Abdul Wadood activity https://gitlab.com/abdwdd 2026-03-17T04:59:34Z tag:gitlab.com,2026-03-17:5210955003 Abdul Wadood commented on merge request !227215 at GitLab.org / GitLab 2026-03-17T04:59:34Z abdwdd Abdul Wadood

@smaglangit could you please do the initial backend review as it's blocking some future MRs?

tag:gitlab.com,2026-03-16:5210058996 Abdul Wadood commented on issue #592091 at GitLab.org / GitLab 2026-03-16T20:54:56Z abdwdd Abdul Wadood

All of these shouldn't reside in the namespaces table, and we should address this technical debt.

Column Index
custom_project_templates_group_id index_namespaces_on_custom_project_templates_group_id_and_type
file_template_project_id index_namespaces_on_file_template_project_id
push_rule_id index_namespaces_on_push_rule_id
runners_token_encrypted index_namespaces_on_runners_token_encrypted
ldap_sync_last_successful_update_at index_namespaces_on_ldap_sync_last_successful_update_at
tag:gitlab.com,2026-03-16:5208447678 Abdul Wadood commented on merge request !225485 at GitLab.org / GitLab 2026-03-16T13:49:26Z abdwdd Abdul Wadood

@alipniagov approved!

tag:gitlab.com,2026-03-16:5208446403 Abdul Wadood approved merge request !225485: Draft: Do Not Merge - Allow service account creation for subgroups and projects at GitLab.org / GitLab 2026-03-16T13:49:10Z abdwdd Abdul Wadood

What does this MR do and why?

Remove feature flags:

  • allow_projects_to_create_service_accounts
  • allow_subgroups_to_create_service_accounts

...to enable service account management in subgroups and projects by default.

Both FFs are globally enabled on .com within %18.10.

These two FFs are tightly coupled, so I see a very little value in splitting this MR per-FF: it will touch the same files all again.

References

allow_projects_to_create_service_accounts rollout issue: https://gitlab.com/gitlab-org/gitlab/-/work_items/586190+

allow_subgroups_to_create_service_accounts rollout issue: https://gitlab.com/gitlab-org/gitlab/-/work_items/582313+

How to set up and validate locally

N/A - FF removal.
Rely on CI and reviews.

MR acceptance checklist

Evaluate this MR against the MR acceptance checklist. It helps you analyze changes to reduce risks in quality, performance, reliability, security, and maintainability.

Related to #586190

tag:gitlab.com,2026-03-16:5206569358 Abdul Wadood commented on merge request !226211 at GitLab.org / GitLab 2026-03-16T06:19:50Z abdwdd Abdul Wadood

@julie_huang Thanks! Feel free to merge for the release. 🚀

tag:gitlab.com,2026-03-16:5206568760 Abdul Wadood approved merge request !226211: Draft: Add `provider` column to `ai_self_hosted_models` table at GitLab.org / GitLab 2026-03-16T06:19:35Z abdwdd Abdul Wadood

What does this MR do and why?

  • Adds provider column to ai_self_hosted_models table
  • Adds provider enum to SelfHostedModel Rails model

Context:

We want to grow our roster of provider options and need to start storing that information in the backend on the self-hosted model record. Currently that information is inferred by the frontend using the stored model identifier, but this is a brittle approach that doesn't scale well.

Screenshot_2026-03-13_at_2.22.08_pm

References

Issue: #507967

Screenshots or screen recordings

Before After

How to set up and validate locally

Pre-requisites:

An active Ultimate license.

  1. Start GDK in self-managed mode GITLAB_SIMULATE_SAAS=0 gdk start
  2. Run the Duo setup rake task GITLAB_SIMULATE_SAAS=0 bundle exec 'rake gitlab:duo:setup' to seed instance with necessary Duo data
  3. Navigate to the self-hosted models page http://gdk.test:3000/admin/gitlab_duo/model_selection/models and ensure there are no regressions by ensuring that you can create, edit, delete and assign self-hosted models.

MR acceptance checklist

Evaluate this MR against the MR acceptance checklist. It helps you analyze changes to reduce risks in quality, performance, reliability, security, and maintainability.

tag:gitlab.com,2026-03-16:5206349411 Abdul Wadood commented on issue #585315 at GitLab.org / GitLab 2026-03-16T04:16:37Z abdwdd Abdul Wadood

We're adding an index on namespace_details.deletion_scheduled_at in Migrate adjourned group deletion to namespace s... (!226560 - merged), which we moved out of state_metadata.

tag:gitlab.com,2026-03-16:5206337531 Abdul Wadood commented on merge request !227215 at GitLab.org / GitLab 2026-03-16T04:11:16Z abdwdd Abdul Wadood

@abime I have addressed your comments. 🏓

tag:gitlab.com,2026-03-16:5206337506 Abdul Wadood commented on merge request !227215 at GitLab.org / GitLab 2026-03-16T04:11:16Z abdwdd Abdul Wadood

Good catch! I have updated the specs.

tag:gitlab.com,2026-03-16:5206337473 Abdul Wadood commented on merge request !227215 at GitLab.org / GitLab 2026-03-16T04:11:15Z abdwdd Abdul Wadood

For now, I just want to replicate the behavior of the service, which only archived the project, and the policy also didn't check if the parent was scheduled for deletion. This will also be addressed in the migration to state management.

tag:gitlab.com,2026-03-16:5206337454 Abdul Wadood commented on merge request !227215 at GitLab.org / GitLab 2026-03-16T04:11:14Z abdwdd Abdul Wadood

This is because of this override https://gitlab.com/gitlab-org/gitlab/-/blob/0be623bd96762dbbe89cdb67abc2e53cb92d385e/app/models/project.rb#L3136. We'll remove it once we migrate to state management.

tag:gitlab.com,2026-03-16:5206337428 Abdul Wadood commented on merge request !227215 at GitLab.org / GitLab 2026-03-16T04:11:14Z abdwdd Abdul Wadood

A part of this will be covered under Fix mismatch between namespace states and proje... (#593442).

But this would be completely addressed in Iteration 4: Replace legacy state checks with n... (gitlab-org#17958).

Another example of this inconsistency that currently exists is if you archive a project, then schedule it for deletion, but then restore the project, the project doesn't stay archived, it goes back to the unarchived state. We're addressing such inconsistencies in the above epic.

tag:gitlab.com,2026-03-16:5206337396 Abdul Wadood commented on merge request !227215 at GitLab.org / GitLab 2026-03-16T04:11:13Z abdwdd Abdul Wadood

This doesn't warrant a feature flag; we're just shifting the responsibility from the service to the policy to enforce the same behavior. We also have good spec coverage.

tag:gitlab.com,2026-03-16:5206337039 Abdul Wadood pushed to project branch 588699-treat-scheduled-deletion-as-archived at GitLab.org / GitLab 2026-03-16T04:10:58Z abdwdd Abdul Wadood

Abdul Wadood (6fab4e2a) at 16 Mar 04:10

Add policy spec for projects scheduled for deletion

... and 1037 more commits

tag:gitlab.com,2026-03-16:5206154671 Abdul Wadood commented on merge request !227402 at GitLab.org / GitLab 2026-03-16T02:10:47Z abdwdd Abdul Wadood

@aslota database LGTM!

tag:gitlab.com,2026-03-16:5206154358 Abdul Wadood approved merge request !227402: Update name and onlyAvailable argument removal milestone to 19.0 at GitLab.org / GitLab 2026-03-16T02:10:31Z abdwdd Abdul Wadood

What does this MR do and why?

Delay the removal of deprecated name and onlyAvailable arguments in WorkItemTypes GraphQL API from milestone 18.11 to 19.0 to avoid multi-version compatibility issues during rolling deployments as discussed here.

This ensures that if the frontend stops using the name and onlyAvailable fields in 18.11 via #593234, there's an additional milestone buffer before backend removal to prevent scenarios where old frontends query new backends that no longer support these arguments.

References

Screenshots or screen recordings

NA

How to set up and validate locally

NA

MR acceptance checklist

Evaluate this MR against the MR acceptance checklist. It helps you analyze changes to reduce risks in quality, performance, reliability, security, and maintainability.

tag:gitlab.com,2026-03-15:5205882937 Abdul Wadood commented on task #525999 at GitLab.org / GitLab 2026-03-15T22:27:47Z abdwdd Abdul Wadood

Completed in Backfill `project_authorizations_for_migrations... (!190716 - merged).

tag:gitlab.com,2026-03-15:5205878446 Abdul Wadood commented on merge request !226983 at GitLab.org / GitLab 2026-03-15T22:22:28Z abdwdd Abdul Wadood

@chen-gitlab I have dropped some comments. 🏓

tag:gitlab.com,2026-03-15:5205878436 Abdul Wadood commented on merge request !226983 at GitLab.org / GitLab 2026-03-15T22:22:28Z abdwdd Abdul Wadood

I would suggest avoiding a fixed set of routes.

Instead, we can get the namespace_id/project_id like we try to get the organization_id in lib/gitlab/current/organization.rb.

@rutgerwessels do you have any suggestions here since you've worked on the above?

tag:gitlab.com,2026-03-15:5205878432 Abdul Wadood commented on merge request !226983 at GitLab.org / GitLab 2026-03-15T22:22:28Z abdwdd Abdul Wadood

We will have some GET requests as well, where we would be writing to the DB. So instead of relying on these HTTP methods, we should just check if the namespace is read-only or not.