Correctly caches log tree calls in Apollo client

See merge request gitlab-org/gitlab!19382

Display the IID of a Feature Flag

See merge request gitlab-org/gitlab!19046

Initialize EKS clusters form for instance and group level clusters

See merge request gitlab-org/gitlab!18690

We are initializing the form to create K8 clusters in three places

- Project level clusters
- Group level clusters
- Instance level clusters

This commit moves duplicated initialization code into a single
fn

Support Read and Delete operations in Feature Flag Scope API

See merge request gitlab-org/gitlab!18200

Docs - Flesh out Praefect documentation

See merge request gitlab-org/gitlab!19418

Docs: fix a typo

See merge request gitlab-org/gitlab!19497

Add a note about feature not being available yet

See merge request gitlab-org/gitlab!19225

- Since the self-monitoring project was not released in the planned
milestone, but the documentation did get released, update the docs now
to note that the feature is not yet available.

Backend: Add checkbox to enable/disable grafana authentication

See merge request gitlab-org/gitlab!19234

In example use "remove_concurrent_index" instead of "remove_index"

See merge request gitlab-org/gitlab!19306

Fix the CI config for the 'review-stop' job

See merge request gitlab-org/gitlab!19459

Merge branch '35337-images-stopped-loading-signed-urls-of-uploads-expire-earlier-than-redirect-cache' into 'master'

Resolve "images stopped loading: signed URLs of uploads expire earlier than redirect cache"

See merge request gitlab-org/gitlab!19494

Expose cycle analytics summary separately

See merge request gitlab-org/gitlab!19229

This change exposes the summary counts for group level cycle analytics
via a separate endpoint while making sure the old CA implementation is
still working.

This allows the flag to be referenced in other portions of the app
(coming soon), such as issues, MRs and epics.

Setting a long-term cache header works fine when not using object
storage, or when using object storage with proxy_download enabled. When
using object storage _without_ proxy_download, we send a redirect that's
cached for six months to a signed URL in object storage, but the signed
URL only works for ten minutes. So once those ten minutes are up, the
browser will keep following the (cached) redirect to a dead link.

In future, we could be more sophisticated by setting the TTL based on
the settings we're using. For now, just set this back to the pre-12.4
state of not caching content for longer than 5 minutes.

Add guard to avoid misleading error message

See merge request gitlab-org/gitlab!19487

Update plantuml.md with URI rewrite for nginx

See merge request gitlab-org/gitlab!19163

Revert "Merge branch 'mc/backstage/needs-canonical-form' into 'master'"

See merge request gitlab-org/gitlab!19479

Update praefect git_data_dirs setup

Closes gitaly#2111

See merge request gitlab-org/gitlab!19040

Add API for suggesting unique alternatives if username exists

See merge request gitlab-org/gitlab!19303

This commit adds public API support for FF scope

This reverts merge request !17539

Replace end-to-end test by rspec test

Closes #33549

See merge request gitlab-org/gitlab!19445

Extend gRPC timeouts for Rake tasks

Closes #22398

See merge request gitlab-org/gitlab!19461

Edited notifications docs

See merge request gitlab-org/gitlab!19341

* Edited and restructured notification docs to improve
clarity.
* Issue: https://gitlab.com/gitlab-org/gitlab/issues/31710

Remove non-api tests from api spec

Closes #32903

See merge request gitlab-org/gitlab!19219

Fix file locking spec and quarantine locks directory spec

See merge request gitlab-org/gitlab!19380

Test was quarantined due to a bug: https://gitlab.com/gitlab-org/gitlab/issues/35254
Changed hardcoded "Adminstrator" in expectation to a username

In https://gitlab.com/gitlab-org/gitlab/merge_requests/16926 we added
gRPC timeouts for calls that did not previously have timeouts to prevent
Sidekiq queries from getting stuck. In addition, we also made long
timeouts 55 seconds for non-Sidekiq requests, but this meant Rake tasks
also fell into this bucket. Rake backup tasks with large repositories
would fail because the CreateBundle RPC would time out after 55 seconds.

To avoid this trap, we flip the logic of long_timeout: instead of
checking for Sidekiq (or other background jobs), we only lower the
timeout to 55 seconds if we're servicing a Web request in Puma or
Unicorn.

Closes https://gitlab.com/gitlab-org/gitlab/issues/22398

Update github_integration.md

See merge request gitlab-org/gitlab!19452