service: allow per-service configuration of protocol (mongodb:// for example)

[joeapearson]

This is a feature request. In summary I'd like to be able to configure the output of minikube service to allow more accurate information to be displayed for our developers.

Environment:

Minikube version (use minikube version): v0.27.0

• OS: macOS 10.13.4
• VM Driver: virtualbox
• ISO version: v0.26.0
• Install tools: n/a
• Others: n/a

What happened:

Using mongodb as an example, but this applies to any external service you'd like to expose that isn't http.

1. Create a mongodb deployment
2. Create a mongodb service
3. Now list your running services as follows:

$ minikube service list --namespace default
|-------------|----------------------|--------------------------------|
|  NAMESPACE  |         NAME         |              URL               |
|-------------|----------------------|--------------------------------|
| default     | mongodb              | http://192.168.99.102:30321    |
|-------------|----------------------|--------------------------------|

What you expected to happen:

It'd be great if there were a way of configuring the service endpoint to display as mongodb://<ip>:<port> and be configurable on a service-by-service basis.

Very happy to submit a pull request for this although I could use some help in identifying an appropriate place to add it since this would be my first contribution.

[dlorenc]

Can you get this to work using the --format flag to "minikube service"?

[joeapearson]

@dlorenc that would only work for a single service though I believe. If you had, for example, an http service and a mongo service running you'd want differing schemes for both.

[dlorenc]

Hmm, is there any way to detect from the service itself what protocol to use? I don't think it would be a very good idea to force users to configure this themselves.

[joeapearson]

Agreed that forcing users to configure it adds extra complexity. But don't you think an option would be fine?

Possibly could use the service port name, although that wouldn't fare so well for services that expose more than one endpoint of the same scheme.

For example, this abbreviated service config:

apiVersion: v1
kind: Service
spec:
  ports:
  - name: mongodb
    port: 27017
  - name: https
    port: 443
  - name: http
    port: 80

…would yield:

$ minikube service list
|-------------|----------------------|--------------------------------|
|  NAMESPACE  |         NAME         |              URL               |
|-------------|----------------------|--------------------------------|
| default     | https                | https://<cluster ip>:<port>    |
| default     | http                 | http://<cluster ip>:<port>     |
| default     | mongodb              | mongodb://<cluster ip>:<port>  |
|-------------|----------------------|--------------------------------|

[dlorenc]

Maybe I'm misunderstanding - for a single service example like you showed, you can do that with the --format flag. Does that make sense?

[joeapearson]

Updated my earlier comment to make it a bit clearer I hope

[nomicode]

I have a service configured to expose two ports. one uses HTTP and the other uses the PostgreSQL wire protocol. so I am running into this issue too

instead of adding something to Minikube, would it make sense to change the column name to NETWORK ADDRESS and drop the http:// prefix?

the http:// is not required even for HTTP services, because the HTTP protocol is inferred by every HTTP client I am aware of

[tstromberg]

This appears somewhat related to #3112

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[fejta-bot]

Stale issues rot after 30d of inactivity.
Mark the issue as fresh with /remove-lifecycle rotten.
Rotten issues close after an additional 30d of inactivity.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle rotten

[fejta-bot]

Rotten issues close after 30d of inactivity.
Reopen the issue with /reopen.
Mark the issue as fresh with /remove-lifecycle rotten.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/close

[k8s-ci-robot]

@fejta-bot: Closing this issue.

In response to this:

Rotten issues close after 30d of inactivity.
Reopen the issue with /reopen.
Mark the issue as fresh with /remove-lifecycle rotten.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.