[HIVE-18237] missing results for insert_only table after DP insert

• Type: Sub-task
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 3.0.0
• Component/s: Transactions
• Assignee: Steve Yeom Assign to me
• Reporter: Zoltan Haindrich
• Created: 06/Dec/17 12:21
• Updated: 22/May/18 23:15
• Resolved: 30/Jan/18 23:48

Description

set hive.stats.column.autogather=false;

set hive.exec.dynamic.partition.mode=nonstrict;
set hive.exec.max.dynamic.partitions.pernode=200;
set hive.exec.max.dynamic.partitions=200;
set hive.support.concurrency=true;
set hive.txn.manager=org.apache.hadoop.hive.ql.lockmgr.DbTxnManager;

create table i0 (p int,v int);
insert into i0 values
        (0,0),
        (2,2),
        (3,3);

create table p0 (v int) partitioned by (p int) stored as orc 
  tblproperties ("transactional"="true", "transactional_properties"="insert_only");

explain insert overwrite table p0 partition (p) select * from i0 where v < 3;
insert overwrite table p0 partition (p) select * from i0 where v < 3;
select count(*) from p0 where v!=1;

The table p0 should contain 2 rows at this point; but the result is 0.

• seems to be specific to insert_only tables
• the existing data appears if an insert into is executed.

Issue Links

• blocks: HIVE-13567 Enable auto-gather column stats by default
• is duplicated by: HIVE-18592 DP insert on insert only table causes StatTask to fail
• relates to: HIVE-18235 Columnstats gather on mm tables: re-enable disabled test