timeseries jstests that spawn parallel shells do not handle FCV upgrade/downgrade properly

    • Type: Bug
    • Resolution: Unresolved
    • Priority: Major - P3
    • None
    • Affects Version/s: None
    • Component/s: None
    • None
    • Catalog and Routing
    • ALL
    • 200
    • None
    • None
    • None
    • None
    • None
    • None
    • None

      As part of the viewless timeseries project a new retriable error was introduced that gets returned if the targeted timeseries is undergoing an FCV upgrade/downgrade called InterruptedDueToTimeseriesUpgradeDowngrade.

      This error is automatically handled in normal jstesting suites that use retriable writes. However, this handling does not occur in tests that spawn a parallel shell since that shell will not load the overrides necessary to handle the error. As such, it will propagate the error upwards and fail the error since it occurs outside of the runCommand method.

      A search reveals that these jstests in the jstests/core/timeseries/write folder are susceptible to this failure:

      • timeseries_delete_hint.js
      • timeseries_delete_with_meta_concurrent.js
      • timeseries_update_concurrent_arbitrary_updates_not_enabled.js
      • timeseries_update_concurrent.js
      • timeseries_update_hint.js

            Assignee:
            Unassigned
            Reporter:
            Jordi Olivares Provencio
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

              Created:
              Updated: