Backport of 'Back off StoreScansWorker retries on lease contention' to 19.5

What does this MR do and why?

This MR backports !262568 (merged) to 19-5-stable-ee.

The MR is a git cherry-pick -x of the squash commit fba32c3a. The pick applied with no conflicts. The patch is identical to the default branch commit (same git patch-id).

What is backported

  1. Security::StoreGroupedScansService calls in_lock with retries: 0. The job tries for the store lease one time and does not sleep in the Sidekiq thread.
  2. When another job holds the lease, the service raises the new Security::StoreGroupedScansService::LeaseTakenError. This class is a subclass of Gitlab::SidekiqMiddleware::RetryError, so the error does not go to Sentry or count against the execution SLI.
  3. Security::StoreScansWorker gets a sidekiq_retry_in block. For LeaseTakenError, the retry delays are 1, 2, 4, 8 and 16 minutes. For all other errors, Sidekiq uses its default delay.
  4. Security::StoreScansWorker retries 5 times instead of 3. spec/workers/every_sidekiq_worker_spec.rb has the new retry count.

Why

The store lease is per pipeline and report file type, with a TTL of 30 minutes. Before this change, a job that did not get the lease waited for it in its Sidekiq thread for about 30 seconds. Then the worker used its next retry with the default delay. All 4 attempts ran in about 3 to 4 minutes. If the lease holder took longer than that, the waiting job failed permanently and did not store its scans.

With this change, the retries cover more than the lease TTL (31 minutes in total). The waiting job does not block a Sidekiq thread.

The related confidential issue https://gitlab.com/gitlab-org/gitlab/-/work_items/633845 has the details.

Risk

  • Security::StoreScansWorker is the only production caller of this code path. No web request depends on the old wait behavior.
  • Other errors also get the 2 extra retries, with the default delay.

Default branch status

Item Value
Default branch MR !262568 (merged)
Squash commit fba32c3a
Merged 2026-10-21
First deployed to GitLab.com production 2026-10-21

Testing on 19-5-stable-ee

I ran these specs on the backport commit, on top of 19-5-stable-ee:

Spec Result
ee/spec/services/security/store_grouped_scans_service_spec.rb, ee/spec/workers/security/store_scans_worker_spec.rb, ee/spec/services/security/store_grouped_sbom_scans_service_spec.rb and ee/spec/services/security/store_scans_service_spec.rb 70 examples, 0 failures

RuboCop found no offenses in the 5 changed files.

MR acceptance checklist

This checklist encourages us to confirm any changes have been analyzed to reduce risks in quality, performance, reliability, security, and maintainability.

  • This MR is backporting a bug fix, documentation update, or spec fix, previously merged in the default branch.
  • The MR that fixed the bug on the default branch has been deployed to GitLab.com (not applicable for documentation or spec changes).
  • The MR title is descriptive (e.g. "Backport of 'title of default branch MR'"). This is important, since the title will be copied to the patch blog post.
  • Required labels have been applied to this merge request
  • This MR has been approved by a maintainer (only one approval is required).
  • Ensure the E2E Omnibus GitLab EE pipeline (triggered by thee2e:test-on-omnibus-ee job) has succeeded. If it has failed (or shows warnings), investigate the failures. If you determine the failures are unrelated, you may proceed. If you need assistance investigating, request help in the #s_developer_experience Slack channel to confirm the failures are unrelated to the merge request.

Note to the merge request author and maintainer

If you have questions about the patch release process, please:

Edited by Gregory Havenga

Merge request reports

Loading
Loading