Flyte enables production-grade orchestration for machine learning workflows and data processing created to accelerate local workflows to production.

Flyte

Anyone have a good way of testing the grpc end points (maybe with grpcurl) to debug networking issues? I upped the admin logging level, and hoping to see more logs to see if we're hitting admin correctly but can't tell if there's just no logs or if we're not hitting it :sweat_smile:

flytectl uses the grpc endpoints under the hood so something like `flytectl get projects` should be a simple test

Something like ...
`flytectl get projects --logger.level=6 --admin.insecure  --admin.endpoint dns:///host:81`  or `flytectl get projects --logger.level=6 --admin.insecure  --admin.endpoint {ip}:81`

And if I have logging level at 20  in my admin config I should be able to see admin logging requests right? I just want to make sure since the logs don't seem to outputting anything (which would suggest we're not hitting it)

if you have auth enabled then the <https://github.com/flyteorg/flyteadmin/blob/26e52192570aa9b6e9de6cd9b36af8f9770ae8e4/pkg/server/service.go#L87|logging interceptor> should <https://github.com/flyteorg/flyteadmin/blob/e32b2ab3a5d3c1e02052ad1106ca2d28cdc127dd/auth/handlers.go#L193|log> on every endpoint

what if I do not have auth enabled :disappointed:

I was able to hit `get projects` but dont have logs for it (without auth)

you could try setting the grpc verbosity env var: <https://github.com/grpc/grpc/blob/master/doc/environment_variables.md>

I did! The logging shows with pyflyte but not flytectl

where are you setting that env var? can you set it for the admin service deployment?

Ah, good call, I was setting it on the client, I'll do that