Server Job fails on different nodeSelectors

[flashpixx]

Describe the bug

I run the helm chart with a nodeSelector, so the server-job runs on a different node than in my case cassandra, but the job fails every time. I set these values:

cadence:
  server:
    nodeSelector:
      <my node key: <my node value>

Steps to reproduce the issue:

Apply the values above

Expected behavior

Server-Job should run also on different kind of nodes, it my helpful if the nodeSelector can be set as well for the server-job

Additional context

I try this with helm chart version 0.23.0 and 0.22.1

[akijakya]

Hi!

I tried to reproduce the bug with chart version 0.22.1 on AWS by creating a cluster with nodes of different instance types, and using the node.kubernetes.io/instance-type label to put Cadence pods on a node of a specific instance type and for me it worked multiple times. My specific values were:

cadence:
  server:
    nodeSelector:
      node.kubernetes.io/instance-type: m4.large

Could you provide more details (node description, pod description and helm chart values) so we can take a look what might went wrong for you?