Update vulnerability metric to use constant

See merge request gitlab-org/gitlab!81080

Issue: https://gitlab.com/gitlab-org/gitlab/-/issues/352915

Track work item title updates

See merge request gitlab-org/gitlab!80532

Break out Sidekiq process supervision into library module

See merge request gitlab-org/gitlab!80617

Add mutation to promote timeline event from a note

See merge request gitlab-org/gitlab!80633

Fix Vale errors in Service Desk doc

See merge request gitlab-org/gitlab!81151

Remove fork_project_form feature flag

See merge request gitlab-org/gitlab!77181

Simplify container repository delete tags services

See merge request gitlab-org/gitlab!81187

Avoid idling in transaction when fetching source for merge requests

See merge request gitlab-org/gitlab!80876

Updated elastic.md docs

See merge request gitlab-org/gitlab!81234

Replaced the fork form paths with provide/inject

Upgrade golang to 1.17.7

See merge request gitlab-org/gitlab!81208

E2E: Fix incorrectly assigned failures object in allure metadata

See merge request gitlab-org/gitlab!81231

Resolve "Resize observer visible in dark mode"

See merge request gitlab-org/gitlab!81117

Add feature flag tests to the definition of done

See merge request gitlab-org/gitlab!81217

Fix GraphQL/OrderedArguments offense (Part 5/6)

See merge request gitlab-org/gitlab!76027

Fix GraphQL/FieldDefinitions offense (Part 4/4)

See merge request gitlab-org/gitlab!76017

Fix GraphQL/OrderedArguments offense (Part 5/6)

See merge request gitlab-org/gitlab!76027

Previously when a merge request was created, the `after_create` Rails
callback would fire the `FetchSourceBranch` Gitaly RPC. However, this
call is made inside a database transaction, which may have an idle in
transaction timeout.

On GitLab.com, this timeout is set to 30 seconds, so any
`FetchSourceBranch` calls taking longer than 30 seconds causes the
database connection to be terminated quietly. When Rails attempts to
make another database write (e.g. writing the metrics), the query
fails with a 500 error after the RPC completed, usually because a
subsequent query ran into an error.

Instead of failing after the idle in transaction, this commit will
cause the query to be killed if the total Web request time expires (60
seconds or so). While this isn't ideal, this is better than the
alternative because we should never hold open a database transaction
for a long Gitaly call. Gitaly team has a few optimizations for
`FetchSourceBranch` in the works, but allowing this RPC to run for up
to 60 seconds during a Web request is better than the status quo.

Relates to https://gitlab.com/gitlab-org/gitlab/-/issues/336657

Changelog: fixed

Clean up forcibly_show_trial_status_popover experiment

See merge request gitlab-org/gitlab!80413

forcibly_show_trial_status_popover experiment has been called unsuccessful
and should be removed and returned back to control behavior

Changelog: removed

Added support for level 3 in widget extensions

See merge request gitlab-org/gitlab!80486

Add development seeds for triage ops

See merge request gitlab-org/gitlab!80751

Track security training provider clicks

See merge request gitlab-org/gitlab!80864

Remove link to outdated demo video

See merge request gitlab-org/gitlab!81225

Clearly state that 5gb namespace limit is not yet shown on page

See merge request gitlab-org/gitlab!81115

Add API endpoint to delete topics

See merge request gitlab-org/gitlab!80725

Switched title to h1 on issues

See merge request gitlab-org/gitlab!75248

Changelog: changed

Changes in Atlassian OmniAuth Provider

See merge request gitlab-org/gitlab!81152

Topics: order by `non_private_projects_count`

See merge request gitlab-org/gitlab!80478

Updates for https://gitlab.com/gitlab-org/gitlab/-/issues/353232

See merge request gitlab-org/gitlab!81155

Upgrade Elasticsearch in CI to 7.17.0

See merge request gitlab-org/gitlab!80993