[jira] [Commented] (HIVE-15422) HiveInputFormat::pushProjectionsAndFilters paths comparison generates huge number of objects for partitioned dataset

["Hive QA (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HIVE-15422?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15744614#comment-15744614 ] 

Hive QA commented on HIVE-15422:
--------------------------------



Here are the results of testing the latest attachment:
https://issues.apache.org/jira/secure/attachment/12842945/HIVE-15422.1.patch

{color:green}SUCCESS:{color} +1 due to 1 test(s) being added or modified.

{color:red}ERROR:{color} -1 due to 10 failed/errored test(s), 10783 tests executed
*Failed tests:*
{noformat}
TestMiniLlapLocalCliDriver - did not produce a TEST-*.xml file (likely timed out) (batchId=144)
	[vectorized_rcfile_columnar.q,vector_elt.q,explainuser_1.q,multi_insert.q,tez_dml.q,vector_bround.q,schema_evol_orc_acid_table.q,vector_when_case_null.q,orc_ppd_schema_evol_1b.q,vector_join30.q,vectorization_11.q,cte_3.q,update_tmp_table.q,vector_decimal_cast.q,groupby_grouping_id2.q,vector_decimal_round.q,tez_smb_empty.q,orc_merge6.q,vector_decimal_trailing.q,cte_5.q,tez_union.q,cbo_rp_subq_not_in.q,vector_decimal_2.q,columnStatsUpdateForStatsOptimizer_1.q,vector_outer_join3.q,schema_evol_text_vec_part_all_complex.q,tez_dynpart_hashjoin_2.q,auto_sortmerge_join_12.q,offset_limit.q,tez_union_multiinsert.q]
TestVectorizedColumnReaderBase - did not produce a TEST-*.xml file (likely timed out) (batchId=251)
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[auto_sortmerge_join_2] (batchId=44)
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[sample2] (batchId=5)
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[sample4] (batchId=15)
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[sample6] (batchId=61)
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[sample7] (batchId=60)
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[sample9] (batchId=38)
org.apache.hadoop.hive.cli.TestMiniLlapCliDriver.testCliDriver[transform_ppr2] (batchId=135)
org.apache.hadoop.hive.cli.TestMiniLlapLocalCliDriver.testCliDriver[stats_based_fetch_decision] (batchId=151)
{noformat}

Test results: https://builds.apache.org/job/PreCommit-HIVE-Build/2556/testReport
Console output: https://builds.apache.org/job/PreCommit-HIVE-Build/2556/console
Test logs: http://104.198.109.242/logs/PreCommit-HIVE-Build-2556/

Messages:
{noformat}
Executing org.apache.hive.ptest.execution.TestCheckPhase
Executing org.apache.hive.ptest.execution.PrepPhase
Executing org.apache.hive.ptest.execution.ExecutionPhase
Executing org.apache.hive.ptest.execution.ReportingPhase
Tests exited with: TestsFailedException: 10 tests failed
{noformat}

This message is automatically generated.

ATTACHMENT ID: 12842945 - PreCommit-HIVE-Build

> HiveInputFormat::pushProjectionsAndFilters paths comparison generates huge number of objects for partitioned dataset
> --------------------------------------------------------------------------------------------------------------------
>
>                 Key: HIVE-15422
>                 URL: https://issues.apache.org/jira/browse/HIVE-15422
>             Project: Hive
>          Issue Type: Improvement
>            Reporter: Rajesh Balamohan
>            Priority: Minor
>         Attachments: HIVE-15422.1.patch, Profiler_Snapshot_HIVE-15422.png
>
>
> When executing the following query in LLAP (single instance) in a 5 node cluster, lots of GC pressure was observed.
> {noformat}
> select a.type, a.city , a.frequency, b.city, b.country, b.lat, b.lon
> from (select  'depart' as type, origin as city, count(origin) as frequency
> from flights
>   group by origin
>   order by frequency desc, type) as a 
> left join airports as b on a.city = b.iata
> order by frequency desc;
> {noformat}
> Flights table has got around 7000+ partitions in S3. Profiling revealed large amount of objects created just in path comparisons in HiveInputFormat.  HIVE-15405 reduces number of path comparisons at FileUtils, but it still ends up doing lots of comparisons in HiveInputFormat::pushProjectionsAndFilters.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)