increase test timeout in network_interface_integration_test

XMLWordPrintableJSON

    • Type: Bug
    • Resolution: Done
    • Priority: Major - P3
    • 9.0.0-rc0
    • Affects Version/s: None
    • Component/s: None
    • None
    • Networking & Observability
    • Fully Compatible
    • ALL
    • Hide

      One way to reproduce the failure is to add a sleep between when B's request is created and when it is sent.

      This isn't necessarily the way that the failure happened in the BF, but it probably is.

      See also the attached markdown transcript of my conversation with an AI about this.

      Show
      One way to reproduce the failure is to add a sleep between when B's request is created and when it is sent. This isn't necessarily the way that the failure happened in the BF, but it probably is. See also the attached markdown transcript of my conversation with an AI about this.
    • N&O 2026-05-11
    • 0
    • None
    • None
    • None
    • None
    • None
    • None
    • None

      BF-42961 is most likely due to a very rare test-only race condition caused when the test driver is suspended by Linux for many milliseconds.

      We could add another fail point to improve the synchronization between the test driver and the code under test, but I'm not confident that would really prevent the race.

      Instead, I'm increasing the relevant timeouts so that if this really is due to super rare process scheduling behavior, it will be much less likely to happen. 

        1. ai-transcript.md
          19 kB

            Assignee:
            David Goffredo
            Reporter:
            David Goffredo
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

              Created:
              Updated:
              Resolved: