[Cluster launcher] Add more commands to "Useful commands" CLI output

[architkulkarni]

Why are these changes needed?

Updates the "useful commands" printed on ray up to add

• ray down
• ray get-head-ip
• ray dashboard
• ray job submit

Removes the "manual ssh command", because "ray attach" should be used instead.

In summary, this PR implements the conclusion of the discussion here: #34685 (comment)

This PR also fixes a bug where cluster-name was not printed in the example commands in places where it should have been. (This is stored in the variable modifiers in the code)

Related issue number

Closes #34685

Checks

• I've signed off every commit(by using the -s flag, i.e., git commit -s) in this PR.
• I've run scripts/format.sh to lint the changes in this PR.
• I've included any doc changes needed for https://docs.ray.io/en/master/.
  • I've added any new APIs to the API Reference. For example, if I added a
    method in Tune, I've added it in doc/source/tune/api/ under the
    corresponding .rst file.
• I've made sure the tests are passing. Note that there might be a few flaky tests, see the recent failures at https://flakey-tests.ray.io/
• Testing Strategy
  • Unit tests
  • Release tests
  • This PR is not tested :(

[scottsun94]

2 high-level questions:

1. There are two pieces of ray job submission info and they look different..

2. The output in the "next steps" section seems to use the local ip. The commands that use the local ip won't work right?

3. If ray dashboard is not installed, will we hide the related output?

[architkulkarni]

2 high-level questions:

1. There are two pieces of ray job submission info and they look different..

2. The output in the "next steps" section seems to use the local ip. The commands that use the local ip won't work right?
3. If ray dashboard is not installed, will we hide the related output?

1. That's a separate issue, right now the output from the head node (which is the output of running ray start --head) is streamed to the user laptop (the one running ray up). The output from the head node lists commands which only make sense on the head node. We can add a message in the output to clarify this in a future PR.
2. Yup, fixed it
3. Both cluster launcher and dashboard require ray[default] so I don't think this situation will happen.

[scottsun94]

That's a separate issue, right now the output from the head node (which is the output of running ray start --head) is streamed to the user laptop (the one running ray up). The output from the head node lists commands which only make sense on the head node. We can add a message in the output to clarify this in a future PR.

Yes. That makes sense. We need a better solution here. It's really confusing. Can we create an issue to track it? cc: @jjyao

Yup, fixed it

Sorry. What did you fix? It seems that this was the same issues as the one above.

Both cluster launcher and dashboard require ray[default] so I don't think this situation will happen.

Got it. Thanks!

[architkulkarni]

That's a separate issue, right now the output from the head node (which is the output of running ray start --head) is streamed to the user laptop (the one running ray up). The output from the head node lists commands which only make sense on the head node. We can add a message in the output to clarify this in a future PR.

Yes. That makes sense. We need a better solution here. It's really confusing. Can we create an issue to track it? cc: @jjyao

I will create an issue and link it here.

Yup, fixed it

Sorry. What did you fix? It seems that this was the same issues as the one above.

I was referring to this commit: 18398de So for this command, the screenshot in the description is slightly out of date (it predates this commit). Basically, it is correct to write "localhost" because we instructed the user to do port-forwarding first.

[architkulkarni]

@scottsun94 added the issue! #40833

[scottsun94]

LGTM. Thanks!

[hongchaodeng]

LGTM!