You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues-all@impala.apache.org by "ASF subversion and git services (Jira)" <ji...@apache.org> on 2026/07/23 16:08:00 UTC

[jira] [Commented] (IMPALA-15045) TestHS2.test_long_polling_full_sleep fails on UBSAN ARM builds

    [ https://issues.apache.org/jira/browse/IMPALA-15045?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18098520#comment-18098520 ] 

ASF subversion and git services commented on IMPALA-15045:
----------------------------------------------------------

Commit f325f7406410a2356568a5f20752e56797989e64 in impala's branch refs/heads/master from Joe McDonnell
[ https://gitbox.apache.org/repos/asf?p=impala.git;h=f325f7406 ]

IMPALA-15045: Deflake TestHS2.test_long_polling_full_sleep

TestHS2.test_long_polling_full_sleep has a check of the
timing of a single call to GetOperationStatus. On some
slow build types (particularly UBSAN on ARM), the call is
taking longer than the expected 100ms. The highest time seen
in the failing tests is a little above 200ms. To avoid this
flakiness, this increases the threshold to 250ms for slow
build types. This is still well short of the runtime of
the query.

Testing:
 - Ran locally

Change-Id: I2c24ca5e3caad74136ed7c6d0ad16c24ea627966
Reviewed-on: http://gerrit.cloudera.org:8080/24627
Tested-by: Impala Public Jenkins <im...@cloudera.com>
Reviewed-by: Arnab Karmakar <ar...@gmail.com>
Reviewed-by: Michael Smith <mi...@cloudera.com>


> TestHS2.test_long_polling_full_sleep fails on UBSAN ARM builds
> --------------------------------------------------------------
>
>                 Key: IMPALA-15045
>                 URL: https://issues.apache.org/jira/browse/IMPALA-15045
>             Project: IMPALA
>          Issue Type: Bug
>          Components: Test
>    Affects Versions: Impala 5.0.0
>            Reporter: Joe McDonnell
>            Assignee: Joe McDonnell
>            Priority: Critical
>              Labels: broken-build, flaky
>             Fix For: Impala 5.0.0
>
>
> Builds on UBSAN ARM are hitting a failure in hs2.test_hs2.TestHS2.test_long_polling_full_sleep:
> {noformat}
> hs2/hs2_test_suite.py:68: in add_session
>     add_session_helper(self, protocol_version, conf_overlay, close_session,
>         close_session = True
>         conf_overlay = {'long_polling_time_ms': '50'}
>         fn         = <function TestHS2.test_long_polling_full_sleep at 0xffff81896a20>
>         protocol_version = 5
>         self       = <tests.hs2.test_hs2.TestHS2 object at 0xffff7eb8a090>
> hs2/hs2_test_suite.py:50: in add_session_helper
>     fn()
>         close_session = True
>         close_session_req = TCloseSessionReq(sessionHandle=TSessionHandle(sessionId=THandleIdentifier(guid=b'\xe1l=\xba\x80\x01B\xaa\x86\xf1yM\xab\xfb\xff\xb5', secret=b'\xf8\xff\xf7\xdf\xd4\xa1@\xad\x90k\xb8\xae\xcc2\x81d')))
>         conf_overlay = {'long_polling_time_ms': '50'}
>         fn         = <function needs_session.<locals>.session_decorator.<locals>.add_session.<locals>.<lambda> at 0xffff7eac5b20>
>         open_session_req = TOpenSessionReq(client_protocol=5, username='jenkins', password=None, configuration={'long_polling_time_ms': '50'})
>         protocol_version = 5
>         resp       = TOpenSessionResp(status=TStatus(statusCode=0, infoMessages=None, sqlState=None, errorCode=None, errorMessage=None), se..., 'WRITE_KUDU_UTC_TIMESTAMPS': '0', 'http_addr': 'impala-ec2-rhel88-m7g-4xlarge-ondemand-038a.vpc.cloudera.com:25000'})
>         self       = <tests.hs2.test_hs2.TestHS2 object at 0xffff7eb8a090>
> hs2/hs2_test_suite.py:69: in <lambda>
>     lambda: fn(self))
>         fn         = <function TestHS2.test_long_polling_full_sleep at 0xffff81896a20>
>         self       = <tests.hs2.test_hs2.TestHS2 object at 0xffff7eb8a090>
> /data/jenkins/workspace/impala-asf-master-core-ubsan-arm/repos/Impala/tests/hs2/test_hs2.py:422: in test_long_polling_full_sleep
>     assert end_time - start_time < 0.1
> E   assert (1779841368.259929 - 1779841368.1474683) < 0.1
>         end_time   = 1779841368.259929
>         execute_statement_resp = TExecuteStatementResp(status=TStatus(statusCode=0, infoMessages=None, sqlState=None, errorCode=None, errorMessage=None...t=b'\xf8\xff\xf7\xdf\xd4\xa1@\xad\x90k\xb8\xae\xcc2\x81d'), operationType=0, hasResultSet=True, modifiedRowCount=None))
>         get_operation_status_resp = TGetOperationStatusResp(status=TStatus(statusCode=0, infoMessages=None, sqlState=None, errorCode=None, errorMessage=None), operationState=1, sqlState=None, errorCode=None, errorMessage=None)
>         self       = <tests.hs2.test_hs2.TestHS2 object at 0xffff7eb8a090>
>         start_time = 1779841368.1474683
>         statement  = 'SELECT count(sleep(100)) from functional.alltypestiny'{noformat}
> This is a timing issue like other failures for the long polling tests. We should adjust the expected time.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-all-unsubscribe@impala.apache.org
For additional commands, e-mail: issues-all-help@impala.apache.org