Ensure Issues::BuildService is findable on reload from issues controller

See merge request gitlab-org/gitlab!40535

Update index.md make sure file `SAST.gitlab-ci.yml` correct location

See merge request gitlab-org/gitlab!40240

Update docs around conan search command

See merge request gitlab-org/gitlab!40427

Amplify warning to DAST docs about host override

See merge request gitlab-org/gitlab!40442

Move tags to the same describe block

See merge request gitlab-org/gitlab!40466

[RelativePositioning] Add rebalancing worker for issues

See merge request gitlab-org/gitlab!40124

Docs: Document Real-Time sidebar as production-ready

See merge request gitlab-org/gitlab!40500

Docs: clarify how to enable Squash and Merge

Closes #220347

See merge request gitlab-org/gitlab!40258

Remove error/warning logs from Webpack manifest

See merge request gitlab-org/gitlab!33918

This is much better from a performance point of view - batches of 500
take a long time (55sec) but we can run batches of 100 in under 60ms.

Revise note on token expiration and enforcement

See merge request gitlab-org/gitlab!40527

Create IssueLink for Vulnerabilities that do not have them

See merge request gitlab-org/gitlab!39986

As mentioned in https://gitlab.com/gitlab-org/gitlab/-/issues/224457 we
need to create missing `Vulnerabilities::IssueLink` objects for
Vulnerabilities that do not have them yet but have an associated
`Vulnerabilities::Feedback` with type of `issue`.

Resolve "Clean up `pageLoading` computed in `insights.vue`"

Closes #235733

See merge request gitlab-org/gitlab!40096

Do not disable the dropdown when no charts have
been requested

Also, simplify the `pageLoading` logic

Use new SAST variable for SAST images

See merge request gitlab-org/gitlab!39875

This is justified because even at the MAX_ISSUE_COUNT, we would only
ever use 0.1% of the full range width by assuming this value. Extremely
large projects such as gitlab-org/gitlab have on the order of around
100k issues, and would only need 1% of the full range.

We can thus assume such a value as gap_size even allowing for 10x growth
in gitlab's issue numbers, at which point managing locking and
concurrent updates during rebalancing would be a much more challenging
problem.

This moves the issue presence check to the first place (as it should be)
and doesn't check for feature availability until we know we need to.

This involves a refactoring of the service so that it is less of a
single procedural thicket.

Safer use of flags, by checking them in the service and the worker.

Allow groups to be used as feature flag gates.

This adds a new feature flag: `rebalance_issues`

This flag is scoped by project, and is read in the issue create and
update services.

This adds a new method in the issues base_service that rebalances the
issue set when the issue is placed in the extreme 0.0001 fringes of the
relative positioning range.

This adds the required worker queue definition for sidekiq, with its
configuration and generated defintion.