You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@impala.apache.org by "Tim Armstrong (JIRA)" <ji...@apache.org> on 2017/11/01 21:35:00 UTC

[jira] [Resolved] (IMPALA-6093) TestHashJoinTimer failed on local filesystem and ASAN builds

     [ https://issues.apache.org/jira/browse/IMPALA-6093?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Tim Armstrong resolved IMPALA-6093.
-----------------------------------
       Resolution: Fixed
    Fix Version/s: Impala 2.11.0

We haven't seen it reoccur recently. We don't have anything actionable now, but if it happens again we will have a profile so can reopen this and continue investigating.

> TestHashJoinTimer failed on local filesystem and ASAN builds
> ------------------------------------------------------------
>
>                 Key: IMPALA-6093
>                 URL: https://issues.apache.org/jira/browse/IMPALA-6093
>             Project: IMPALA
>          Issue Type: Bug
>          Components: Infrastructure
>    Affects Versions: Impala 2.11.0
>            Reporter: Tim Armstrong
>            Assignee: Tim Armstrong
>            Priority: Critical
>              Labels: flaky
>             Fix For: Impala 2.11.0
>
>
> {noformat}
> 04:26:05  TestHashJoinTimer.test_hash_join_timer[test cases: ['select /*+straight_join*/ count(*) from (select distinct * from functional.alltypes where int_col >= sleep(5)) a cross join  functional.alltypes b where a.id > b.id and b.id=99', 'NESTED LOOP JOIN'] | exec_option: {'batch_size': 0, 'num_nodes': 0, 'disable_codegen_rows_threshold': 0, 'disable_codegen': False, 'abort_on_error': 1, 'exec_single_node_rows_threshold': 0} | table_format: text/none] 
> 04:26:05 query_test/test_hash_join_timer.py:140: in test_hash_join_timer
> 04:26:05     assert (asyn_build), "Join is not prepared asynchronously"
> 04:26:05 E   UnboundLocalError: local variable 'asyn_build' referenced before assignment
> {noformat}
> Saw this at commit: 2187d36... IMPALA-6045: Make build scripts more friendly to Ubuntu 16.04
> The test seems fundamentally flaky, since the behaviour will change if there are enough other threads were running simultaneously and taking up threads that the query couldn't start an async build thread. It's worth thinking about how we can make this less flaky without losing coverage.



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