Run Pachctl tool in separate cluster

• Run docker image containing the pachctl tool in the container (as published in Dockerfile.pachctl )

kubectl run -it --rm pachctl --image=lenisha/pachctl --generator=run-pod/v1

• To point pachctl tool to API in different cluster use

   pachctl config update context default --pachd-address "xxxxxx:30750"
   pachctl version

Where xxxxxx - is address for the API LoadBalancer exposed for pachd service

• Verify

  pachctl config get context default
  or
  cat ~/.pachyderm/config.json

Access PACHD Dashboard

• Expose dash-lb service kubectl apply -f dash.yaml

• Check IP of LoadBalancer kubectl get svc

• Access Dashboard

http://<dash-host>:8080/app?port=8081

where port parameter points to exposed grpc port for websocket communication

Run GRPCURL

• Dockerfile has also utility called grpcurl to help send the grpc calls, run it to get version using .proto definition

kubectl run -it --rm pachctl --image=lenisha/pachctl --generator=run-pod/v1
# grpcurl -proto version.proto -plaintext  xxxxx:30750 versionpb.API.GetVersion

To install grpcutil locally see Dockerfile

References

pachyderm/pachyderm#3232 https://docs.pachyderm.io/en/latest/pachctl/pachctl_config_update_context.html?highlight=address https://docs.pachyderm.io/en/latest/reference/config_spec.html pachyderm/pachyderm#3466 -