The rubocop todo files have been regenerated using the command:

`bundle exec rake rubocop:todo:generate`

Nonexistent files and the files with no rubocop offenses have been
removed from the todo list.

The grace period for some cops has been retained as they are failing on
the master branch. The corresponding files for these cops have been
added to their todo files.

The grace period for these cops is safe to remove after 1 week of no
warning in #f_rubocop.
See
https://docs.gitlab.com/ee/development/rubocop_development_guide.html#cop-grace-period.

As part of the squash processes, finalized BBM classes from
squashed migrations can be removed from the codebase, including
each respective spec

Changelog: removed

This changes are not relevant anymore and can
be removed. This commit reverts
https://gitlab.com/gitlab-org/gitlab/-/merge_requests/46444
without DB changes.

Changelog: removed

This is a step towards the feature of supporting the GitLab for Slack
app integration to be configured at group and instance-level.

https://gitlab.com/gitlab-org/gitlab/-/issues/391526

This change adds services and controller support for group-level and
instance-level Slack app installation via OAuth with Slack.

The changes are hidden behind a feature flag.

We should flag the usage of pluck without a proper limit
to avoid database performance degradation.

https://gitlab.com/gitlab-com/gl-infra/production/-/issues/17168

Changelog: changed

Stock ActiveRecord STI works by storing the class name in the database,
which is heavy and unnecessary. EnumInheritance introduces a way to use
STI through model defined enums.

Changelog: other

This commit removes excludes from RuboCop TODOs and rspec_order_todo.yml
for removed/non-existent files.

This commit removes excludes from RuboCop TODOs and rspec_order_todo.yml
for removed/non-existent files.

Related Issue: https://gitlab.com/gitlab-org/gitlab/-/issues/362031

Changelog: fixed

We still have many ignored Rubocop Offenses for
Database/MultipleDatabases in specs. This is a MR
to address validate and fix all of them.

Addressing Issue: https://gitlab.com/gitlab-org/gitlab/-/issues/362031

Changelog: other

We still have many ignored Rubocop Offenses for
Database/MultipleDatabases in specs. This is a MR
to address validate and fix all of them.

Addressing Issue: https://gitlab.com/gitlab-org/gitlab/-/issues/362031

Changelog: other

Merge branch 'secMaster' into 'master'

See merge request gitlab-org/security/gitlab!2945

Changelog: security

Many of the files have been deleted or removed, but
not cleaned up from the Rubocop todos before.

Changelog: other

Removed in https://gitlab.com/gitlab-org/gitlab/-/merge_requests/92395

Adding support for a decomposed database to the Backup rake
tasks gitlab:backup:create and gitlab:backup:restore

Issue: https://gitlab.com/gitlab-org/gitlab/-/issues/364292
Changelog: added

Relates to https://gitlab.com/gitlab-org/gitlab/-/issues/245327

- config/initializers/sidekiq.rb this method is safe for multiple
  databases
- lib/gitlab/database.rb is a legitmate use
- lib/gitlab/database/load_balancing/load_balancer.rb uses methods that
  are safe with multiple databases
- lib/gitlab/database/migrations/observers/query_log.rb also uses a
  method that is safe with multiple databases
- lib/tasks/gitlab/db/validate_config.rake is OK because it's executed
  in the context of looping through each database.

This switches from disconnecting only the ActiveRecord::Base connection
to disconnecting all connections

Addresses the Database/MultipleDatabases offense here.

- config/application.rb. This is safe the methods here are
  multi-database safe
- ee/lib/tasks/geo.rake. This is safe because we are looping though each
  database

Allow allowlist the ActiveRecord::Base.configurations method as it's a
read-only method to get all DB configurations.

Adds support to change table helper to specify which database to create
records in.

These places can be switched to use ApplicationRecord as they
exclusively read, and write non-ci tables

All references has long since been removed

Add a cop to prevent usage of disable_referential_integrity
as this method can lead to data consistency problems and
create unsafe situations where triggers on all tables are
disabled.

Geo does not yet support syncing both primary and ci databases, so all
accesses here refer to the primary database.

Add a Gitlab::Database.all_uncached method which implements
ActiveRecord::Base.uncached behavior that works with multiple
databases.

The new method wraps the given block in an #uncached call for each base
model, so that the behavior will be correct on any database connection.

Add for specified methods to be allowed on ActiveRecord::Base in the
Database/MultipleDatabases cop. Also add `no_touching` as the first
method in the allowlist.

Use the connection from the associated relation class rather than
ActiveRecord::Base when doing an index check.

https://gitlab.com/gitlab-org/gitlab/-/issues/350191

Relates to https://gitlab.com/gitlab-org/gitlab/-/issues/350190

Relates to https://gitlab.com/gitlab-org/gitlab/-/issues/350651

Relates to https://gitlab.com/gitlab-org/gitlab/-/issues/350651

Make the :connection argument for WithLockRetries, by removing the
default of ActiveRecord::Base. This prevents accidental use of the wrong
connection once we begin using multiple databases.

This removes most of the things related to Sherlock.
Dependent gems and docs will be removed separately.

Remove some unneeded todos for the Database/MultipleDatabases cop,
because the listed files no longer violate the rule.

Change the existing BackgroundMigrationWorker to use SharedModel with
the correct connection for running BG migration jobs. Also make the
class and specs generic, in preparation for adding a new worker which
will process jobs specific to the ci database.

This commit moves all RuboCop configuration from the large
`.rubocop_manual_todo.yml` into smaller more focused files within
`.rubocop_todo/**/*.yml`.

This step enables us to split `.rubocop_todo.yml` later which is very
likely needed if we start using `rake rubocop:todo:generate` because it
will produce a huge `.rubocop_todo.yml` (due to much higher "exclude
limit").