Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-103245

Only use larger kCancelCommandTimeout in tests that require it

    • Type: Icon: Improvement Improvement
    • Resolution: Fixed
    • Priority: Icon: Major - P3 Major - P3
    • 8.2.0-rc0, 8.1.1
    • Affects Version/s: None
    • Component/s: None
    • None
    • Networking & Observability
    • Fully Compatible
    • v8.1
    • N&O 2025-04-28, N&O 2025-05-12
    • None
    • 3
    • None
    • None
    • None
    • None
    • None
    • None

      In SERVER-103048, the timeout used for _killOperations commands sent as part of NetworkInterfaceTL's cancellation logic was increased whenever the TestingProctor is enabled. This was done to fix BF-36510, which involved a C++ integration test that would occasionally hit this timeout and fail on one buildvariant. However, the TestingProctor is enabled in all testing environments, including jstests and other C++ integration tests, which is a much larger scope than is needed to address BF-36510. We should consider modifying the fix applied in SERVER-103048 to use a failpoint so that we can only enable it where necessary (e.g. in the CancelRemotely test), as this will ensure the rest of our test suite is properly exercising the behavior that users will see in production.

            Assignee:
            joseph.prince@mongodb.com Joseph Prince
            Reporter:
            patrick.freed@mongodb.com Patrick Freed
            Votes:
            0 Vote for this issue
            Watchers:
            1 Start watching this issue

              Created:
              Updated:
              Resolved: