[jira] [Created] (HIVE-9821) Having the consistent physical execution plan , which using explain command with disable CBO and enable CBO.

2015-02-27 Thread asko (JIRA)
asko created HIVE-9821:
--

 Summary: Having the consistent physical execution plan  , which 
using explain command  with disable CBO and enable CBO.
 Key: HIVE-9821
 URL: https://issues.apache.org/jira/browse/HIVE-9821
 Project: Hive
  Issue Type: Bug
  Components: CBO
Affects Versions: 0.14.0
Reporter: asko
Priority: Critical


bq. Test case is( JOIN sub tree had been flatten after CBO in final plan stage 
of calcite optimizer) :
{quote}
--set  hive.cbo.enable=true;

--ANALYZE TABLE customer COMPUTE STATISTICS for columns;
--ANALYZE TABLE orders COMPUTE STATISTICS for columns;
--ANALYZE TABLE lineitem COMPUTE STATISTICS for columns;
--ANALYZE TABLE region COMPUTE STATISTICS for columns;
--ANALYZE TABLE supplier COMPUTE STATISTICS for columns;
--ANALYZE TABLE partsupp COMPUTE STATISTICS for columns;
--ANALYZE TABLE part COMPUTE STATISTICS for columns;
--ANALYZE TABLE nation COMPUTE STATISTICS for columns;

explain select
  o_year, sum(case when nation = 'BRAZIL' then volume else 0.0 end) / 
sum(volume) as mkt_share
from
  (
select
  year(o_orderdate) as o_year, l_extendedprice * (1-l_discount) as volume,
  n2.n_name as nation
from
nation n1 join region r
  on n1.n_regionkey = r.r_regionkey and r.r_name = 'AMERICA'
join customer c
  on c.c_nationkey = n1.n_nationkey
join orders o
  on c.c_custkey = o.o_custkey
join lineitem l
  on l.l_orderkey = o.o_orderkey and o.o_orderdate = '1995-01-01'
 and o.o_orderdate  '1996-12-31'
join part p
  on p.p_partkey = l.l_partkey and p.p_type = 'ECONOMY ANODIZED STEEL'
join supplier s
  on s.s_suppkey = l.l_suppkey
join  nation n2
  on s.s_nationkey = n2.n_nationkey
  ) all_nation
group by o_year
order by o_year;
{quote}
bq. This test from had modified q8 in TPC-H_full . Uncomment  could enable CBO. 
twice run results are same :
{quote}
STAGE DEPENDENCIES:
  Stage-1 is a root stage
  Stage-2 depends on stages: Stage-1, Stage-7
  Stage-3 depends on stages: Stage-2, Stage-10
  Stage-4 depends on stages: Stage-3
  Stage-5 depends on stages: Stage-4
  Stage-7 is a root stage
  Stage-9 is a root stage
  Stage-10 depends on stages: Stage-9, Stage-12
  Stage-12 is a root stage
  Stage-0 depends on stages: Stage-5

STAGE PLANS:
  Stage: Stage-1
Map Reduce
  Map Operator Tree:
  TableScan
alias: l
Statistics: Num rows: 27137974 Data size: 759863296 Basic stats: 
COMPLETE Column stats: NONE
Filter Operator
  predicate: ((l_partkey is not null and l_suppkey is not null) and 
l_orderkey is not null) (type: boolean)
  Statistics: Num rows: 3392247 Data size: 94982919 Basic stats: 
COMPLETE Column stats: NONE
  Select Operator
expressions: l_orderkey (type: int), l_partkey (type: int), 
l_suppkey (type: int), l_extendedprice (type: double), l_discount (type: double)
outputColumnNames: _col0, _col1, _col2, _col3, _col4
Statistics: Num rows: 3392247 Data size: 94982919 Basic stats: 
COMPLETE Column stats: NONE
Reduce Output Operator
  key expressions: _col1 (type: int)
  sort order: +
  Map-reduce partition columns: _col1 (type: int)
  Statistics: Num rows: 3392247 Data size: 94982919 Basic 
stats: COMPLETE Column stats: NONE
  value expressions: _col0 (type: int), _col2 (type: int), 
_col3 (type: double), _col4 (type: double)
  TableScan
alias: p
Statistics: Num rows: 928322 Data size: 24136384 Basic stats: 
COMPLETE Column stats: NONE
Filter Operator
  predicate: ((p_type = 'ECONOMY ANODIZED STEEL') and p_partkey is 
not null) (type: boolean)
  Statistics: Num rows: 232081 Data size: 6034109 Basic stats: 
COMPLETE Column stats: NONE
  Select Operator
expressions: p_partkey (type: int)
outputColumnNames: _col0
Statistics: Num rows: 232081 Data size: 6034109 Basic stats: 
COMPLETE Column stats: NONE
Reduce Output Operator
  key expressions: _col0 (type: int)
  sort order: +
  Map-reduce partition columns: _col0 (type: int)
  Statistics: Num rows: 232081 Data size: 6034109 Basic stats: 
COMPLETE Column stats: NONE
  Reduce Operator Tree:
Join Operator
  condition map:
   Inner Join 0 to 1
  keys:
0 _col1 (type: int)
1 _col0 (type: int)
  outputColumnNames: _col0, _col2, _col3, _col4
  Statistics: Num rows: 3731471 Data size: 104481213 Basic stats: 
COMPLETE Column stats: NONE
  File Output Operator
compressed: false
table:
input format: 

[jira] [Commented] (HIVE-9510) Throwing null point exception , when get join distinct row count from RelMdUtil.java class

2015-01-29 Thread asko (JIRA)

[ 
https://issues.apache.org/jira/browse/HIVE-9510?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14298018#comment-14298018
 ] 

asko commented on HIVE-9510:


Thanks! When i set org.apache.calcite.plan.RelOptPlanner.level=FINE ,  it could 
be executed correctly. So i  don't agree with you say.What do you think?

 Throwing null point exception , when get join distinct row count from 
 RelMdUtil.java class
 --

 Key: HIVE-9510
 URL: https://issues.apache.org/jira/browse/HIVE-9510
 Project: Hive
  Issue Type: Bug
  Components: CBO
Reporter: asko
Assignee: Julian Hyde
 Attachments: log.txt, log3_cbo5


 Setting log level in logging.properties file as following:
 {noformat}
 handlers=java.util.logging.ConsoleHandler.level=INFO
 org.apache.calcite.plan.RelOptPlanner.level=ALL
 java.util.logging.ConsoleHandler.level=ALL
 {noformat}
 Running Q3 in TPCH-full  after modifying , in order to  test join reorder,
 but running  failed.
 QL:
 {code:sql}
 set  hive.cbo.enable=true;
 --ANALYZE TABLE customer COMPUTE STATISTICS for columns;
 --ANALYZE TABLE orders COMPUTE STATISTICS for columns;
 --ANALYZE TABLE lineitem COMPUTE STATISTICS for columns;
 --Q3
 -- the query
 select 
   l_orderkey, sum(l_extendedprice*(1-l_discount)) as revenue, o_orderdate, 
 o_shippriority 
 from 
   lineitem l join orders o 
 on l.l_orderkey = o.o_orderkey
   join customer c
 on c.c_mktsegment = 'BUILDING' and c.c_custkey = o.o_custkey 
 where 
   o_orderdate  '1995-03-15' and l_shipdate  '1995-03-15' 
 group by l_orderkey, o_orderdate, o_shippriority 
 order by revenue desc, o_orderdate 
 limit 10;
 {code}
 LOG:
 see log.txt



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)