- 09 Mar, 2022 40 commits
-
-
Vitali Tatarintev authored
gitaly_client: Don't check for pre-receive-errors in merge responses See merge request gitlab-org/gitlab!82287
-
Vitali Tatarintev authored
Replace issuable sort dropdown See merge request gitlab-org/gitlab!81430
-
Serhii Yarynovskyi authored
-
Kushal Pandya authored
Migrate Epic Tabs to use Button Group See merge request gitlab-org/gitlab!81978
-
Max Woolf authored
Remove feature flag project_owners_list_project_pending_deletion See merge request gitlab-org/gitlab!82372
-
Mikołaj Wawrzyniak authored
Prefer OncallShifts::ReadService to inherit from BaseProjectService See merge request gitlab-org/gitlab!82356
-
Sarah Yasonik authored
-
Andrejs Cunskis authored
E2E: Check if license present instead of checking for nil See merge request gitlab-org/gitlab!82468
-
Sanad Liaquat authored
-
Andrejs Cunskis authored
Rewrite snippet tests verification flow See merge request gitlab-org/gitlab!81769
-
Matija Čupić authored
Update Auto DevOps docker versions to 20.x See merge request gitlab-org/gitlab!82353
-
Furkan Ayhan authored
Support Vault EE namespaces See merge request gitlab-org/gitlab!80590
-
Aleksander Zak authored
Currently the vault integration does not support Vault EE namespaces. Only the vault server URL, role name and auth path can be configured with the variables. It is possible to just prefix the engine's path with the namespace, but it is not possible to specify the namespace for the authentication - it always uses the root namespace. This change makes the namespace configurable wiht the VAULT_NAMESPACE variable (similar to how `VAULT_SERVER_URL`, `VAULT_AUTH_ROLE` and `VAULT_AUTH_PATH are being used already). It will fall-back to an empty string in case the variable is not defined by the user. Fallback to an empty string instead of 'root' when no vault namespace specified by the user. Changelog: added
-
Marc Shaw authored
Remove default_merge_ref_for_diffs feature flag See merge request gitlab-org/gitlab!82069
-
Adam Hegyi authored
Fix MAX_BATCH_SIZE to match BATCH_SIZE See merge request gitlab-org/gitlab!82466
-
release-tools approver bot authored
Update Gitaly version See merge request gitlab-org/gitlab!82469
-
Max Woolf authored
Creates audit event when approval rule is deleted See merge request gitlab-org/gitlab!82297
-
Max Woolf authored
Fix scoped label text color in emails See merge request gitlab-org/gitlab!82360
-
Igor Drozdov authored
gitaly: Adapt to structured errors returned by UserSquash See merge request gitlab-org/gitlab!81905
-
Fabio Pitino authored
Improve matcher for publishing event See merge request gitlab-org/gitlab!80358
-
Albert Salim authored
-
Rajat Jain authored
Changelog: changed EE: true
-
Etienne Baqué authored
Extend Group API to get billable members count for a given plan See merge request gitlab-org/gitlab!82049
-
Vamsi Vempati authored
- Extended billableMembersCount field to accept requestedHostedPlan - Added tests for billableMembersCount - Updates GraphQL docs Changelog: changed EE: true
-
Matthias Käppler authored
Extract service to unassign runner from project See merge request gitlab-org/gitlab!81532
-
Pedro Pombeiro authored
-
GitLab Release Tools Bot authored
-
Martin Wortschack authored
Merge branch '353755-make-spinners-in-app-views-users-_overview-html-haml-pajamas-compliant' into 'master' Make spinners in users/_overview partial Pajamas-compliant See merge request gitlab-org/gitlab!82379
-
Alper Akgun authored
Add ability to check when free user cap setting reached on SaaS See merge request gitlab-org/gitlab!80518
-
Doug Stull authored
- for control of free user cap logic Changelog: added EE: true
-
Max Woolf authored
Ensure Vale lefthook rule returns error codes when necessary See merge request gitlab-org/gitlab!82370
-
Alexandru Croitor authored
-
Heinrich Lee Yu authored
Allow to destroy related epic links on internal API See merge request gitlab-org/gitlab!82035
-
Patrick Steinhardt authored
Gitaly is in the process of migrating UserSquash to use detailed errors like we already do for UserMergeBranch. This change will break the current calling convention, where Gitaly will return some errors by returning successfully, but with an `git_error` field set in the response. Instead, Gitaly will return a proper error for all conditions but give clients the ability to distinguish error cases via error details. Prepare the OperationService for this change by handling those new error details. Theoretically, this allows us to implement more fine-grained handling of these errors now: `ResolveRevisionErrors` contain the revision that Gitaly tried to resolve but couldn't, and `MergeConflictErrors` contain information about the files that have been conflicting. To ease the transition though we're treating those errors exactly the same as we did previously, where we raised a `GitError`. We can still iterate on this at a later point. The end result should be that there is no real user-visible change in how we handle UserSquash.
-
Patrick Steinhardt authored
Gitaly has started to return detailed errors in the context of UserMergeBranch, which was the first proof of concept to see whether this new model makes sense for error handling. With the introduction of structured error handling we also had to convert Rails to handle these new types of errors, which led to the introduction of the new helper function `#decode_detailed_error`. This function both is responsible for decoding the error, and translating it into Rails-specific errors. While it makes sense to have it decode errors, the concrete translation into Rails-specific errors is something that is specific for every RPC: UserMergeBranch can return one set of error details, while other RPCs may return a different set of error details. Mixing this all into a single function that is supposed to be shared between RPCs thus does not make much sense. Refactor the code to move the RPC-specific error translation into the RPC's own error handling. This prepares for the introduction of error detail handling in UserSquash.
-
Patrick Steinhardt authored
Tests for the UserCommitFiles RPC are currently part of the tests for UserSquash. This seems to be unintentional and doesn't make a whole lot of sense. Move out tests for UserCommitFiles into its own standalone context.
-
Kati Paizee authored
Clarify that admin must set webhook_timeout See merge request gitlab-org/gitlab!82464
-
Ben Bodenmiller authored
-
Kati Paizee authored
Note issue between GitLab Pages and Let's Encrypt See merge request gitlab-org/gitlab!82459
-
Dylan Griffith authored
Add test coverage for interruptible with child pipelines See merge request gitlab-org/gitlab!82347
-