Check the uniqueness of the values of pipeline enum

See merge request gitlab-org/gitlab-ce!23504

CE port Refactor Gitlab::Checks::ChangeAccess class

See merge request gitlab-org/gitlab-ce!23533

Fallback to admin KUBE_TOKEN for project clusters only

See merge request gitlab-org/gitlab-ce!23527

Fix collapsing discussion replies

Closes #54646

See merge request gitlab-org/gitlab-ce!23462

Revert "Merge branch 'auto-devops-support-for-group-security-dashboard' into 'master'"

See merge request gitlab-org/gitlab-ce!23530

CE Port2 of "Add empty and disabled web terminal tab"

See merge request gitlab-org/gitlab-ce!23509

Update -font-size variable name

See merge request gitlab-org/gitlab-ce!23519

Disable format determination based on path extension

Closes #54278

See merge request gitlab-org/gitlab-ce!23454

Make docs about expect_next_instance_of consistent

See merge request gitlab-org/gitlab-ce!23507

This reverts merge request !23165

Approximate table counts based on TABLESAMPLE

See merge request gitlab-org/gitlab-ce!22650

Fix lack of documentation on how to fetch a snippet's content using API

See merge request gitlab-org/gitlab-ce!23448

We do not want group level clusters to fall back to what was old
behaviour for project level clusters. So instead we will not return any
KUBE_TOKEN if we cannot find a suitable kubernetes_namespace for the
project, in the group level cluster case.

Add test cases to assert above

Fix documentation for /api/v4/projects/:id/repository/archive

Closes #45992

See merge request gitlab-org/gitlab-ce!23501

Rollback will not work because Envoy will still fetch latest commit from the…

See merge request gitlab-org/gitlab-ce!23493

Docs : Auto Devops migrate: add two release info

See merge request gitlab-org/gitlab-ce!23500

[ci skip]

The numbers in
https://gitlab.com/gitlab-org/gitlab-ce/merge_requests/22650#note_117645395
are quite promising with that target.

This improves readability quite a bit.

A tablesample count executes in two phases:
* Estimate table sizes based on reltuples.
* Based on the estimate:
  * If the table is considered 'small', execute an exact relation count.
  * Otherwise, count on a sample of the table using TABLESAMPLE.

The size of the sample is chosen in a way that we always roughly scan
the same amount of rows (see TABLESAMPLE_ROW_TARGET).

The goal is to reflect separation of concern also in specs.