Skip to content

test: stabilize SQL Server reminder lifecycle tests on net10 #11057

Closed as not planned

Description

@ReubenBond

The Microsoft SQL Server provider tests (net10.0) job in CI run 33809237932 consistently timed out while waiting for reminder topology reconciliation. The corresponding net8 job passed, and the failing code is unrelated to the single-file metadata changes in PR #11054.

Affected tests include:

  • ReminderService_CleanupIsIsolated
  • ReminderService_StartupReadiness
  • ReminderService_RemovalReachesQuiescence
  • ReminderService_ExactDueRecovery

Representative failure:

System.OperationCanceledException: Reminder topology stabilization was canceled during 'stable topology refresh start'.
Expected active silos and observed active silos matched, but one silo remained at reconciliationCompleted=False until the scenario timeout.

The stack terminates in ReminderTopologyStabilizer.WaitForStableTopologyAsync and ReminderServiceLifecycleHarness.StabilizeTopologyAsync. Investigate why the first silo remains on a 100% ring range with reconciliation generation 1 while the second silo completes reconciliation on net10 SQL Server runs.

Activity

  1. ReubenBond commented on Sep 3, 2026

    @ReubenBond
    MemberAuthor

    Duplicate of #11048.

  2. locked and limited conversation to collaborators on Oct 4, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    area-testingOrleans testing related issues

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions