- the typo in this factory name was precluding us from properly
creating dynamic code to remove duplciation.

This rake task had been broken for a while. This fixes the breakages,
adds a test to help avoid future breakages, and adds a few ergonomic
improvements to the task itself.

In preparation for embedding specific metrics in issues
https://gitlab.com/gitlab-org/gitlab-ce/issues/62971,
this commit moves the BaseService for metrics dashboards
to a new services subdirectory. This is purely for the sake
of organization and maintainability.

Extends the quick actions "commands applied" banner to show
the quick action preview text, but with everything in past tense.

Otherwise certain machines return IPv6 first, which is non-deterministic

Add a zoom link added / removed system note when a zoom link is being
added / removed to the issue description.

This adds diirect monitoring for sidekiq metrics. This is done via
sidekiq middleware and a sampler to pull from sidekiqs api.

We dropped MySQL support and a lot of mysql specific code has been
removed in https://gitlab.com/gitlab-org/gitlab-ce/merge_requests/29608.

This comes in from the other direction and removes any `if postgresql?`
branches.

Added usage ping counter when the user makes
a search through the navbar search component.

The SanitizationFilter was running before the WikiFilter. Since
WikiFilter can modify links, we could see links that _should_ be stopped
by SanatizationFilter being rendered on the page. I (kerrizor) had
previously addressed the bug in: https://gitlab.com/gitlab-org/gitlab-ee/commit/7bc971915bbeadb950bb0e1f13510bf3038229a4
However, an additional exploit was discovered after that was merged.
Working through the issue, we couldn't simply shuffle the order of
filters, due to some implicit assumptions about the order of filters, so
instead we've extracted the logic that sanitizes a Nokogiri-generated
Node object, and applied it to the WikiLinkFilter as well.

On moving filters around:
Once we start moving around filters, we get cascading failures; fix one,
another one crops up. Many of the existing filters in the WikiPipeline
chain seem to assume that other filters have already done their work,
and thus operate on a "transform anything that's left" basis;
WikiFilter, for instance, assumes any link it finds in the markdown
should be prepended with the wiki_base_path.. but if it does that, it
also turns `href="@user"` into `href="/path/to/wiki/@user"`, which the
UserReferenceFilter doesn't see as a user reference it needs to
transform into a user profile link. This is true for all the reference
filters in the WikiPipeline.

Without this change, we get a NoMethodError on nil

Note that Performance/UnfreezeString recommends unary plus over
"".dup, but unary plus has lower precedence so we have to use
parenthesis

Using the sed script from
https://gitlab.com/gitlab-org/gitlab-ce/issues/59758

This was shown in specs but surely this will be happening in application
code as well if this method is passes a frozen string.

We were also trying to force_encode a OmniAuth::AuthHash which had the
very confusing behaviour of returning nil when it was sent a method that
it did not define. Fix that by only force_encoding a String.

Currently, SidekiqMemoryKiller does not feed worker class name in the
json structured logging. This commit extends the json parameter.

These are not required because MySQL is not
supported anymore

Changes all calls to data_source_exists? to table_exists?
since that is the intent of these calls

* Limits raw requests to 300 per minute and per raw path.
* Add a new attribute to ApplicationSettings so user can change this
value on their instance.
* Uses Gitlab::ActionRateLimiter to limit the raw requests.
* Add a new method into ActionRateLimiter to log the event into auth.log

Related to https://gitlab.com/gitlab-org/gitlab-ce/issues/48717

Signed-off-by: NIstvan szalai <istvan.szalai@savoirfairelinux.com>

https://gitlab.com/gitlab-org/gitlab-ce/issues/58792

- Fix `O(n)` complexity of `append_or_update_attribute`,
  we append objects to an array and re-save project
- Remove the usage of `keys.include?` as it performs `O(n)`
  search, instead use `.has_key?`
- Remove the usage of `.keys.first` as it performs a copy
  of all keys, instead use `.first.first`

A project can have the same `LfsObject` linked with up to three
`LfsObjectsProject` records. Each of these records would be for a
different repository, recorded in the `repository_type` property. The
different repositories at time of writing are "project", "wiki", and
"design". See https://gitlab.com/gitlab-org/gitlab-ee/merge_requests/13894

This change exports the list of `repository_type`s as a JSON mapping of
oid => repository_types, which are imported to recreate the correct
`LfsObjectsProject` records.

https://gitlab.com/gitlab-org/gitlab-ee/issues/11090

Although `flat_map` is equivalent to `map` + `flatten(1)`
(note the level 1) we can apply this same refactoring to all cases.

This will help diagnose the source of excessive I/O from Rugged
calls. To implement this, we need to obtain the full list of arguments
sent to each request method.

None of this code can be reached any more, so it can all be removed

Often live traces are removed even though the archived trace
doesn't exist. This commit checkes the existence strictly.

Add cr remarks

Improve specs according to the review

Fix schema

Add cr remarks

Fix naming

Add cr remarks

To allow project filtering

Prepare summary for accepting multiple groups

Modify deploys group summary class

Add filtering by project name in issues summary

Fix rubocop offences

Add changelog entry

Change name to id in project filtering

Fix rebase problem

Add project extraction

It seems that we missed the backward compatibility support
for profiles in the existing folder.

This commit also fixes some specs to be idempotent
and work in a temporary directory which not always
seems to be the case.

This commit also brings the profile_spec.rb which seems
to be missing.

This makes these tests available for other implementations

Previously, in Omnibus, Workhorse expected to listen via the Redis
shared state cluster for the `workhorse:notifications` publish/subscribe
channel, but the Rails code was using the Sidekiq queue cluster for
this. To fix this inconsistency, we make the Rails code use the
persistent cluster, since we don't want Workhorse to be looking at
anything Sidekiq-related.

This backports remaining import/export spec changes from EE to CE.

This backports all changes made to an import/export fixture file from EE
to CE.

This line is already present further down in the YAML file. In EE the
line is already removed.

PATs are accepted using the OAuth2 compliant header
"Authorization: Bearer {token}" in order to allow for
OAuth requests while 2FA is enabled.

This adds a counter to count page creation, which is reflected in the
usage-data we collect.

The number created is stored in Redis, avoiding DB access.

See https://gitlab.com/gitlab-com/gl-infra/infrastructure/issues/6069

If `GitalyClient#can_use_disk?` returned `false`, it was never cached
properly and led to excessive number of Gitaly calls. Instead of using
`cached_value.present?`, we need to check `cached_value.nil?`.

Closes https://gitlab.com/gitlab-org/gitlab-ce/issues/64802

* Remove `controller` and `action` labels from duration histogram.
* Create a new simple counter for `controller` and `action`.
* Adjust histogram buckets to observe smaller response times.