Update Mimir Prometheus to baf7468

[gotjosh]

What this PR does

The whole list of changes is included as part of Mimir Prometheus PR in grafana/mimir-prometheus#622, grafana/mimir-prometheus#623 and grafana/mimir-prometheus#625

I don't think there's any change that's worth mentioning, given that they aren't user-facing but do let me know if you think there's something worth adding to the changelog.

As part of the second mimir-prometheus PR there is a change that I have noted in the changelog. Matrix results were returning null instead of [] in some cases.

I have managed to reproduce this by running:

# Main branch
$ ~> curl "http://localhost:61793/prometheus/api/v1/query_range?query=bottomk(2,%20notExists)&start=1714745905&end=1714745910&step=30s" \
     -H 'X-Scope-OrgID: tenant-1' | jq
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100    65  100    65    0     0  24780      0 --:--:-- --:--:-- --:--:-- 32500
{
  "status": "success",
  "data": {
    "resultType": "matrix",
    "result": null
  }
}


# This PR

$ ~> curl "http://localhost:61235/prometheus/api/v1/query_range?query=bottomk(2,%20notExists)&start=1714745710&end=1714745715&step=30s" \
     -H 'X-Scope-OrgID: tenant-1' | jq
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100    63  100    63    0     0   6039      0 --:--:-- --:--:-- --:--:--  6300
{
  "status": "success",
  "data": {
    "resultType": "matrix",
    "result": []
  }
}

Against a querier in both main and this branch.

Which issue(s) this PR fixes or relates to

Related to https://github.com/grafana/mimir-squad/issues/1845

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.

[pracucci]

LGTM. I've reviewed mimir-prometheus changes, and checked protobuf release note (I can't see anything suspicious there).

[gotjosh]

I'm investigating the CI failure as is very relevant to the CI changes - looks like it's failing to execute the restoration because it does not return any samples when querying the ALERTS_FOR_STATE metric.

[gotjosh]

I think this is a typo all the others are capitalised.

[pracucci]

It's OK editing it, but we're not very strict with the CHANGELOG. Anyway, it's good you improved it.

[gotjosh]

The CI failures are real; they relate to prometheus/prometheus#14047, which is going to be shortly merged.

What are my options here? I could technically fix it in mimir-Prometheus very quickly and re-import or just wait?

I'd like this PR to be merged today so that this gets rolled out next week 😭

[pstibrany]

I'd like this PR to be merged today so that this gets rolled out next week 😭

It looks like that PR will indeed be merged shortly. I'd suggest to wait for it and if it's not merged by monday, do the fix in mimir-prometheus. We can update our weekly release after it's been cut.

[ArthurSens]

Looks like that PR is blocking your work, I'm just going to add the missing test case to the contributor's PR. Do you want to add a review over there? :)