/orgs/apache/packages/container/flink-kubernetes-operator/28318947?tag=c2148b1 > > > > Im currently facing the issue for scaling task manager pods. > > On the documentation : > > https://nightlies.apache.org/flink/flink-kubernetes-operator-docs-main/docs/custom-resource/reference/#taskmanagerspec > > It is mentioned that, we should be able to use replicas as a parameter, > but neither the CRD definition nor the TaskManagerSpec.java is updated for > the image. > > Getting the following image due to it. > error: error validating "workingbk.yaml": error validating data: > ValidationError(FlinkDeployment.spec.taskManager): unknown field "replicas" > in org.apache.flink.v1beta1.FlinkDeployment.spec.taskManager; if you choose > to ignore these errors, turn validation off with --validate=false > > Is this going to be resolved in 1.0.1 release? >
