Add weekly and monthly config files for
deploy token metrics

Changelog: other

This change moves github collaborators import
out of an advance worker stage and before pull
requests are imported.

Changelog: changed

We have some legacy database entries in `ci_trigger_request.variables`
that serialized values as `Hashie::Mash`. This results in a 500 error
when attempting to load old jobs. Allow this class to be deserialized
since it's effectively a `Hash`.

Relates to https://gitlab.com/gitlab-org/gitlab/-/issues/440316

Changelog: fixed

The group 'memory' was renamed to 'application_performance'

'environment' is a typo. It should be 'environments'

MR: gitlab.com/gitlab-org/gitlab/-/merge_requests/143077

Changelog: other

Changing `/-/profile` -> `/-/user_settings/profile`

Changelog: other

https://gitlab.com/gitlab-org/gitlab/-/issues/231317

Changelog: fixed

Changelog: removed

The group 'import' was renamed to 'import_and_integrate'

The group 'import' was renamed to 'import_and_integrate'

The group 'import' was renamed to 'import_and_integrate'

This completes the rollout of
`merge_request_pipelines_list_disable_coverage`. When rendering pipeline
lists on merge requests, we were serializing unused coverage data. This
is expensive, especially for pipelines with lots of downstream
pipelines, as each coverage report iterates over all builds in a
pipeline, and each pipeline has its own coverage data.

To be extra safe, we rolled this out behind a feature flag. This flag
has been live on GitLab.com now for over 24h without issues. Since this
is a very high traffic page the flag should be safe to remove now.

Changelog: performance

Add migration to create ci_job_token_group_scope_link
Add ci_job_token_group factory
Add loose foregin keys for ci_job_token_group_scope_links
Inject target group validation to allowlist

Changelog: added

This will direct all AI gateway traffic to go through
cloud.gitlab.com/ai instead of codesuggestions.gitlab.com.

Rollout issue:
https://gitlab.com/gitlab-com/gl-infra/production/-/issues/17405

Changelog: changed

https://gitlab.com/gitlab-org/gitlab/-/merge_requests/88882 introduced
a monkey patch for `Enumeration#next` to improve stack traces for
Gitaly calls due to https://bugs.ruby-lang.org/issues/16829.

This patch relies on checking that `gitlab_patch_backtrace_marker` is
present in the stack trace. However, for a Rust extension like
`prometheus-client-mmap`, the Ruby interpreter omits the details of
the calls made inside the extension. For example, in
https://gitlab.com/gitlab-org/omnibus-gitlab/-/issues/8373#note_1748497473,
we see `StopIteration` is raised when the extension iterates through
all strings in its `WeakMap`, but the backtrace does not contain the
expected signature since `Enumerator#next` is called from the Rust
code. As a result, the backtrace would always be unnecessarily
appended to the `StopIteration` exception, resulting in millions of
wasted memory allocations.

This commit reduces memory allocations in two ways:

1. Use a constant for the regular expression. This eliminates the need
to call `parse_regexp` in the Ruby interpreter repeatedly each time
`Enumerator#next` is called.

2. Don't append stack traces for `StopIteration`. A Enumerator raises
`StopIteration` when the end is reached, and `Kernel#loop` rescues this
exception. Previously a lot of care went into ensuring that nested
enums preserved the full backtrace when `StopIteration` occurred, but
this generally isn't needed for debugging. If a C or Rust extension
iterates through an array repeatedly, we saw millions of temporary
objects being created to build the complete stack trace for no real
use.

Relates to https://gitlab.com/gitlab-org/gitlab/-/issues/439920

Changelog: performance

Those CI/CD templates will be removed as soon as CI/CD components
are available to self-managed customers.

Changelog: added

https://gitlab.com/gitlab-org/gitlab/-/issues/429557

This feature flag enables improving canceling redundant pipelines
with `interruptible`. It's been stable on .com.

Changelog: added

This commit introduces a ff which will allow us to read
Ci::PipelineVariables from our new partitioned table.

Changelog: performance

Allows user to create a group in an organization from the UI

Switching to event based metrics to track pipeline creation

Pass pending options via the callers

Read from application settings where needed

Update specs to reflect the change

Fix rubocop offense

Add reason for inline disabling of cop

Add setting as reason

Fix rubocop offense

Add reason for inline disabling of cop

Fix other cop offenses

Fix wrong key for dns rebind protection

Remove silent enabled option

Remove allow_local_requests key

Fix failing specs for UrlBlocker change

Remove dns_rebind_protection wherever applicable

This is a gitlab.com-only migration since QueuesMetadata config fallback
is Queues.

Changelog: other

This reverts merge request !143118

https://gitlab.com/gitlab-org/gitlab/-/merge_requests/143285

Changelog: changed

This commit adds a worker which enables
syncing of audit events from PostgreSQL to
ClickHouse database

Update sidekiq queue yml

Add migration to create ci_job_token_group_scope_link
Add ci_job_token_group factory
Add loose foregin keys for ci_job_token_group_scope_links
Add sharding_key

The following sorting options are supported:

1. storage_size_desc
2. latest_activity_desc
3. latest_activity_asc
4. path_asc
5. path_desc
6. stars_desc
7. stars_asc
8. created_asc
9. created_at_asc
10. created_date
11. created_desc
12. created_at_desc
13. id_asc
14. id_desc
15. name_asc
16. name_desc
17. updated_asc
18. updated_at_asc
19. updated_desc
20. updated_at_desc

Changelog: added

Changelog: added
EE: true

Changelog: added
EE: true

Currently, Environment#stop_actions only picks up the stop jobs of
the latest_successful_deployments in the latest_successful_pipeline.
This changes Environment#stop_actions to pick up the stop jobs of
the latest_finished_deployments in the latest_finished_pipeline.

Changelog: changed

Changelog: changed

Added a new rate limit and logging on downstream
pipelines triggered by Ci::Bridge jobs. The change is behind
a Feature Flag 'ci_rate_limit_downstream_pipelines'.
This MR also adds a new service class for triggering
downstream pipelines from a bridge job.