Gopal V created HIVE-14707:
------------------------------
Summary: ACID: Insert shuffle sort-merges on blank KEY
Key: HIVE-14707
URL: https://issues.apache.org/jira/browse/HIVE-14707
Project: Hive
Issue Type: Bug
Components: Transactions
Affects Versions: 2.2.0
Reporter: Gopal V
The ACID insert codepath uses a sorted shuffle, while they key used for shuffle
is always 0 bytes long.
{code}
hive (sales_acid)> explain insert into sales values(1, 2, '3400-0000-0000-009',
1, null);
STAGE PLANS:
Stage: Stage-1
Tez
DagId: gopal_20160906172626_80261c4c-79cc-4e02-87fe-3133be404e55:2
Edges:
Reducer 2 <- Map 1 (SIMPLE_EDGE)
...
Vertices:
Map 1
Map Operator Tree:
TableScan
alias: values__tmp__table__2
Statistics: Num rows: 1 Data size: 28 Basic stats: COMPLETE
Column stats: NONE
Select Operator
expressions: tmp_values_col1 (type: string),
tmp_values_col2 (type: string), tmp_values_col3 (type: string), tmp_values_col4
(type: string), tmp_values_col5 (type: string)
outputColumnNames: _col0, _col1, _col2, _col3, _col4
Statistics: Num rows: 1 Data size: 28 Basic stats: COMPLETE
Column stats: NONE
Reduce Output Operator
sort order:
Map-reduce partition columns: UDFToLong(_col1) (type:
bigint)
Statistics: Num rows: 1 Data size: 28 Basic stats:
COMPLETE Column stats: NONE
value expressions: _col0 (type: string), _col1 (type:
string), _col2 (type: string), _col3 (type: string), _col4 (type: string)
Execution mode: vectorized, llap
LLAP IO: no inputs
{code}
Note the missing "+" / "-" in the Sort Order fields.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
