| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
| |
When `needs:` is defined, the value of it is
not respected when returning a list of artifacts
to the runner from the job.
|
|
|
|
|
|
|
| |
Currently we write out empty CSS classes (`class=""`) every time we
create a new tag. This adds 9 unnecessary bytes per span element. In a
recent trace, I have counted 11950 span elements. So we transported 105
unnecessary kilobytes!
|
|
|
|
|
|
| |
See https://gitlab.com/gitlab-com/gl-infra/production/issues/1023
This also has been suggested by Tomasz in the MR
|
| |
|
|
|
|
|
|
|
|
| |
`detailed_metric.vue` was attempting to read the `duration` field even
when the metric did not exist in the JSON payload. We now check for the
existence of the metrics before attempting to display the value.
Closes https://gitlab.com/gitlab-org/gitlab-ce/issues/65491
|
| |
|
| |
|
| |
|
|
|
|
|
| |
- Remove setTimeout calls
- Add helper function to encapsulate setup
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously, both InfluxSampler and RubySampler were relying on the
`GC::Profiler.total_time` data which is the sum over the list
of captured GC events. Also, both samplers asynchronously called
`GC::Profiler.clear` which led to incorrect metric data because
each sampler has the wrong assumption it is the only object who calls
`GC::Profiler.clear` and thus could rely on the gathered results between
such calls.
We should ensure that `GC::Profiler.total_time` is called only in one
place making it possible to rely on accumulated data between such wipes.
Also, we need to track the amount of profiler reports we lost.
|
| |
|
|
|
|
|
|
|
| |
* Remove PNGs in favor of dynamically rendered graphs
* Eliminate <b> styling because this is no longer supported
* Change <br /> to <br> because the latter is only supported
* Add quotes around subgraph titles
|
| |
|
| |
|
| |
|
|
|
|
| |
Changes * to + to be more precise with acceptable branch names
|
|
|
|
|
|
|
|
|
|
| |
Mermaid v8.2.3 added support for styling in subgraphs, but if there are
multiple words in a title, quotes are required.
* https://github.com/knsv/mermaid/pull/845
* https://github.com/knsv/mermaid/issues/391#issuecomment-495602059
Closes https://gitlab.com/gitlab-org/gitlab-ce/issues/65338
|
| |
|
|
|
|
|
|
| |
Change sidebar's renderBlock() to return true only if
any of the block's <detail-row>s will be rendered
Change block to render only if renderBlock() returns true
|
| |
|
|
|
|
| |
Needs shift keypress
|
| |
|
|
|
|
|
| |
Move it to DiffPositionableNote concern which will be re-used
in EE in DraftNote model.
|
|
|
|
| |
CE port of https://gitlab.com/gitlab-org/gitlab-ee/merge_requests/14520
|
| |
|
|
|
|
| |
We removed the breadcrumbs in 0def36777e3f9e5f57fd0e1ebd023b43229622fc.
|
| |
|
|
|
|
|
|
| |
This task was removed in 2014! The spec wasn't, probably because it
didn't fail as it was never executed. See
0fc5c80207a1ebcfbdae1d0d16b9aeb984bac271.
|
| |
|
| |
|
|
|
|
| |
RSpec only executes files ending in _spec.rb.
|
|
|
|
|
|
|
|
| |
The RSpec/FilePath cop checks that a spec file's path is correct, but
only if the file describes a constant. We want to check, for any file
with a top-level `describe`, whether the file path ends in
`_spec.rb`. We don't care what comes before that; just that it will be
executed by RSpec at all.
|
|
|
|
| |
Documents how to enable security approvals
|
|
|
|
| |
- Update package.json and yarn.lock
|
|
|
|
|
| |
This reverts commit 28f2225bdfee4d353f07a766c6c8b29ba6181397, reversing
changes made to 96ae5bd83da31350e9856a290127d7aa1469710a.
|
|
|
|
| |
Introducing Docker Registry replication
|
|
|
|
| |
It's used only if merge_request_iid param is provided
|
|
|
|
|
| |
If a model has a default relative position we don't need to test that
it is moved to the end, it will instead get the default value assigned.
|
|
|
|
| |
Signed-off-by: Rémy Coutable <remy@rymai.me>
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Remove project from NotesFinder constructor
Add project parameter to specs
Also look for methods in private scope
Fix specs to match new NotesFinder constructor
|
|
|
|
|
|
|
|
|
| |
After measuring the response and performance of the workers on
https://gitlab.com/gitlab-org/gitlab-ce/issues/64092, we've decided 1.5
is fine for now. The time can be decreased later in the future when
enforcing the quota limits
Related to https://gitlab.com/gitlab-org/gitlab-ce/issues/64092
|
|
|
|
|
|
|
|
| |
Concurrent calls to UserMergeToRef RPC updating a single ref
can lead to an opaque fail that is being rescued at Gitaly.
So this commit adds an exclusive lease to the mergeability
check process with the key as the current MR ID.
|
| |
|
|
|
|
|
|
| |
This implements the support for `needs:` keyword
as part of GitLab CI. That makes some of the jobs
to be run out of order.
|
|
|
|
| |
Signed-off-by: Rémy Coutable <remy@rymai.me>
|
|
|
|
|
|
|
| |
Starts from START_POSITION when there are no existing
positions.
Also improves the test to actually test the behavior
|
|
|
|
| |
And remove it from quarantine
|
| |
|
|
|
|
|
|
|
| |
Make package-and-qa manual so the default `on_success` doesn't need
to be set on package-and-qa-always.
Remove unnecessary `except: []`
|