This refactors the AutocompleteController according to the guidelines
and boundaries discussed in
https://gitlab.com/gitlab-org/gitlab-ce/issues/49653. Specifically,
ActiveRecord logic is moved to different finders, which are then used in
the controller. View logic in turn is moved to presenters, instead of
directly using ActiveRecord's "to_json" method.

The finder MoveToProjectFinder is also adjusted according to the
abstraction guidelines and boundaries, resulting in a much more simple
finder.

By using finders (and other abstractions) more actively, we can push a
lot of logic out of the controller. We also remove the need for various
"before_action" hooks, though this could be achieved without using
finders as well.

The various finders related to AutcompleteController have also been
moved into a namespace. This removes the need for calling everything
"AutocompleteSmurfFinder", instead you can use
"Autocomplete::SmurfFinder".

This whitelists all existing places where we use "destroy_all".

This has been tested on .com. So far no errors have been seen.

Closes https://gitlab.com/gitlab-org/gitaly/issues/1310
Closes https://gitlab.com/gitlab-org/gitaly/issues/1246

Acceptance testing done through:
https://gitlab.com/gitlab-org/gitaly/issues/1312

Relatively short AT period, but given its not a hard RPC, nor anything
funky is going on, I felt that this was decent enough to remove the
feature flag.

Closes https://gitlab.com/gitlab-org/gitaly/issues/1243

New translations added to `locale/cs_CZ` will automatically be picked up.

Fix bugs in Gitlab::Template::Finders preventing instances from BaseTemplate.all from loading content

Introduced by https://gitlab.com/gitlab-org/gitlab-ce/merge_requests/20087,
this has been tested on .com now, and is stable.

Closes https://gitlab.com/gitlab-org/gitaly/issues/1286
Closes https://gitlab.com/gitlab-org/gitaly/issues/1233

This way we can globally set group only ee events too.
Use nil to indicate global.

allowing override for EE.

This adds an endpoint to get the common ancestor of 2 refs from the API.

it will decide the method for disable statement_timeout upon
per transaction or per session, based on how it's called.

When calling with a block, block will be executed and it will use
session based statement_timeout, otherwise will default to existing
behavior.

By default statement_timeout will only be enabled during transaction
lifetime, therefore not leaking outside of it.

With `transaction: false` it will set for entire session, but requires
a block to passed. It yields control and cleans up session after block
finishes, also preventing leaking outside of it.

Inline diff comments did not have the proper position, so even though
they had line codes the merge request validation would fail. Now
we cache the line position for each parent comment and use that.

Closes #50052

This changes the BackgroundMigration worker so it checks for the health
of the DB before performing a background migration. This in turn allows
us to reduce the minimum interval, without having to worry about blowing
things up if we schedule too many migrations.

In this setup, the BackgroundMigration worker will reschedule jobs as
long as the database is considered to be in an unhealthy state. Once the
database has recovered, the migration can be performed.

To determine if the database is in a healthy state, we look at the
replication lag of any replication slots defined on the primary. If the
lag is deemed to great (100 MB by default) for too many slots, the
migration is rescheduled for a later point in time.

The health checking code is hidden behind a feature flag, allowing us to
disable it if necessary.

System hooks were not being triggered when projects were archived or
unarchived. Reuse the Projects::UpdateService to automatically
trigger system hooks and increase code reuse.

Resolve "Hashed storage: extend "Enable hashed storage for all new projects" to "for all new and renamed projects""