Unify mimirtool authentication options and add extra-headers support for commands that depend on MimirClient

[LeszekBlazewski]

What this PR does

1. Unifies all authentication options for mimirtool commands so that the Configuration options work the same for subcommands which use MimirClient.
2. Adds extra headers support to all commands from mimirtool which depend on MimirClient so that any supported command can be used like in the use case that was described in Mimirtool: add the ability to include extra headers with requests made to the server #7141.
3. Adjusted the Prometheus api client for analyse prometheus command so that its auth behaves similarly to other mimirtools commands.

Which issue(s) this PR fixes or relates to

There is a discussion opened by my team member: #10152 (comment) and I have created an issue from it: #10178. This PR addresses the issues he is facing.

Related PRs to the changes I am proposing:

• Mimirtool: add the ability to include extra headers with requests made to the server #7141
• mimirtool: clarify what auth flags do #6798
• Adds bearer token support for mimirtool's analyze ruler/prometheus co… #9587
• chore: mimirtool prometheus analyze set tenant header #6737

Thanks for making such a great product and all the tooling around it! Working with mimir and mimirtool is a breeze and helps a lot with our monitoring setup. We have discovered the above limitations during optimising our stored metrics in mimir. For us the basic auth username configured on the nginx proxy that is in front of mimir does not match the tenant name to which we push metrics and that prevents us from using some of the subcommands but others work and with that PR all of them should be unified. The proposed changes should be backwards compatible so anyone using mimirtool shouldn't be forced to change the passed ENVS / flags.

Checklist

• Tests updated.
• Documentation added.
• CHANGELOG.md updated - the order of entries should be [CHANGE], [FEATURE], [ENHANCEMENT], [BUGFIX].
• about-versioning.md updated with experimental features.

[CLAassistant]

All committers have signed the CLA.

[chencs]

The CHANGELOG has just been cut to prepare for the next release. Please rebase main and eventually move the CHANGELOG entry added / updated in this PR to the top of the CHANGELOG.md document. Thanks!

[LeszekBlazewski]

@chencs Rebased PR and adjusted changelog as requested.

[LeszekBlazewski]

I am not sure from whom should I request a review here and it does not seem possible to tag @grafana/mimir-maintainers team as a whole. Maybe @dimitarvdimitrov you could take a look as I have seen some of your work on mimirtool?

If there is anything missing/wrong with this PR, please let me know.

[colega]

Thank you for your contribution, this looks very good! I just left a small nit about the changelog entry, waiting for confirmation there before merging.

[colega]

Thank you.

[LeszekBlazewski]

@colega it looks like before the CI checks can run, they need to be approved by a maintainer: source

By default, all first-time contributors require approval to run workflows.

[colega]

Oh, but I approved the PR. Now I see what happened, I'm trying "the new merge experience" and it seems that the approval button is missing in it:

New:

Classic:

[LeszekBlazewski]

@colega I checked the CI logs and I don't see how my changes could fail the TestInstantQuerySplittingCorrectness test . The test passes locally. On my local fork that test also failed but after I have rerun it, it passed OK: source. Is there any chance that those tests could be flaky and CI run dependant? Or there is indeed something off with the changes I am proposing? It seems I am not able to trigger a rerun for the failed ci / test (2,4) job.

I have rebased the PR with main and force pushed which disabled the auto merge and it looks like CI needs another approval - sorry for that.

[colega]

Hey, yes, TestInstantQuerySplittingCorrectness is known to be flaky, there's an issue with a long discussion about how to tackle it, so I'll just restart the build if it fails again. Sorry for the inconvenience.

[LeszekBlazewski]

I think that the new merge experience got in a way again 😅 since the CI workflow is not running and must be approved again.