[
https://issues.apache.org/jira/browse/HIVE-17973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16243612#comment-16243612
]
Zoltan Haindrich edited comment on HIVE-17973 at 11/8/17 9:49 AM:
------------------------------------------------------------------
the goal of the test is to check that hive doesn't let user1 to create a table
with a location under an unowned path.
earlier; it thrown some exception which was not present in the output - because
it contains the hdfs url...so the "masker" filters it out.
this semanticexception is unrelated to that goal; and shouldn't be accepted...
Earlier I've bisected this test to be broken by
5250ef450430fcdeed0a2cb7a770f48647987cd3 (HIVE-12408); but until now, I haven't
checked it thorough - I've just checked it and its true :)
the original exception was (which have been that sole masked line):
FAILED: HiveAccessControlException Permission denied: Principal [name=user1,
type=USER] does not have following privileges for operation CREATETABLE
[[OBJECT OWNERSHIP] on Object [type=DFS_URI,
name=hdfs://localhost:35753/tmp/ct_noperm_loc_foo0]]
I'm still not sure how that commit broke it...
Thanks for pinging me to check it! :)
I'll open a jira with these infos; (opened: HIVE-18012)
was (Author: kgyrtkirk):
the goal of the test is to check that hive doesn't let user1 to create a table
with a location under an unowned path.
earlier; it thrown some exception which was not present in the output - because
it contains the hdfs url...so the "masker" filters it out.
this semanticexception is unrelated to that goal; and shouldn't be accepted...
Earlier I've bisected this test to be broken by
5250ef450430fcdeed0a2cb7a770f48647987cd3 (HIVE-12408); but until now, I haven't
checked it thorough - I've just checked it and its true :)
the original exception was (which have been that sole masked line):
FAILED: HiveAccessControlException Permission denied: Principal [name=user1,
type=USER] does not have following privileges for operation CREATETABLE
[[OBJECT OWNERSHIP] on Object [type=DFS_URI,
name=hdfs://localhost:35753/tmp/ct_noperm_loc_foo0]]
I'm still not sure how that commit broke it...
Thanks for pinging me to check it! :)
I'll open a jira with these infos;
> Fix small bug in multi_insert_union_src.q
> -----------------------------------------
>
> Key: HIVE-17973
> URL: https://issues.apache.org/jira/browse/HIVE-17973
> Project: Hive
> Issue Type: Bug
> Reporter: liyunzhang
> Assignee: liyunzhang
> Priority: Trivial
> Attachments: HIVE-17973.2.patch, HIVE-17973.2.patch,
> HIVE-17973.patch, HVIE-17973.3.patch
>
>
> in ql\src\test\queries\clientpositive\multi_insert_union_src.q,
> There are two problems in the query file
> 1. It is strange to drop src_multi1 twice
> 2. {{src1}} is not created but used src1(Maybe we create src1 in other qfile)
> {code}
> set hive.mapred.mode=nonstrict;
> drop table if exists src2;
> drop table if exists src_multi1;
> drop table if exists src_multi1;
> set hive.stats.dbclass=fs;
> CREATE TABLE src2 as SELECT * FROM src;
> create table src_multi1 like src;
> create table src_multi2 like src;
> explain
> from (select * from src1 where key < 10 union all select * from src2 where
> key > 100) s
> insert overwrite table src_multi1 select key, value where key < 150 order by
> key
> insert overwrite table src_multi2 select key, value where key > 400 order by
> value;
> from (select * from src1 where key < 10 union all select * from src2 where
> key > 100) s
> insert overwrite table src_multi1 select key, value where key < 150 order by
> key
> insert overwrite table src_multi2 select key, value where key > 400 order by
> value;
> select * from src_multi1;
> select * from src_multi2;
> {code}
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
