EE: true

In
https://ops.gitlab.net/gitlab-com/database-team/gitlab-com-database-testing/-/jobs/17362259
we can see that after truncating our migration names for Housekeeper we
are getting duplicate file names. This is because we have the following
long migration names that share the same prefix:

```
BackfillProtectedEnvironmentApprovalRulesProtectedEnvironmentProjectId
BackfillProtectedEnvironmentApprovalRulesProtectedEnvironmentGroupId
```

In order to work around this problem we should ensure that our truncated
names always end up unique. We do this by taking a hash of the file
name before truncating and appending this 5 digit number to the end of
the file name. This should be a consistent hash so that each time we run
Housekeeper we get the same name.

This reverts commit a3720f8b, reversing
changes made to 5e3f7661.

We will log response sizes over the GITLAB_LOG_DECOMPRESSED_RESPONSE_BYTESIZE if set

This commit fixes this keep when a migration has already be marked
as finalized and does not create a new migration file for it.

Moves the `Gitlab/DocUrl` cop to the
`Gitlab/DocumentationLinks/HardcodedUrl` namespace.

This commit updates our rubocop configuration file automatically
when running UpdateTableSizes keep. This way new table identified
as large or over_limit will be prevented from adding new indexes
or new columns to them.

We use the reviewer roulette when generating merge requests using
housekeeper. Here we fix a bug where the role like `maintainer::backend`
was being picked for another project like `customers-app`. Instead, we
should be picking `maintainer::backend` for the `gitlab` project.

This commit adds a new keep to automatically updates table_size classification.
It will be used in DB Lab clone to automatically update the table sizes
in our database dictionaries.

In the GitLab monolith codebase, we should use the `Gitlab::HTTP`
wrapper instead of `Gitlab::HTTP_V2` because we need to pass options
that are related to the GitLab instance setting.

This commit adds a new keep which will help to automate the conversion
of our remaining integer columns to bigint.

It adds a new Housekeeper Keep to update workers data
consistency from always to sticky. These changes aim
to help with database resource saturation.

https://gitlab.com/gitlab-org/gitlab/-/issues/462611

Changelog: changed

Previously the keep would return changes sorted alphabetically by name
of migration file. This change returns the changes by most to least
overdue.

This commit introduces a new method to get upcoming milestones from
a given date. This will be used to schedule migration in n+3 milestone.

Replace usages of MIN_SCHEMA_GITLAB_VERSION with data sourced
from the upgrade_path.yml file.

Related to: https://gitlab.com/gitlab-org/gitlab/-/issues/481816

We're using the housekeeper gem here to generate the `.rubocop_todo`
files and create an MR.

Run `bundle exec gitlab-housekeeper -d -k Keeps::GenerateRubocopTodos`
to test out the changes locally.

Reduce branch name length
Refactor description and link
Divy MRs evenly to team members

Signed-off-by: Rémy Coutable <remy@rymai.me>

Signed-off-by: Rémy Coutable <remy@rymai.me>

Signed-off-by: Rémy Coutable <remy@rymai.me>

Signed-off-by: Rémy Coutable <remy@rymai.me>

Signed-off-by: Rémy Coutable <remy@rymai.me>

Signed-off-by: Rémy Coutable <remy@rymai.me>

Add gdk start db to dry run instruction

Update MR description

It generates migration to remove indexes as post migrations. It
also remove the duplicated index from duplicate indexes list

Changelog: changed