You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@impala.apache.org by "Thomas Tauber-Marshall (JIRA)" <ji...@apache.org> on 2017/11/06 16:24:00 UTC

[jira] [Created] (IMPALA-6164) test_always_false_filter failure on ASAN

Thomas Tauber-Marshall created IMPALA-6164:
----------------------------------------------

             Summary: test_always_false_filter failure on ASAN
                 Key: IMPALA-6164
                 URL: https://issues.apache.org/jira/browse/IMPALA-6164
             Project: IMPALA
          Issue Type: Bug
          Components: Infrastructure
    Affects Versions: Impala 2.11.0
            Reporter: Thomas Tauber-Marshall
            Assignee: Tianyi Wang
            Priority: Blocker


Seen in an asf-master ASAN job at git hash 6b37a793a0af74ba01c67bb5ac7285af7c5c4d52:
{noformat}
04:15:21 ____________________ TestAlwaysFalseFilter.test_skip_split _____________________
04:15:21 
04:15:21 self = <test_always_false_filter.TestAlwaysFalseFilter object at 0x347a810>
04:15:21 cursor = <impala.hiveserver2.HiveServer2Cursor object at 0x3246310>
04:15:21 
04:15:21     @SkipIfBuildType.not_dev_build
04:15:21     @pytest.mark.execute_serially
04:15:21     @CustomClusterTestSuite.with_args("--skip_file_runtime_filtering=true")
04:15:21     def test_skip_split(self, cursor):
04:15:21       """IMPALA-5789: Test that always false filter filters out splits when file-level
04:15:21         filtering is disabled. The filtering is not enabled in seq-based file formats."""
04:15:21       cursor.execute("SET RUNTIME_FILTER_MODE=GLOBAL")
04:15:21       cursor.execute("SET RUNTIME_FILTER_WAIT_TIME_MS=10000")
04:15:21       query = """select STRAIGHT_JOIN * from alltypes inner join
04:15:21                 (select * from alltypessmall where smallint_col=-1) v
04:15:21                 on v.year = alltypes.year"""
04:15:21       # Manually iterate through file formats instead of creating a test matrix to prevent
04:15:21       # the cluster from restarting multiple times.
04:15:21       for table_suffix in ['_avro', '_rc', '_seq']:
04:15:21         cursor.execute("use functional" + table_suffix)
04:15:21         cursor.execute(query)
04:15:21         profile = cursor.get_profile()
04:15:21         assert re.search("Files rejected: [^0] \([^0]\)", profile) is None
04:15:21         assert re.search("Splits rejected: [^0] \([^0]\)", profile) is None
04:15:21       for table_suffix in ['', '_parquet']:
04:15:21         cursor.execute("use functional" + table_suffix)
04:15:21         cursor.execute(query)
04:15:21         profile = cursor.get_profile()
04:15:21         assert re.search("Files rejected: [^0] \([^0]\)", profile) is None
04:15:21 >       assert re.search("Splits rejected: 8 \(8\)", profile) is not None
04:15:21 E       assert None is not None
04:15:21 E        +  where None = <function search at 0x1076cf8>('Splits rejected: 8 \\(8\\)', 'Query (id=b1414c2cf5045bec:4e93cc3200000000):\n  DEBUG MODE WARNING: Query profile created while running a DEBUG buil...ryContextSwitches: 7 (7)\n           - TotalRawHdfsReadTime(*): 4.999ms\n           - TotalReadThroughput: 0.00 /sec\n')
04:15:21 E        +    where <function search at 0x1076cf8> = re.search
04:15:21 
04:15:21 custom_cluster/test_always_false_filter.py:53: AssertionError
{noformat}



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)