hey guys, I have a flyteagent pod running, but when I ran a workflow that uses

flytekit.sensor

, the error being thrown is:

RuntimeExecutionError: max number of system retry attempts [4/3] exhausted. Last known status message: failed at Node[n0]. RuntimeExecutionError: failed during plugin execution, caused by: failed to execute handle for plugin [agent-service]: rpc error: code = Unavailable desc = name resolver error: produced zero addresses

the error is thrown from flyte propeller I believe, do you guys have any advice or insights what went wrong? And is there a way to actually test to see if flyteagent is running, or to test the connection from propeller to agent?

is there a way to actually test to see if flyteagent is running

you could run

kubectl get pods

to see if agent is running

get pods show that it’s running, but how do i test the connection? (flytepython310) yaou@yaou-mn1 li-flyte % k get pods flyteagent-789b664d5-48f29 1/1 Running 0 3d20h flyteagent-789b664d5-zh6lg 1/1 Running 0 3d13h

@gray-ram-51379 This is a bit strange, I've only seen this error when there's some error in the `endpoint`configured in your local

config.yaml

. Can you check that what you're using there is correctly resolved by your DNS? I don't think the connection between propeller and the agent uses DNS so it's maybe from your client to flyteadmin

Ok, i think what solved my problem was updating the port name to

flyteagent

in service.yaml:

ports:
  - name: {{ .Values.flyteagent.ports.name }}
    port: {{ .Values.flyteagent.ports.containerPort }}
    protocol: TCP
    targetPort: {{ .Values.flyteagent.ports.name }}

in values.yaml:

ports:
    containerPort: 8000
    name: flyteagent (had to update this, this is the string that worked for me)