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

Flyte

We are setting up flyte in GCP, is ingress and FQDN a mandatory requirement for a production ready setup? can we can run the jobs without having a FQDN?

They are not mandatory, but thr ui uses rest api and cli uses Grpc. If you are using the single binary setup you should be able to
Simply port forward 

i can confirm that you can run flyte on GCP entirely internal to the k8s cluster and port-forward as necessary if registering tasks/workflows or viewing the console.

Also could anyone please refer me to some doc which talks about how a continuous integration and continuous deployment look like for a flyte based ml-workflows ?

<@U04MG1WLA4T> did you ever find a solution to this?  Running into the same problem.

<@U04P6HHMCG0> I was able to utilise the port forwarding inorder to make flyte cli work with remote cluster. For UI console port forwarding didn't help I started using dns

Thanks <@U04MG1WLA4T>.  I did the same with DNS.  Thanks for confirming.

&gt; I was able to utilise the port forwarding inorder to make flyte cli work with remote cluster
I have been struggling to get `flytectl` to work with the DNS name and the ingress so I’m going to try this instead since I just got it working on the simple `flyte-binary` deployment.

I also faced error connecting flytectl with dns host, for now I am skipping certificate verification by adding below config in my config.yaml file to make it work
```admin:
 # For GRPC endpoints you might want to use dns:///flyte.myexample.com
 insecure: false
 insecureSkipVerify: true```