[IMPALA-10357] A test case in test_basic_joins() seems flaky

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Duplicate
• Affects Version/s: Impala 4.0.0
• Labels: broken-build flaky-test
• Epic Color: ghx-label-14
• Assignee: Tim Armstrong
• Reporter: Fang-Yu Rao
• Created: 25/Nov/20 00:23
• Updated: 25/Nov/20 05:05
• Resolved: 25/Nov/20 05:05

Description

We found in a recent build that a test case in test_basic_joins() does not fail with the expected error. This test reads the test case at joins.test and verifies if the returned result is "Debug Action: FAIL".

Error Message

query_test/test_join_queries.py:71: in test_basic_joins     self.run_test_case('QueryTest/joins', new_vector) common/impala_test_suite.py:668: in run_test_case     self.__verify_exceptions(test_section['CATCH'], str(e), use_db) common/impala_test_suite.py:485: in __verify_exceptions     (expected_str, actual_str) E   AssertionError: Unexpected exception string. Expected: Debug Action: FAIL E   Not found in actual: ImpalaBeeswaxException: Query aborted:ExecQueryFInstances rpc query_id=9f4e869839247e75:d01b009300000000 failed: Exec() rpc failed: Aborted: ExecQueryFInstances RPC to 127.0.0.1:27002 is cancelled in state SENT

Stacktrace

query_test/test_join_queries.py:71: in test_basic_joins
    self.run_test_case('QueryTest/joins', new_vector)
common/impala_test_suite.py:668: in run_test_case
    self.__verify_exceptions(test_section['CATCH'], str(e), use_db)
common/impala_test_suite.py:485: in __verify_exceptions
    (expected_str, actual_str)
E   AssertionError: Unexpected exception string. Expected: Debug Action: FAIL
E   Not found in actual: ImpalaBeeswaxException: Query aborted:ExecQueryFInstances rpc query_id=9f4e869839247e75:d01b009300000000 failed: Exec() rpc failed: Aborted: ExecQueryFInstances RPC to 127.0.0.1:27002 is cancelled in state SENT

Standard Error

set debug_action="2:PREPARE:FAIL|COORD_BEFORE_EXEC_RPC:JITTER@100@0.3";

-- 2020-11-23 23:27:30,175 INFO     MainThread: Started query 1a49038843b6b4da:30f6228100000000
-- executing against localhost:21000


select 1 from alltypestiny a join alltypestiny b on a.id != b.id;

-- 2020-11-23 23:27:30,188 INFO     MainThread: Started query 9f4e869839247e75:d01b009300000000
-- executing against localhost:21000

According to the output to standard error, two related queries are of query id's "1a49038843b6b4da:30f6228100000000" and "9f4e869839247e75:d01b009300000000".

Using these two query id's, we excerpted the related log entries from two log files of impalad's.

Since this test was recently revised in IMPALA-9309, maybe tarmstrong could offer some insight into it. Assigned it to Tim for now but feel free to re-assign it as you see appropriate. Also let me know if some other logs are needed for investigation. Thanks!

Issue Links

• duplicates: IMPALA-10336 test_type_conversions_hive3 fails because incorrect error is returned to client