Mailing-List: contact issues-help@hive.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@hive.apache.org
Date: Fri, 1 Dec 2017 05:46:00 +0000 (UTC)
From: "Pengcheng Xiong (JIRA)" <jira@apache.org>
To: issues@hive.apache.org
Message-ID: <JIRA.13120668.1511545802000.362554.1512107160203@Atlassian.JIRA>
In-Reply-To: <JIRA.13120668.1511545802000@Atlassian.JIRA>
References: <JIRA.13120668.1511545802000@Atlassian.JIRA> <JIRA.13120668.1511545802212@jira-lw-us.apache.org>
Subject: [jira] [Commented] (HIVE-18144) Runtime type inference error when
 join three table for different column type
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
archived-at: Fri, 01 Dec 2017 05:46:07 -0000


    [ https://issues.apache.org/jira/browse/HIVE-18144?page=3Dcom.atlassian=
.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=3D1627=
3968#comment-16273968 ]=20

Pengcheng Xiong commented on HIVE-18144:
----------------------------------------

LGTM. I saw you put 2.1 and 2.2 as affected versions. How about master?

> Runtime type inference error when join three table for different column t=
ype=20
> -------------------------------------------------------------------------=
----
>
>                 Key: HIVE-18144
>                 URL: https://issues.apache.org/jira/browse/HIVE-18144
>             Project: Hive
>          Issue Type: Bug
>          Components: Query Planning
>    Affects Versions: 2.1.1, 2.2.0
>            Reporter: Wang Haihua
>            Assignee: Wang Haihua
>         Attachments: HIVE-18144.1.patch
>
>
> Union operation with three or more table, which has different column type=
s, may cause type inference error when Task execution.
> E.g, e.g. t1(with column int) union all t2(with column int) union all t3(=
with column bigint), finally should be {{bigint}},
> RowSchema of union t1 with t2, we call {{leftOp}}, should be int, then le=
ftOp union t3 should finally be bigint.
> This mean RowSchema of leftOp would be {{bigint}} instead of {{int}}
> However we see in {{SemanticAnalyzer.java}}, leftOp RowSchema is finally =
{{int}} which was wrong:=20
> {code}
> (_col0: int|{t01-subquery1}diff_long_type,_col1: int|{t01-subquery1}id2,_=
col2: bigint|{t01-subquery1}id3)}}
> {code}
> Impacted code  in SemanticAnalyzer.java:
> {code}
>       if(!(leftOp instanceof UnionOperator)) {
>         Operator oldChild =3D leftOp;
>         leftOp =3D (Operator) leftOp.getParentOperators().get(0);
>         leftOp.removeChildAndAdoptItsChildren(oldChild);
>       }
>       // make left a child of right
>       List<Operator<? extends OperatorDesc>> child =3D
>           new ArrayList<Operator<? extends OperatorDesc>>();
>       child.add(leftOp);
>       rightOp.setChildOperators(child);
>       List<Operator<? extends OperatorDesc>> parent =3D leftOp
>           .getParentOperators();
>       parent.add(rightOp);
>       UnionDesc uDesc =3D ((UnionOperator) leftOp).getConf();
>       // Here we should set RowSchema of leftOp to unionoutRR's, or else =
the RowSchema of leftOp is wrong.
>       // leftOp.setSchema(new RowSchema(unionoutRR.getColumnInfos()));
>       uDesc.setNumInputs(uDesc.getNumInputs() + 1);
>       return putOpInsertMap(leftOp, unionoutRR);
> {code}
> Operation for reproduce=EF=BC=9A
> {code}
> create table test_union_different_type(id bigint, id2 bigint, id3 bigint,=
 name string);
> set hive.auto.convert.join=3Dtrue;
> insert overwrite table test_union_different_type select 1, 2, 3, "test_un=
ion_different_type";
> select
>   t01.diff_long_type as diff_long_type,
>   t01.id2 as id2,
>   t00.id as id,
>   t01.id3 as id3
> from test_union_different_type t00
> left join
>   (
>     select 1 as diff_long_type, 30 as id2, id3 from test_union_different_=
type
>     union ALL
>     select 2 as diff_long_type, 20 as id2, id3 from test_union_different_=
type
>     union ALL
>     select id as diff_long_type, id2, 30 as id3 from test_union_different=
_type
>   ) t01
> on t00.id =3D t01.diff_long_type
> ;
> {code}
> Stack trace:
> {code}
> Diagnostic Messages for this Task:
> Error: java.lang.RuntimeException: org.apache.hadoop.hive.ql.metadata.Hiv=
eException: Hive Runtime Error while processing row {"id":1,"id2":null,"id3=
":null,"name":null}
>   at org.apache.hadoop.hive.ql.exec.mr.ExecMapper.map(ExecMapper.java:169=
)
>   at org.apache.hadoop.mapred.MapRunner.run(MapRunner.java:54)
>   at org.apache.hadoop.mapred.MapTask.runOldMapper(MapTask.java:453)
>   at org.apache.hadoop.mapred.MapTask.run(MapTask.java:343)
>   at org.apache.hadoop.mapred.YarnChild$2.run(YarnChild.java:164)
>   at java.security.AccessController.doPrivileged(Native Method)
>   at javax.security.auth.Subject.doAs(Subject.java:415)
>   at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInform=
ation.java:1693)
>   at org.apache.hadoop.mapred.YarnChild.main(YarnChild.java:158)
> Caused by: org.apache.hadoop.hive.ql.metadata.HiveException: Hive Runtime=
 Error while processing row {"id":1,"id2":null,"id3":null,"name":null}
>   at org.apache.hadoop.hive.ql.exec.MapOperator.process(MapOperator.java:=
499)
>   at org.apache.hadoop.hive.ql.exec.mr.ExecMapper.map(ExecMapper.java:160=
)
>   ... 8 more
> Caused by: org.apache.hadoop.hive.ql.metadata.HiveException: Unexpected e=
xception from MapJoinOperator : org.apache.hadoop.io.LongWritable cannot be=
 cast to org.apache.hadoop.io.IntWritable
>   at org.apache.hadoop.hive.ql.exec.MapJoinOperator.process(MapJoinOperat=
or.java:465)
>   at org.apache.hadoop.hive.ql.exec.Operator.forward(Operator.java:879)
>   at org.apache.hadoop.hive.ql.exec.TableScanOperator.process(TableScanOp=
erator.java:130)
>   at org.apache.hadoop.hive.ql.exec.MapOperator$MapOpCtx.forward(MapOpera=
tor.java:149)
>   at org.apache.hadoop.hive.ql.exec.MapOperator.process(MapOperator.java:=
489)
>   ... 9 more
> Caused by: java.lang.ClassCastException: org.apache.hadoop.io.LongWritabl=
e cannot be cast to org.apache.hadoop.io.IntWritable
>   at org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableIntO=
bjectInspector.get(WritableIntObjectInspector.java:36)
>   at org.apache.hadoop.hive.serde2.lazy.LazyUtils.writePrimitiveUTF8(Lazy=
Utils.java:239)
>   at org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe.serialize(LazySim=
pleSerDe.java:292)
>   at org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe.serializeField(La=
zySimpleSerDe.java:247)
>   at org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe.doSerialize(LazyS=
impleSerDe.java:231)
>   at org.apache.hadoop.hive.serde2.AbstractEncodingAwareSerDe.serialize(A=
bstractEncodingAwareSerDe.java:55)
>   at org.apache.hadoop.hive.ql.exec.FileSinkOperator.process(FileSinkOper=
ator.java:714)
>   at org.apache.hadoop.hive.ql.exec.Operator.forward(Operator.java:879)
>   at org.apache.hadoop.hive.ql.exec.SelectOperator.process(SelectOperator=
.java:95)
>   at org.apache.hadoop.hive.ql.exec.Operator.forward(Operator.java:879)
>   at org.apache.hadoop.hive.ql.exec.CommonJoinOperator.internalForward(Co=
mmonJoinOperator.java:647)
>   at org.apache.hadoop.hive.ql.exec.CommonJoinOperator.genUniqueJoinObjec=
t(CommonJoinOperator.java:660)
>   at org.apache.hadoop.hive.ql.exec.CommonJoinOperator.genUniqueJoinObjec=
t(CommonJoinOperator.java:663)
>   at org.apache.hadoop.hive.ql.exec.CommonJoinOperator.checkAndGenObject(=
CommonJoinOperator.java:759)
>   at org.apache.hadoop.hive.ql.exec.MapJoinOperator.process(MapJoinOperat=
or.java:452)
>   ... 13 more
> {code}


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