[ https://issues.apache.org/jira/browse/BEAM-10054?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Kenneth Knowles updated BEAM-10054: ----------------------------------- Status: Open (was: Triage Needed) > Direct Runner execution stalls with test pipeline > ------------------------------------------------- > > Key: BEAM-10054 > URL: https://issues.apache.org/jira/browse/BEAM-10054 > Project: Beam > Issue Type: Bug > Components: sdk-py-core > Reporter: Maximilian Michels > Assignee: Maximilian Michels > Priority: P2 > > Internally, we have a test pipeline which runs with the DirectRunner. When > upgrading from 2.18.0 to 2.21.0 the test failed with the following exception: > {noformat} > tp = Exception('Monitor task detected a pipeline stall.',), value = None, tb > = None > def raise_(tp, value=None, tb=None): > """ > A function that matches the Python 2.x ``raise`` statement. This > allows re-raising exceptions with the cls value and traceback on > Python 2 and 3. > """ > if value is not None and isinstance(tp, Exception): > raise TypeError("instance exception may not have a separate > value") > if value is not None: > exc = tp(value) > else: > exc = tp > if exc.__traceback__ is not tb: > raise exc.with_traceback(tb) > > raise exc > E Exception: Monitor task detected a pipeline stall. > {noformat} > I was able to bisect the error. This commit introduced the failure: > https://github.com/apache/beam/commit/ea9b1f350b88c2996cafb4d24351869e82857731 > If the following conditions evaluates to False, the pipeline runs correctly: > https://github.com/apache/beam/commit/ea9b1f350b88c2996cafb4d24351869e82857731#diff-2bb845e226f3a97c0f0f737d0558c5dbR1273 -- This message was sent by Atlassian Jira (v8.3.4#803005)