[
https://issues.apache.org/jira/browse/HIVE-22040?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16894833#comment-16894833
]
Jason Dere commented on HIVE-22040:
-----------------------------------
Hi [~xiepengjie], thanks for submitting the patch. What version of Hive are you
working with? The patch does not apply on Hive master because the source files
are now in standalone-metastore/metastore-server and
standalone-metastore/metastore-common on master.
I also was not able to get TestDropPartitions to run properly on master branch
.. it seems like there have been some changes to how this test is run, that I
am not aware of. I would recommend adding the following qfile test to
TestMiniLlapCliDriver, to simulate the failure scenario you mentioned:
{noformat}
create table delete_parent_path (c1 int) partitioned by (year string, month
string, day string) location 'hdfs:///tmp/delete_parent_path';
alter table delete_parent_path add partition (year='2019', month='07',
day='01');
dfs -rm -r hdfs:///tmp/delete_parent_path/year=2019/month=07;
alter table delete_parent_path drop partition (year='2019', month='07',
day='01');
{noformat}
However I noticed that this no longer fails on Hive master. It seems like this
has been fixed by HIVE-17472.
I do agree with the change that you made in Warehouse.isEmpty(), if you could
either remove the existing test case or replace it with the one I suggested.
> Drop partition throws exception with 'Failed to delete parent: File does not
> exist' when the partition's parent path does not exists
> ------------------------------------------------------------------------------------------------------------------------------------
>
> Key: HIVE-22040
> URL: https://issues.apache.org/jira/browse/HIVE-22040
> Project: Hive
> Issue Type: Improvement
> Components: Metastore
> Affects Versions: 1.2.1, 2.0.0, 3.0.0
> Reporter: xiepengjie
> Assignee: xiepengjie
> Priority: Major
> Attachments: HIVE-22040.01.patch, HIVE-22040.patch
>
>
> I create a manage table with multi partition columns, when i try to drop
> partition throws exception with 'Failed to delete parent: File does not
> exist' when the partition's parent path does not exist. The partition's
> metadata in mysql has been deleted, but the exception is still thrown. it
> will fail if connecting hiveserver2 with jdbc by java, this problem also
> exists in master branch, I think it is very unfriendly and we should fix it.
> Example:
> – First, create manage table with nulti partition columns, and add partitions:
> {code:java}
> drop table if exists t1;
> create table t1 (c1 int) partitioned by (year string, month string, day
> string);
> alter table t1 add partition(year='2019', month='07', day='01');{code}
> – Second, delete the path of partition 'month=07':
> {code:java}
> hadoop fs -rm -r
> /user/hadoop/xiepengjietest.db/drop_partition/year=2019/month=07{code}
> -- Third, when i try to drop partition, the metastore throws exception with
> 'Failed to delete parent: File does not exist' .
> {code:java}
> alter table t1 drop partition partition(year='2019', month='07', day='01');
> {code}
> exception like this:
> {code:java}
> Error: Error while processing statement: FAILED: Execution Error, return code
> 1 from org.apache.hadoop.hive.ql.exec.DDLTask. Failed to delete parent: File
> does not exist:
> /user/hadoop/xiepengjietest.db/drop_partition/year=2019/month=07
> at
> org.apache.hadoop.hdfs.server.namenode.FSDirStatAndListingOp.getContentSummaryInt(FSDirStatAndListingOp.java:493)
> at
> org.apache.hadoop.hdfs.server.namenode.FSDirStatAndListingOp.getContentSummary(FSDirStatAndListingOp.java:140)
> at
> org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getContentSummary(FSNamesystem.java:3995)
> at
> org.apache.hadoop.hdfs.server.namenode.NameNodeRpcServer.getContentSummary(NameNodeRpcServer.java:1202)
> at
> org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolServerSideTranslatorPB.getContentSummary(ClientNamenodeProtocolServerSideTranslatorPB.java:883)
> at
> org.apache.hadoop.hdfs.protocol.proto.ClientNamenodeProtocolProtos$ClientNamenodeProtocol$2.callBlockingMethod(ClientNamenodeProtocolProtos.java)
> at
> org.apache.hadoop.ipc.ProtobufRpcEngine$Server$ProtoBufRpcInvoker.call(ProtobufRpcEngine.java:616)
> at org.apache.hadoop.ipc.RPC$Server.call(RPC.java:982)
> at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2115)
> at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2111)
> at java.security.AccessController.doPrivileged(Native Method)
> at javax.security.auth.Subject.doAs(Subject.java:422)
> at
> org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1867)
> at org.apache.hadoop.ipc.Server$Handler.run(Server.java:2111)
> (state=08S01,code=1)
> {code}
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
