When running some multi-instance component tests repeatedly, I noted a potential race condition that seems to prevent some component instances from starting up properly. For instance, in the worker many-to-many or many-to-one tests, every now and then a fourth instance of the four worker instances will not properly begin emitting messages, likely indicating that the start AsyncResultHandler is never called.
When running some multi-instance component tests repeatedly, I noted a potential race condition that seems to prevent some component instances from starting up properly. For instance, in the worker many-to-many or many-to-one tests, every now and then a fourth instance of the four worker instances will not properly begin emitting messages, likely indicating that the start
AsyncResultHandler
is never called.