http://git-wip-us.apache.org/repos/asf/hive/blob/55887646/ql/src/test/results/clientpositive/perf/spark/query18.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/spark/query18.q.out b/ql/src/test/results/clientpositive/perf/spark/query18.q.out index e77a918..1d73576 100644 --- a/ql/src/test/results/clientpositive/perf/spark/query18.q.out +++ b/ql/src/test/results/clientpositive/perf/spark/query18.q.out @@ -106,15 +106,15 @@ STAGE PLANS: predicate: ((c_birth_month) IN (9, 5, 12, 4, 1, 10) and c_current_addr_sk is not null and c_current_cdemo_sk is not null and c_customer_sk is not null) (type: boolean) Statistics: Num rows: 80000000 Data size: 68801615852 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: c_customer_sk (type: int), c_current_cdemo_sk (type: int), c_current_addr_sk (type: int), c_birth_year (type: int) - outputColumnNames: _col0, _col1, _col2, _col4 + expressions: c_customer_sk (type: int), c_current_cdemo_sk (type: int), c_current_addr_sk (type: int), CAST( c_birth_year AS decimal(12,2)) (type: decimal(12,2)) + outputColumnNames: _col0, _col1, _col2, _col3 Statistics: Num rows: 80000000 Data size: 68801615852 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator key expressions: _col2 (type: int) sort order: + Map-reduce partition columns: _col2 (type: int) Statistics: Num rows: 80000000 Data size: 68801615852 Basic stats: COMPLETE Column stats: NONE - value expressions: _col0 (type: int), _col1 (type: int), _col4 (type: int) + value expressions: _col0 (type: int), _col1 (type: int), _col3 (type: decimal(12,2)) Execution mode: vectorized Map 13 Map Operator Tree: @@ -145,15 +145,15 @@ STAGE PLANS: predicate: ((cd_education_status = 'College') and (cd_gender = 'M') and cd_demo_sk is not null) (type: boolean) Statistics: Num rows: 465450 Data size: 179296539 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: cd_demo_sk (type: int), cd_dep_count (type: int) - outputColumnNames: _col0, _col3 + expressions: cd_demo_sk (type: int), CAST( cd_dep_count AS decimal(12,2)) (type: decimal(12,2)) + outputColumnNames: _col0, _col1 Statistics: Num rows: 465450 Data size: 179296539 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: 465450 Data size: 179296539 Basic stats: COMPLETE Column stats: NONE - value expressions: _col3 (type: int) + value expressions: _col1 (type: decimal(12,2)) Execution mode: vectorized Map 15 Map Operator Tree: @@ -224,7 +224,7 @@ STAGE PLANS: predicate: (cs_bill_cdemo_sk is not null and cs_bill_customer_sk is not null and cs_item_sk is not null and cs_sold_date_sk is not null) (type: boolean) Statistics: Num rows: 287989836 Data size: 38999608952 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: cs_sold_date_sk (type: int), cs_bill_customer_sk (type: int), cs_bill_cdemo_sk (type: int), cs_item_sk (type: int), cs_quantity (type: int), cs_list_price (type: decimal(7,2)), cs_sales_price (type: decimal(7,2)), cs_coupon_amt (type: decimal(7,2)), cs_net_profit (type: decimal(7,2)) + expressions: cs_sold_date_sk (type: int), cs_bill_customer_sk (type: int), cs_bill_cdemo_sk (type: int), cs_item_sk (type: int), CAST( cs_quantity AS decimal(12,2)) (type: decimal(12,2)), CAST( cs_list_price AS decimal(12,2)) (type: decimal(12,2)), CAST( cs_coupon_amt AS decimal(12,2)) (type: decimal(12,2)), CAST( cs_sales_price AS decimal(12,2)) (type: decimal(12,2)), CAST( cs_net_profit AS decimal(12,2)) (type: decimal(12,2)) outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6, _col7, _col8 Statistics: Num rows: 287989836 Data size: 38999608952 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator @@ -232,7 +232,7 @@ STAGE PLANS: sort order: + Map-reduce partition columns: _col0 (type: int) Statistics: Num rows: 287989836 Data size: 38999608952 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: int), _col2 (type: int), _col3 (type: int), _col4 (type: int), _col5 (type: decimal(7,2)), _col6 (type: decimal(7,2)), _col7 (type: decimal(7,2)), _col8 (type: decimal(7,2)) + value expressions: _col1 (type: int), _col2 (type: int), _col3 (type: int), _col4 (type: decimal(12,2)), _col5 (type: decimal(12,2)), _col6 (type: decimal(12,2)), _col7 (type: decimal(12,2)), _col8 (type: decimal(12,2)) Execution mode: vectorized Reducer 10 Reduce Operator Tree: @@ -249,7 +249,7 @@ STAGE PLANS: sort order: + Map-reduce partition columns: _col2 (type: int) Statistics: Num rows: 316788826 Data size: 42899570777 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: int), _col3 (type: int), _col4 (type: int), _col5 (type: decimal(7,2)), _col6 (type: decimal(7,2)), _col7 (type: decimal(7,2)), _col8 (type: decimal(7,2)) + value expressions: _col1 (type: int), _col3 (type: int), _col4 (type: decimal(12,2)), _col5 (type: decimal(12,2)), _col6 (type: decimal(12,2)), _col7 (type: decimal(12,2)), _col8 (type: decimal(12,2)) Reducer 11 Reduce Operator Tree: Join Operator @@ -258,14 +258,14 @@ STAGE PLANS: keys: 0 _col2 (type: int) 1 _col0 (type: int) - outputColumnNames: _col1, _col3, _col4, _col5, _col6, _col7, _col8, _col14 + outputColumnNames: _col1, _col3, _col4, _col5, _col6, _col7, _col8, _col11 Statistics: Num rows: 348467716 Data size: 47189528877 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator key expressions: _col3 (type: int) sort order: + Map-reduce partition columns: _col3 (type: int) Statistics: Num rows: 348467716 Data size: 47189528877 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: int), _col4 (type: int), _col5 (type: decimal(7,2)), _col6 (type: decimal(7,2)), _col7 (type: decimal(7,2)), _col8 (type: decimal(7,2)), _col14 (type: int) + value expressions: _col1 (type: int), _col4 (type: decimal(12,2)), _col5 (type: decimal(12,2)), _col6 (type: decimal(12,2)), _col7 (type: decimal(12,2)), _col8 (type: decimal(12,2)), _col11 (type: decimal(12,2)) Reducer 12 Reduce Operator Tree: Join Operator @@ -274,18 +274,18 @@ STAGE PLANS: keys: 0 _col3 (type: int) 1 _col0 (type: int) - outputColumnNames: _col1, _col4, _col5, _col6, _col7, _col8, _col14, _col16 + outputColumnNames: _col1, _col4, _col5, _col6, _col7, _col8, _col11, _col13 Statistics: Num rows: 383314495 Data size: 51908482889 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: _col16 (type: string), _col1 (type: int), _col4 (type: int), _col5 (type: decimal(7,2)), _col6 (type: decimal(7,2)), _col7 (type: decimal(7,2)), _col8 (type: decimal(7,2)), _col14 (type: int) - outputColumnNames: _col1, _col3, _col6, _col7, _col8, _col9, _col10, _col16 + expressions: _col13 (type: string), _col1 (type: int), _col4 (type: decimal(12,2)), _col5 (type: decimal(12,2)), _col6 (type: decimal(12,2)), _col7 (type: decimal(12,2)), _col8 (type: decimal(12,2)), _col11 (type: decimal(12,2)) + outputColumnNames: _col1, _col3, _col6, _col7, _col8, _col9, _col10, _col13 Statistics: Num rows: 383314495 Data size: 51908482889 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator key expressions: _col3 (type: int) sort order: + Map-reduce partition columns: _col3 (type: int) Statistics: Num rows: 383314495 Data size: 51908482889 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: string), _col6 (type: int), _col7 (type: decimal(7,2)), _col8 (type: decimal(7,2)), _col9 (type: decimal(7,2)), _col10 (type: decimal(7,2)), _col16 (type: int) + value expressions: _col1 (type: string), _col6 (type: decimal(12,2)), _col7 (type: decimal(12,2)), _col8 (type: decimal(12,2)), _col9 (type: decimal(12,2)), _col10 (type: decimal(12,2)), _col13 (type: decimal(12,2)) Reducer 2 Reduce Operator Tree: Join Operator @@ -294,14 +294,14 @@ STAGE PLANS: keys: 0 _col2 (type: int) 1 _col0 (type: int) - outputColumnNames: _col0, _col1, _col4, _col6, _col7, _col8 + outputColumnNames: _col0, _col1, _col3, _col5, _col6, _col7 Statistics: Num rows: 88000001 Data size: 75681779077 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: 88000001 Data size: 75681779077 Basic stats: COMPLETE Column stats: NONE - value expressions: _col0 (type: int), _col4 (type: int), _col6 (type: string), _col7 (type: string), _col8 (type: string) + value expressions: _col0 (type: int), _col3 (type: decimal(12,2)), _col5 (type: string), _col6 (type: string), _col7 (type: string) Reducer 3 Reduce Operator Tree: Join Operator @@ -310,14 +310,14 @@ STAGE PLANS: keys: 0 _col1 (type: int) 1 _col0 (type: int) - outputColumnNames: _col0, _col4, _col6, _col7, _col8 + outputColumnNames: _col0, _col3, _col5, _col6, _col7 Statistics: Num rows: 96800003 Data size: 83249958789 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: 96800003 Data size: 83249958789 Basic stats: COMPLETE Column stats: NONE - value expressions: _col4 (type: int), _col6 (type: string), _col7 (type: string), _col8 (type: string) + value expressions: _col3 (type: decimal(12,2)), _col5 (type: string), _col6 (type: string), _col7 (type: string) Reducer 4 Reduce Operator Tree: Join Operator @@ -326,24 +326,20 @@ STAGE PLANS: keys: 0 _col0 (type: int) 1 _col3 (type: int) - outputColumnNames: _col4, _col6, _col7, _col8, _col11, _col16, _col17, _col18, _col19, _col20, _col26 + outputColumnNames: _col3, _col5, _col6, _col7, _col10, _col15, _col16, _col17, _col18, _col19, _col22 Statistics: Num rows: 421645953 Data size: 57099332415 Basic stats: COMPLETE Column stats: NONE - Select Operator - expressions: _col11 (type: string), _col8 (type: string), _col7 (type: string), _col6 (type: string), CAST( _col16 AS decimal(12,2)) (type: decimal(12,2)), CAST( _col17 AS decimal(12,2)) (type: decimal(12,2)), CAST( _col19 AS decimal(12,2)) (type: decimal(12,2)), CAST( _col18 AS decimal(12,2)) (type: decimal(12,2)), CAST( _col20 AS decimal(12,2)) (type: decimal(12,2)), CAST( _col4 AS decimal(12,2)) (type: decimal(12,2)), CAST( _col26 AS decimal(12,2)) (type: decimal(12,2)) - outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6, _col7, _col8, _col9, _col10 - Statistics: Num rows: 421645953 Data size: 57099332415 Basic stats: COMPLETE Column stats: NONE - Group By Operator - aggregations: sum(_col4), count(_col4), sum(_col5), count(_col5), sum(_col6), count(_col6), sum(_col7), count(_col7), sum(_col8), count(_col8), sum(_col9), count(_col9), sum(_col10), count(_col10) - keys: _col0 (type: string), _col1 (type: string), _col2 (type: string), _col3 (type: string), 0L (type: bigint) - mode: hash - outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6, _col7, _col8, _col9, _col10, _col11, _col12, _col13, _col14, _col15, _col16, _col17, _col18 + Group By Operator + aggregations: sum(_col15), count(_col15), sum(_col16), count(_col16), sum(_col17), count(_col17), sum(_col18), count(_col18), sum(_col19), count(_col19), sum(_col3), count(_col3), sum(_col22), count(_col22) + keys: _col5 (type: string), _col6 (type: string), _col7 (type: string), _col10 (type: string), 0L (type: bigint) + mode: hash + outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6, _col7, _col8, _col9, _col10, _col11, _col12, _col13, _col14, _col15, _col16, _col17, _col18 + Statistics: Num rows: 2108229765 Data size: 285496662075 Basic stats: COMPLETE Column stats: NONE + Reduce Output Operator + key expressions: _col0 (type: string), _col1 (type: string), _col2 (type: string), _col3 (type: string), _col4 (type: bigint) + sort order: +++++ + Map-reduce partition columns: _col0 (type: string), _col1 (type: string), _col2 (type: string), _col3 (type: string), _col4 (type: bigint) Statistics: Num rows: 2108229765 Data size: 285496662075 Basic stats: COMPLETE Column stats: NONE - Reduce Output Operator - key expressions: _col0 (type: string), _col1 (type: string), _col2 (type: string), _col3 (type: string), _col4 (type: bigint) - sort order: +++++ - Map-reduce partition columns: _col0 (type: string), _col1 (type: string), _col2 (type: string), _col3 (type: string), _col4 (type: bigint) - Statistics: Num rows: 2108229765 Data size: 285496662075 Basic stats: COMPLETE Column stats: NONE - value expressions: _col5 (type: decimal(22,2)), _col6 (type: bigint), _col7 (type: decimal(22,2)), _col8 (type: bigint), _col9 (type: decimal(22,2)), _col10 (type: bigint), _col11 (type: decimal(22,2)), _col12 (type: bigint), _col13 (type: decimal(22,2)), _col14 (type: bigint), _col15 (type: decimal(22,2)), _col16 (type: bigint), _col17 (type: decimal(22,2)), _col18 (type: bigint) + value expressions: _col5 (type: decimal(22,2)), _col6 (type: bigint), _col7 (type: decimal(22,2)), _col8 (type: bigint), _col9 (type: decimal(22,2)), _col10 (type: bigint), _col11 (type: decimal(22,2)), _col12 (type: bigint), _col13 (type: decimal(22,2)), _col14 (type: bigint), _col15 (type: decimal(22,2)), _col16 (type: bigint), _col17 (type: decimal(22,2)), _col18 (type: bigint) Reducer 5 Execution mode: vectorized Reduce Operator Tree: @@ -355,7 +351,7 @@ STAGE PLANS: Statistics: Num rows: 1054114882 Data size: 142748330969 Basic stats: COMPLETE Column stats: NONE pruneGroupingSetId: true Select Operator - expressions: _col0 (type: string), _col1 (type: string), _col2 (type: string), _col3 (type: string), (_col5 / _col6) (type: decimal(38,18)), (_col7 / _col8) (type: decimal(38,18)), (_col9 / _col10) (type: decimal(38,18)), (_col11 / _col12) (type: decimal(38,18)), (_col13 / _col14) (type: decimal(38,18)), (_col15 / _col16) (type: decimal(38,18)), (_col17 / _col18) (type: decimal(38,18)) + expressions: _col3 (type: string), _col2 (type: string), _col1 (type: string), _col0 (type: string), (_col5 / _col6) (type: decimal(38,18)), (_col7 / _col8) (type: decimal(38,18)), (_col9 / _col10) (type: decimal(38,18)), (_col11 / _col12) (type: decimal(38,18)), (_col13 / _col14) (type: decimal(38,18)), (_col15 / _col16) (type: decimal(38,18)), (_col17 / _col18) (type: decimal(38,18)) outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6, _col7, _col8, _col9, _col10 Statistics: Num rows: 1054114882 Data size: 142748330969 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator
http://git-wip-us.apache.org/repos/asf/hive/blob/55887646/ql/src/test/results/clientpositive/perf/spark/query19.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/spark/query19.q.out b/ql/src/test/results/clientpositive/perf/spark/query19.q.out index 237157e..f6e29e6 100644 --- a/ql/src/test/results/clientpositive/perf/spark/query19.q.out +++ b/ql/src/test/results/clientpositive/perf/spark/query19.q.out @@ -80,7 +80,7 @@ STAGE PLANS: predicate: s_store_sk is not null (type: boolean) Statistics: Num rows: 1704 Data size: 3256276 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: s_store_sk (type: int), s_zip (type: string) + expressions: s_store_sk (type: int), substr(s_zip, 1, 5) (type: string) outputColumnNames: _col0, _col1 Statistics: Num rows: 1704 Data size: 3256276 Basic stats: COMPLETE Column stats: NONE Spark HashTable Sink Operator @@ -171,7 +171,7 @@ STAGE PLANS: predicate: ca_address_sk is not null (type: boolean) Statistics: Num rows: 40000000 Data size: 40595195284 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: ca_address_sk (type: int), ca_zip (type: string) + expressions: ca_address_sk (type: int), substr(ca_zip, 1, 5) (type: string) outputColumnNames: _col0, _col1 Statistics: Num rows: 40000000 Data size: 40595195284 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator @@ -227,7 +227,7 @@ STAGE PLANS: keys: 0 _col0 (type: int) 1 _col2 (type: int) - outputColumnNames: _col3, _col7, _col8, _col13, _col14, _col15, _col16 + outputColumnNames: _col3, _col7, _col8, _col11, _col12, _col13, _col14 Statistics: Num rows: 766650239 Data size: 67634106676 Basic stats: COMPLETE Column stats: NONE Map Join Operator condition map: @@ -235,20 +235,20 @@ STAGE PLANS: keys: 0 _col7 (type: int) 1 _col0 (type: int) - outputColumnNames: _col3, _col8, _col13, _col14, _col15, _col16, _col19 + outputColumnNames: _col3, _col8, _col11, _col12, _col13, _col14, _col16 input vertices: 1 Map 12 Statistics: Num rows: 843315281 Data size: 74397518956 Basic stats: COMPLETE Column stats: NONE Filter Operator - predicate: (substr(_col3, 1, 5) <> substr(_col19, 1, 5)) (type: boolean) + predicate: (_col3 <> _col16) (type: boolean) Statistics: Num rows: 843315281 Data size: 74397518956 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: _col8 (type: decimal(7,2)), _col13 (type: int), _col14 (type: string), _col15 (type: int), _col16 (type: string) - outputColumnNames: _col8, _col13, _col14, _col15, _col16 + expressions: _col8 (type: decimal(7,2)), _col11 (type: int), _col12 (type: string), _col13 (type: int), _col14 (type: string) + outputColumnNames: _col8, _col11, _col12, _col13, _col14 Statistics: Num rows: 843315281 Data size: 74397518956 Basic stats: COMPLETE Column stats: NONE Group By Operator aggregations: sum(_col8) - keys: _col14 (type: string), _col13 (type: int), _col15 (type: int), _col16 (type: string) + keys: _col12 (type: string), _col11 (type: int), _col13 (type: int), _col14 (type: string) mode: hash outputColumnNames: _col0, _col1, _col2, _col3, _col4 Statistics: Num rows: 843315281 Data size: 74397518956 Basic stats: COMPLETE Column stats: NONE @@ -317,14 +317,14 @@ STAGE PLANS: keys: 0 _col1 (type: int) 1 _col0 (type: int) - outputColumnNames: _col2, _col3, _col4, _col9, _col10, _col11, _col12 + outputColumnNames: _col2, _col3, _col4, _col7, _col8, _col9, _col10 Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator key expressions: _col2 (type: int) sort order: + Map-reduce partition columns: _col2 (type: int) Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - value expressions: _col3 (type: int), _col4 (type: decimal(7,2)), _col9 (type: int), _col10 (type: string), _col11 (type: int), _col12 (type: string) + value expressions: _col3 (type: int), _col4 (type: decimal(7,2)), _col7 (type: int), _col8 (type: string), _col9 (type: int), _col10 (type: string) Stage: Stage-0 Fetch Operator http://git-wip-us.apache.org/repos/asf/hive/blob/55887646/ql/src/test/results/clientpositive/perf/spark/query2.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/spark/query2.q.out b/ql/src/test/results/clientpositive/perf/spark/query2.q.out index 07e8c66..f623575 100644 --- a/ql/src/test/results/clientpositive/perf/spark/query2.q.out +++ b/ql/src/test/results/clientpositive/perf/spark/query2.q.out @@ -191,15 +191,15 @@ STAGE PLANS: predicate: (d_date_sk is not null and d_week_seq is not null) (type: boolean) Statistics: Num rows: 73049 Data size: 81741831 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: d_date_sk (type: int), d_week_seq (type: int), d_day_name (type: string) - outputColumnNames: _col0, _col1, _col2 + expressions: d_date_sk (type: int), d_week_seq (type: int), (d_day_name = 'Sunday') (type: boolean), (d_day_name = 'Monday') (type: boolean), (d_day_name = 'Tuesday') (type: boolean), (d_day_name = 'Wednesday') (type: boolean), (d_day_name = 'Thursday') (type: boolean), (d_day_name = 'Friday') (type: boolean), (d_day_name = 'Saturday') (type: boolean) + outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6, _col7, _col8 Statistics: Num rows: 73049 Data size: 81741831 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: 73049 Data size: 81741831 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: int), _col2 (type: string) + value expressions: _col1 (type: int), _col2 (type: boolean), _col3 (type: boolean), _col4 (type: boolean), _col5 (type: boolean), _col6 (type: boolean), _col7 (type: boolean), _col8 (type: boolean) Execution mode: vectorized Map 15 Map Operator Tree: @@ -250,15 +250,15 @@ STAGE PLANS: predicate: (d_date_sk is not null and d_week_seq is not null) (type: boolean) Statistics: Num rows: 73049 Data size: 81741831 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: d_date_sk (type: int), d_week_seq (type: int), d_day_name (type: string) - outputColumnNames: _col0, _col1, _col2 + expressions: d_date_sk (type: int), d_week_seq (type: int), (d_day_name = 'Sunday') (type: boolean), (d_day_name = 'Monday') (type: boolean), (d_day_name = 'Tuesday') (type: boolean), (d_day_name = 'Wednesday') (type: boolean), (d_day_name = 'Thursday') (type: boolean), (d_day_name = 'Friday') (type: boolean), (d_day_name = 'Saturday') (type: boolean) + outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6, _col7, _col8 Statistics: Num rows: 73049 Data size: 81741831 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: 73049 Data size: 81741831 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: int), _col2 (type: string) + value expressions: _col1 (type: int), _col2 (type: boolean), _col3 (type: boolean), _col4 (type: boolean), _col5 (type: boolean), _col6 (type: boolean), _col7 (type: boolean), _col8 (type: boolean) Execution mode: vectorized Map 8 Map Operator Tree: @@ -307,10 +307,10 @@ STAGE PLANS: keys: 0 _col0 (type: int) 1 _col0 (type: int) - outputColumnNames: _col1, _col3, _col4 + outputColumnNames: _col1, _col3, _col4, _col5, _col6, _col7, _col8, _col9, _col10 Statistics: Num rows: 475191764 Data size: 64437789277 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: _col3 (type: int), CASE WHEN ((_col4 = 'Sunday')) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN ((_col4 = 'Monday')) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN ((_col4 = 'Tuesday')) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN ((_col4 = 'Wednesday')) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN ((_col4 = 'Thursday')) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN ((_col4 = 'Friday')) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN ((_col4 = 'Saturday')) THEN (_col1) ELSE (null) END (type: decimal(7,2)) + expressions: _col3 (type: int), CASE WHEN (_col4) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN (_col5) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN (_col6) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN (_col7) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN (_col8) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN (_col9) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN (_col10) THEN (_col1) ELSE (null) END (type: decimal(7,2)) outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6, _col7 Statistics: Num rows: 475191764 Data size: 64437789277 Basic stats: COMPLETE Column stats: NONE Group By Operator @@ -364,10 +364,10 @@ STAGE PLANS: keys: 0 _col0 (type: int) 1 _col0 (type: int) - outputColumnNames: _col1, _col3, _col4 + outputColumnNames: _col1, _col3, _col4, _col5, _col6, _col7, _col8, _col9, _col10 Statistics: Num rows: 475191764 Data size: 64437789277 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: _col3 (type: int), CASE WHEN ((_col4 = 'Sunday')) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN ((_col4 = 'Monday')) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN ((_col4 = 'Tuesday')) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN ((_col4 = 'Wednesday')) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN ((_col4 = 'Thursday')) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN ((_col4 = 'Friday')) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN ((_col4 = 'Saturday')) THEN (_col1) ELSE (null) END (type: decimal(7,2)) + expressions: _col3 (type: int), CASE WHEN (_col4) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN (_col5) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN (_col6) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN (_col7) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN (_col8) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN (_col9) THEN (_col1) ELSE (null) END (type: decimal(7,2)), CASE WHEN (_col10) THEN (_col1) ELSE (null) END (type: decimal(7,2)) outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6, _col7 Statistics: Num rows: 475191764 Data size: 64437789277 Basic stats: COMPLETE Column stats: NONE Group By Operator @@ -407,10 +407,10 @@ STAGE PLANS: 0 _col0 (type: int) 1 _col0 (type: int) 2 (_col0 - 53) (type: int) - outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6, _col7, _col11, _col12, _col13, _col14, _col15, _col16, _col17 + outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6, _col7, _col10, _col11, _col12, _col13, _col14, _col15, _col16 Statistics: Num rows: 574982057 Data size: 77969728401 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: _col0 (type: int), round((_col1 / _col11), 2) (type: decimal(20,2)), round((_col2 / _col12), 2) (type: decimal(20,2)), round((_col3 / _col13), 2) (type: decimal(20,2)), round((_col4 / _col14), 2) (type: decimal(20,2)), round((_col5 / _col15), 2) (type: decimal(20,2)), round((_col6 / _col16), 2) (type: decimal(20,2)), round((_col7 / _col17), 2) (type: decimal(20,2)) + expressions: _col0 (type: int), round((_col1 / _col10), 2) (type: decimal(20,2)), round((_col2 / _col11), 2) (type: decimal(20,2)), round((_col3 / _col12), 2) (type: decimal(20,2)), round((_col4 / _col13), 2) (type: decimal(20,2)), round((_col5 / _col14), 2) (type: decimal(20,2)), round((_col6 / _col15), 2) (type: decimal(20,2)), round((_col7 / _col16), 2) (type: decimal(20,2)) outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6, _col7 Statistics: Num rows: 574982057 Data size: 77969728401 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator http://git-wip-us.apache.org/repos/asf/hive/blob/55887646/ql/src/test/results/clientpositive/perf/spark/query20.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/spark/query20.q.out b/ql/src/test/results/clientpositive/perf/spark/query20.q.out index f9c3ba5..953986d 100644 --- a/ql/src/test/results/clientpositive/perf/spark/query20.q.out +++ b/ql/src/test/results/clientpositive/perf/spark/query20.q.out @@ -162,11 +162,11 @@ STAGE PLANS: keys: 0 _col1 (type: int) 1 _col0 (type: int) - outputColumnNames: _col2, _col6, _col7, _col8, _col9, _col10 + outputColumnNames: _col2, _col5, _col6, _col7, _col8, _col9 Statistics: Num rows: 348467716 Data size: 47189528877 Basic stats: COMPLETE Column stats: NONE Group By Operator aggregations: sum(_col2) - keys: _col10 (type: string), _col9 (type: string), _col6 (type: string), _col7 (type: string), _col8 (type: decimal(7,2)) + keys: _col9 (type: string), _col8 (type: string), _col5 (type: string), _col6 (type: string), _col7 (type: decimal(7,2)) mode: hash outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5 Statistics: Num rows: 348467716 Data size: 47189528877 Basic stats: COMPLETE Column stats: NONE http://git-wip-us.apache.org/repos/asf/hive/blob/55887646/ql/src/test/results/clientpositive/perf/spark/query21.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/spark/query21.q.out b/ql/src/test/results/clientpositive/perf/spark/query21.q.out index 77fde06..6c2efe7 100644 --- a/ql/src/test/results/clientpositive/perf/spark/query21.q.out +++ b/ql/src/test/results/clientpositive/perf/spark/query21.q.out @@ -112,8 +112,8 @@ STAGE PLANS: predicate: (CAST( d_date AS TIMESTAMP) BETWEEN TIMESTAMP'1998-03-09 00:00:00' AND TIMESTAMP'1998-05-08 00:00:00' and d_date_sk is not null) (type: boolean) Statistics: Num rows: 8116 Data size: 9081804 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: d_date_sk (type: int), d_date (type: string) - outputColumnNames: _col0, _col1 + expressions: d_date_sk (type: int), (CAST( d_date AS DATE) < DATE'1998-04-08') (type: boolean), (CAST( d_date AS DATE) >= DATE'1998-04-08') (type: boolean) + outputColumnNames: _col0, _col1, _col2 Statistics: Num rows: 8116 Data size: 9081804 Basic stats: COMPLETE Column stats: NONE Spark HashTable Sink Operator keys: @@ -150,7 +150,7 @@ STAGE PLANS: keys: 0 _col0 (type: int) 1 _col0 (type: int) - outputColumnNames: _col1, _col2, _col3, _col5 + outputColumnNames: _col1, _col2, _col3, _col5, _col6 input vertices: 1 Map 5 Statistics: Num rows: 41342400 Data size: 653203228 Basic stats: COMPLETE Column stats: NONE @@ -159,7 +159,7 @@ STAGE PLANS: sort order: + Map-reduce partition columns: _col1 (type: int) Statistics: Num rows: 41342400 Data size: 653203228 Basic stats: COMPLETE Column stats: NONE - value expressions: _col2 (type: int), _col3 (type: int), _col5 (type: string) + value expressions: _col2 (type: int), _col3 (type: int), _col5 (type: boolean), _col6 (type: boolean) Execution mode: vectorized Local Work: Map Reduce Local Work @@ -193,7 +193,7 @@ STAGE PLANS: keys: 0 _col1 (type: int) 1 _col0 (type: int) - outputColumnNames: _col2, _col3, _col5, _col7 + outputColumnNames: _col2, _col3, _col5, _col6, _col8 Statistics: Num rows: 45476640 Data size: 718523566 Basic stats: COMPLETE Column stats: NONE Map Join Operator condition map: @@ -201,12 +201,12 @@ STAGE PLANS: keys: 0 _col2 (type: int) 1 _col0 (type: int) - outputColumnNames: _col3, _col5, _col7, _col10 + outputColumnNames: _col3, _col5, _col6, _col8, _col10 input vertices: 1 Map 7 Statistics: Num rows: 50024305 Data size: 790375939 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: _col10 (type: string), _col7 (type: string), CASE WHEN ((CAST( _col5 AS DATE) < DATE'1998-04-08')) THEN (_col3) ELSE (0) END (type: int), CASE WHEN ((CAST( _col5 AS DATE) >= DATE'1998-04-08')) THEN (_col3) ELSE (0) END (type: int) + expressions: _col10 (type: string), _col8 (type: string), CASE WHEN (_col5) THEN (_col3) ELSE (0) END (type: int), CASE WHEN (_col6) THEN (_col3) ELSE (0) END (type: int) outputColumnNames: _col0, _col1, _col2, _col3 Statistics: Num rows: 50024305 Data size: 790375939 Basic stats: COMPLETE Column stats: NONE Group By Operator http://git-wip-us.apache.org/repos/asf/hive/blob/55887646/ql/src/test/results/clientpositive/perf/spark/query22.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/spark/query22.q.out b/ql/src/test/results/clientpositive/perf/spark/query22.q.out index 8c9d8a5..2b5e7d0 100644 --- a/ql/src/test/results/clientpositive/perf/spark/query22.q.out +++ b/ql/src/test/results/clientpositive/perf/spark/query22.q.out @@ -185,11 +185,11 @@ STAGE PLANS: keys: 0 _col1 (type: int) 1 _col0 (type: int) - outputColumnNames: _col3, _col8, _col9, _col10, _col11 + outputColumnNames: _col3, _col7, _col8, _col9, _col10 Statistics: Num rows: 50024305 Data size: 790375939 Basic stats: COMPLETE Column stats: NONE Group By Operator aggregations: sum(_col3), count(_col3) - keys: _col8 (type: string), _col9 (type: string), _col10 (type: string), _col11 (type: string), 0L (type: bigint) + keys: _col7 (type: string), _col8 (type: string), _col9 (type: string), _col10 (type: string), 0L (type: bigint) mode: hash outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6 Statistics: Num rows: 250121525 Data size: 3951879695 Basic stats: COMPLETE Column stats: NONE http://git-wip-us.apache.org/repos/asf/hive/blob/55887646/ql/src/test/results/clientpositive/perf/spark/query23.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/spark/query23.q.out b/ql/src/test/results/clientpositive/perf/spark/query23.q.out index bfda977..e85f710 100644 --- a/ql/src/test/results/clientpositive/perf/spark/query23.q.out +++ b/ql/src/test/results/clientpositive/perf/spark/query23.q.out @@ -1,5 +1,5 @@ -Warning: Map Join MAPJOIN[285][bigTable=?] in task 'Stage-1:MAPRED' is a cross product -Warning: Map Join MAPJOIN[286][bigTable=?] in task 'Stage-1:MAPRED' is a cross product +Warning: Map Join MAPJOIN[275][bigTable=?] in task 'Stage-1:MAPRED' is a cross product +Warning: Map Join MAPJOIN[276][bigTable=?] in task 'Stage-1:MAPRED' is a cross product PREHOOK: query: explain with frequent_ss_items as (select substr(i_item_desc,1,30) itemdesc,i_item_sk item_sk,d_date solddate,count(*) cnt @@ -146,15 +146,15 @@ STAGE PLANS: predicate: (ss_customer_sk is not null and ss_sold_date_sk is not null) (type: boolean) Statistics: Num rows: 575995635 Data size: 50814502088 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: ss_sold_date_sk (type: int), ss_customer_sk (type: int), ss_quantity (type: int), ss_sales_price (type: decimal(7,2)) - outputColumnNames: _col0, _col1, _col2, _col3 + expressions: ss_sold_date_sk (type: int), ss_customer_sk (type: int), (CAST( ss_quantity AS decimal(10,0)) * ss_sales_price) (type: decimal(18,2)) + outputColumnNames: _col0, _col1, _col2 Statistics: Num rows: 575995635 Data size: 50814502088 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: 575995635 Data size: 50814502088 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: int), _col2 (type: int), _col3 (type: decimal(7,2)) + value expressions: _col1 (type: int), _col2 (type: decimal(18,2)) Execution mode: vectorized Map 18 Map Operator Tree: @@ -204,15 +204,15 @@ STAGE PLANS: predicate: (ss_customer_sk is not null and ss_sold_date_sk is not null) (type: boolean) Statistics: Num rows: 575995635 Data size: 50814502088 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: ss_sold_date_sk (type: int), ss_customer_sk (type: int), ss_quantity (type: int), ss_sales_price (type: decimal(7,2)) - outputColumnNames: _col0, _col1, _col2, _col3 + expressions: ss_sold_date_sk (type: int), ss_customer_sk (type: int), (CAST( ss_quantity AS decimal(10,0)) * ss_sales_price) (type: decimal(18,2)) + outputColumnNames: _col0, _col1, _col2 Statistics: Num rows: 575995635 Data size: 50814502088 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: 575995635 Data size: 50814502088 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: int), _col2 (type: int), _col3 (type: decimal(7,2)) + value expressions: _col1 (type: int), _col2 (type: decimal(18,2)) Execution mode: vectorized Map 25 Map Operator Tree: @@ -260,14 +260,14 @@ STAGE PLANS: keys: 0 _col0 (type: int) 1 _col0 (type: int) - outputColumnNames: _col1, _col2, _col3 + outputColumnNames: _col1, _col2 Statistics: Num rows: 633595212 Data size: 55895953508 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: 633595212 Data size: 55895953508 Basic stats: COMPLETE Column stats: NONE - value expressions: _col2 (type: int), _col3 (type: decimal(7,2)) + value expressions: _col2 (type: decimal(18,2)) Reducer 15 Reduce Operator Tree: Join Operator @@ -276,24 +276,20 @@ STAGE PLANS: keys: 0 _col1 (type: int) 1 _col0 (type: int) - outputColumnNames: _col2, _col3, _col6 + outputColumnNames: _col2, _col4 Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - Select Operator - expressions: _col6 (type: int), (CAST( _col2 AS decimal(10,0)) * _col3) (type: decimal(18,2)) + Group By Operator + aggregations: sum(_col2) + keys: _col4 (type: int) + mode: hash outputColumnNames: _col0, _col1 Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - Group By Operator - aggregations: sum(_col1) - keys: _col0 (type: int) - mode: hash - outputColumnNames: _col0, _col1 + Reduce Output Operator + key expressions: _col0 (type: int) + sort order: + + Map-reduce partition columns: _col0 (type: int) Statistics: Num rows: 696954748 Data size: 61485550191 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: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: decimal(28,2)) + value expressions: _col1 (type: decimal(28,2)) Reducer 16 Execution mode: vectorized Reduce Operator Tree: @@ -351,14 +347,14 @@ STAGE PLANS: keys: 0 _col0 (type: int) 1 _col0 (type: int) - outputColumnNames: _col1, _col2, _col3 + outputColumnNames: _col1, _col2 Statistics: Num rows: 633595212 Data size: 55895953508 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: 633595212 Data size: 55895953508 Basic stats: COMPLETE Column stats: NONE - value expressions: _col2 (type: int), _col3 (type: decimal(7,2)) + value expressions: _col2 (type: decimal(18,2)) Reducer 22 Reduce Operator Tree: Join Operator @@ -367,24 +363,20 @@ STAGE PLANS: keys: 0 _col1 (type: int) 1 _col0 (type: int) - outputColumnNames: _col2, _col3, _col6 + outputColumnNames: _col2, _col4 Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - Select Operator - expressions: _col6 (type: int), (CAST( _col2 AS decimal(10,0)) * _col3) (type: decimal(18,2)) + Group By Operator + aggregations: sum(_col2) + keys: _col4 (type: int) + mode: hash outputColumnNames: _col0, _col1 Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - Group By Operator - aggregations: sum(_col1) - keys: _col0 (type: int) - mode: hash - outputColumnNames: _col0, _col1 + Reduce Output Operator + key expressions: _col0 (type: int) + sort order: + + Map-reduce partition columns: _col0 (type: int) Statistics: Num rows: 696954748 Data size: 61485550191 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: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: decimal(28,2)) + value expressions: _col1 (type: decimal(28,2)) Reducer 23 Execution mode: vectorized Reduce Operator Tree: @@ -446,15 +438,15 @@ STAGE PLANS: predicate: (ss_customer_sk is not null and ss_sold_date_sk is not null) (type: boolean) Statistics: Num rows: 575995635 Data size: 50814502088 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: ss_sold_date_sk (type: int), ss_customer_sk (type: int), ss_quantity (type: int), ss_sales_price (type: decimal(7,2)) - outputColumnNames: _col0, _col1, _col2, _col3 + expressions: ss_sold_date_sk (type: int), ss_customer_sk (type: int), (CAST( ss_quantity AS decimal(10,0)) * ss_sales_price) (type: decimal(18,2)) + outputColumnNames: _col0, _col1, _col2 Statistics: Num rows: 575995635 Data size: 50814502088 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: 575995635 Data size: 50814502088 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: int), _col2 (type: int), _col3 (type: decimal(7,2)) + value expressions: _col1 (type: int), _col2 (type: decimal(18,2)) Execution mode: vectorized Map 47 Map Operator Tree: @@ -504,15 +496,15 @@ STAGE PLANS: predicate: (ss_customer_sk is not null and ss_sold_date_sk is not null) (type: boolean) Statistics: Num rows: 575995635 Data size: 50814502088 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: ss_sold_date_sk (type: int), ss_customer_sk (type: int), ss_quantity (type: int), ss_sales_price (type: decimal(7,2)) - outputColumnNames: _col0, _col1, _col2, _col3 + expressions: ss_sold_date_sk (type: int), ss_customer_sk (type: int), (CAST( ss_quantity AS decimal(10,0)) * ss_sales_price) (type: decimal(18,2)) + outputColumnNames: _col0, _col1, _col2 Statistics: Num rows: 575995635 Data size: 50814502088 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: 575995635 Data size: 50814502088 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: int), _col2 (type: int), _col3 (type: decimal(7,2)) + value expressions: _col1 (type: int), _col2 (type: decimal(18,2)) Execution mode: vectorized Map 54 Map Operator Tree: @@ -560,14 +552,14 @@ STAGE PLANS: keys: 0 _col0 (type: int) 1 _col0 (type: int) - outputColumnNames: _col1, _col2, _col3 + outputColumnNames: _col1, _col2 Statistics: Num rows: 633595212 Data size: 55895953508 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: 633595212 Data size: 55895953508 Basic stats: COMPLETE Column stats: NONE - value expressions: _col2 (type: int), _col3 (type: decimal(7,2)) + value expressions: _col2 (type: decimal(18,2)) Reducer 44 Reduce Operator Tree: Join Operator @@ -576,24 +568,20 @@ STAGE PLANS: keys: 0 _col1 (type: int) 1 _col0 (type: int) - outputColumnNames: _col2, _col3, _col6 + outputColumnNames: _col2, _col4 Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - Select Operator - expressions: _col6 (type: int), (CAST( _col2 AS decimal(10,0)) * _col3) (type: decimal(18,2)) + Group By Operator + aggregations: sum(_col2) + keys: _col4 (type: int) + mode: hash outputColumnNames: _col0, _col1 Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - Group By Operator - aggregations: sum(_col1) - keys: _col0 (type: int) - mode: hash - outputColumnNames: _col0, _col1 + Reduce Output Operator + key expressions: _col0 (type: int) + sort order: + + Map-reduce partition columns: _col0 (type: int) Statistics: Num rows: 696954748 Data size: 61485550191 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: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: decimal(28,2)) + value expressions: _col1 (type: decimal(28,2)) Reducer 45 Execution mode: vectorized Reduce Operator Tree: @@ -651,14 +639,14 @@ STAGE PLANS: keys: 0 _col0 (type: int) 1 _col0 (type: int) - outputColumnNames: _col1, _col2, _col3 + outputColumnNames: _col1, _col2 Statistics: Num rows: 633595212 Data size: 55895953508 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: 633595212 Data size: 55895953508 Basic stats: COMPLETE Column stats: NONE - value expressions: _col2 (type: int), _col3 (type: decimal(7,2)) + value expressions: _col2 (type: decimal(18,2)) Reducer 51 Reduce Operator Tree: Join Operator @@ -667,24 +655,20 @@ STAGE PLANS: keys: 0 _col1 (type: int) 1 _col0 (type: int) - outputColumnNames: _col2, _col3, _col6 + outputColumnNames: _col2, _col4 Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - Select Operator - expressions: _col6 (type: int), (CAST( _col2 AS decimal(10,0)) * _col3) (type: decimal(18,2)) + Group By Operator + aggregations: sum(_col2) + keys: _col4 (type: int) + mode: hash outputColumnNames: _col0, _col1 Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - Group By Operator - aggregations: sum(_col1) - keys: _col0 (type: int) - mode: hash - outputColumnNames: _col0, _col1 + Reduce Output Operator + key expressions: _col0 (type: int) + sort order: + + Map-reduce partition columns: _col0 (type: int) Statistics: Num rows: 696954748 Data size: 61485550191 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: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: decimal(28,2)) + value expressions: _col1 (type: decimal(28,2)) Reducer 52 Execution mode: vectorized Reduce Operator Tree: @@ -791,7 +775,7 @@ STAGE PLANS: predicate: i_item_sk is not null (type: boolean) Statistics: Num rows: 462000 Data size: 663560457 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: i_item_sk (type: int), i_item_desc (type: string) + expressions: i_item_sk (type: int), substr(i_item_desc, 1, 30) (type: string) outputColumnNames: _col0, _col1 Statistics: Num rows: 462000 Data size: 663560457 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator @@ -811,15 +795,15 @@ STAGE PLANS: predicate: ss_customer_sk is not null (type: boolean) Statistics: Num rows: 575995635 Data size: 50814502088 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: ss_customer_sk (type: int), ss_quantity (type: int), ss_sales_price (type: decimal(7,2)) - outputColumnNames: _col0, _col1, _col2 + expressions: ss_customer_sk (type: int), (CAST( ss_quantity AS decimal(10,0)) * ss_sales_price) (type: decimal(18,2)) + outputColumnNames: _col0, _col1 Statistics: Num rows: 575995635 Data size: 50814502088 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: 575995635 Data size: 50814502088 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: int), _col2 (type: decimal(7,2)) + value expressions: _col1 (type: decimal(18,2)) Execution mode: vectorized Map 30 Map Operator Tree: @@ -910,13 +894,13 @@ STAGE PLANS: Statistics: Num rows: 348477374 Data size: 30742775095 Basic stats: COMPLETE Column stats: NONE Select Operator expressions: _col0 (type: int), _col3 (type: bigint) - outputColumnNames: _col0, _col3 + outputColumnNames: _col1, _col3 Statistics: Num rows: 348477374 Data size: 30742775095 Basic stats: COMPLETE Column stats: NONE Filter Operator predicate: (_col3 > 4L) (type: boolean) Statistics: Num rows: 116159124 Data size: 10247591639 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: _col0 (type: int) + expressions: _col1 (type: int) outputColumnNames: _col1 Statistics: Num rows: 116159124 Data size: 10247591639 Basic stats: COMPLETE Column stats: NONE Group By Operator @@ -953,24 +937,20 @@ STAGE PLANS: keys: 0 _col0 (type: int) 1 _col0 (type: int) - outputColumnNames: _col1, _col2, _col3 + outputColumnNames: _col1, _col2 Statistics: Num rows: 633595212 Data size: 55895953508 Basic stats: COMPLETE Column stats: NONE - Select Operator - expressions: _col3 (type: int), (CAST( _col1 AS decimal(10,0)) * _col2) (type: decimal(18,2)) + Group By Operator + aggregations: sum(_col1) + keys: _col2 (type: int) + mode: hash outputColumnNames: _col0, _col1 Statistics: Num rows: 633595212 Data size: 55895953508 Basic stats: COMPLETE Column stats: NONE - Group By Operator - aggregations: sum(_col1) - keys: _col0 (type: int) - mode: hash - outputColumnNames: _col0, _col1 + Reduce Output Operator + key expressions: _col0 (type: int) + sort order: + + Map-reduce partition columns: _col0 (type: int) Statistics: Num rows: 633595212 Data size: 55895953508 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: 633595212 Data size: 55895953508 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: decimal(28,2)) + value expressions: _col1 (type: decimal(28,2)) Reducer 29 Execution mode: vectorized Local Work: @@ -1100,24 +1080,20 @@ STAGE PLANS: keys: 0 _col1 (type: int) 1 _col0 (type: int) - outputColumnNames: _col3, _col5, _col6 + outputColumnNames: _col3, _col4, _col5 Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - Select Operator - expressions: substr(_col6, 1, 30) (type: string), _col5 (type: int), _col3 (type: string) - outputColumnNames: _col0, _col1, _col2 + Group By Operator + aggregations: count() + keys: _col4 (type: int), _col3 (type: string), _col5 (type: string) + mode: hash + outputColumnNames: _col0, _col1, _col2, _col3 Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - Group By Operator - aggregations: count() - keys: _col1 (type: int), _col0 (type: string), _col2 (type: string) - mode: hash - outputColumnNames: _col0, _col1, _col2, _col3 + Reduce Output Operator + key expressions: _col0 (type: int), _col1 (type: string), _col2 (type: string) + sort order: +++ + Map-reduce partition columns: _col0 (type: int) Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - Reduce Output Operator - key expressions: _col0 (type: int), _col1 (type: string), _col2 (type: string) - sort order: +++ - Map-reduce partition columns: _col0 (type: int) - Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - value expressions: _col3 (type: bigint) + value expressions: _col3 (type: bigint) Reducer 4 Reduce Operator Tree: Join Operator http://git-wip-us.apache.org/repos/asf/hive/blob/55887646/ql/src/test/results/clientpositive/perf/spark/query24.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/spark/query24.q.out b/ql/src/test/results/clientpositive/perf/spark/query24.q.out index 91fe702..2d57eca 100644 --- a/ql/src/test/results/clientpositive/perf/spark/query24.q.out +++ b/ql/src/test/results/clientpositive/perf/spark/query24.q.out @@ -136,12 +136,12 @@ STAGE PLANS: Statistics: Num rows: 852 Data size: 1628138 Basic stats: COMPLETE Column stats: NONE Select Operator expressions: s_store_sk (type: int), s_store_name (type: string), s_state (type: string), s_zip (type: string) - outputColumnNames: _col0, _col1, _col3, _col4 + outputColumnNames: _col0, _col1, _col2, _col3 Statistics: Num rows: 852 Data size: 1628138 Basic stats: COMPLETE Column stats: NONE Spark HashTable Sink Operator keys: 0 _col2 (type: string) - 1 _col4 (type: string) + 1 _col3 (type: string) Execution mode: vectorized Local Work: Map Reduce Local Work @@ -167,7 +167,7 @@ STAGE PLANS: predicate: (ca_address_sk is not null and ca_zip is not null) (type: boolean) Statistics: Num rows: 40000000 Data size: 40595195284 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: ca_address_sk (type: int), ca_state (type: string), ca_zip (type: string), ca_country (type: string) + expressions: ca_address_sk (type: int), ca_state (type: string), ca_zip (type: string), upper(ca_country) (type: string) outputColumnNames: _col0, _col1, _col2, _col3 Statistics: Num rows: 40000000 Data size: 40595195284 Basic stats: COMPLETE Column stats: NONE Map Join Operator @@ -175,8 +175,8 @@ STAGE PLANS: Inner Join 0 to 1 keys: 0 _col2 (type: string) - 1 _col4 (type: string) - outputColumnNames: _col0, _col1, _col3, _col4, _col5, _col7 + 1 _col3 (type: string) + outputColumnNames: _col0, _col1, _col3, _col4, _col5, _col6 input vertices: 1 Map 19 Statistics: Num rows: 44000000 Data size: 44654715780 Basic stats: COMPLETE Column stats: NONE @@ -185,7 +185,7 @@ STAGE PLANS: sort order: + Map-reduce partition columns: _col0 (type: int) Statistics: Num rows: 44000000 Data size: 44654715780 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: string), _col3 (type: string), _col4 (type: int), _col5 (type: string), _col7 (type: string) + value expressions: _col1 (type: string), _col3 (type: string), _col4 (type: int), _col5 (type: string), _col6 (type: string) Execution mode: vectorized Local Work: Map Reduce Local Work @@ -276,62 +276,62 @@ STAGE PLANS: keys: 0 _col0 (type: int) 1 _col1 (type: int) - outputColumnNames: _col1, _col3, _col4, _col5, _col7, _col9, _col11, _col12, _col13 + outputColumnNames: _col1, _col3, _col4, _col5, _col6, _col8, _col10, _col11, _col12 Statistics: Num rows: 88000001 Data size: 75681779077 Basic stats: COMPLETE Column stats: NONE Filter Operator - predicate: (_col13 <> upper(_col3)) (type: boolean) + predicate: (_col12 <> _col3) (type: boolean) Statistics: Num rows: 88000001 Data size: 75681779077 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator - key expressions: _col9 (type: int), _col4 (type: int) + key expressions: _col8 (type: int), _col4 (type: int) sort order: ++ - Map-reduce partition columns: _col9 (type: int), _col4 (type: int) + Map-reduce partition columns: _col8 (type: int), _col4 (type: int) Statistics: Num rows: 88000001 Data size: 75681779077 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: string), _col5 (type: string), _col7 (type: string), _col11 (type: string), _col12 (type: string) + value expressions: _col1 (type: string), _col5 (type: string), _col6 (type: string), _col10 (type: string), _col11 (type: string) Reducer 14 Reduce Operator Tree: Join Operator condition map: Inner Join 0 to 1 keys: - 0 _col9 (type: int), _col4 (type: int) + 0 _col8 (type: int), _col4 (type: int) 1 _col1 (type: int), _col2 (type: int) - outputColumnNames: _col1, _col5, _col7, _col11, _col12, _col14, _col17, _col18 + outputColumnNames: _col1, _col5, _col6, _col10, _col11, _col13, _col16, _col17 Statistics: Num rows: 633595212 Data size: 55895953508 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator - key expressions: _col14 (type: int) + key expressions: _col13 (type: int) sort order: + - Map-reduce partition columns: _col14 (type: int) + Map-reduce partition columns: _col13 (type: int) Statistics: Num rows: 633595212 Data size: 55895953508 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: string), _col5 (type: string), _col7 (type: string), _col11 (type: string), _col12 (type: string), _col17 (type: int), _col18 (type: decimal(7,2)) + value expressions: _col1 (type: string), _col5 (type: string), _col6 (type: string), _col10 (type: string), _col11 (type: string), _col16 (type: int), _col17 (type: decimal(7,2)) Reducer 15 Reduce Operator Tree: Join Operator condition map: Inner Join 0 to 1 keys: - 0 _col14 (type: int) + 0 _col13 (type: int) 1 _col0 (type: int) - outputColumnNames: _col1, _col5, _col7, _col11, _col12, _col14, _col17, _col18, _col20, _col21, _col22, _col23, _col24 + outputColumnNames: _col1, _col5, _col6, _col10, _col11, _col13, _col16, _col17, _col19, _col20, _col21, _col22, _col23 Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator - key expressions: _col14 (type: int), _col17 (type: int) + key expressions: _col13 (type: int), _col16 (type: int) sort order: ++ - Map-reduce partition columns: _col14 (type: int), _col17 (type: int) + Map-reduce partition columns: _col13 (type: int), _col16 (type: int) Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: string), _col5 (type: string), _col7 (type: string), _col11 (type: string), _col12 (type: string), _col18 (type: decimal(7,2)), _col20 (type: decimal(7,2)), _col21 (type: string), _col22 (type: string), _col23 (type: string), _col24 (type: int) + value expressions: _col1 (type: string), _col5 (type: string), _col6 (type: string), _col10 (type: string), _col11 (type: string), _col17 (type: decimal(7,2)), _col19 (type: decimal(7,2)), _col20 (type: string), _col21 (type: string), _col22 (type: string), _col23 (type: int) Reducer 16 Reduce Operator Tree: Join Operator condition map: Inner Join 0 to 1 keys: - 0 _col14 (type: int), _col17 (type: int) + 0 _col13 (type: int), _col16 (type: int) 1 _col0 (type: int), _col1 (type: int) - outputColumnNames: _col1, _col5, _col7, _col11, _col12, _col18, _col20, _col21, _col22, _col23, _col24 + outputColumnNames: _col1, _col5, _col6, _col10, _col11, _col17, _col19, _col20, _col21, _col22, _col23 Statistics: Num rows: 766650239 Data size: 67634106676 Basic stats: COMPLETE Column stats: NONE Group By Operator - aggregations: sum(_col18) - keys: _col11 (type: string), _col12 (type: string), _col1 (type: string), _col5 (type: string), _col7 (type: string), _col20 (type: decimal(7,2)), _col21 (type: string), _col22 (type: string), _col23 (type: string), _col24 (type: int) + aggregations: sum(_col17) + keys: _col10 (type: string), _col11 (type: string), _col1 (type: string), _col5 (type: string), _col6 (type: string), _col19 (type: decimal(7,2)), _col20 (type: string), _col21 (type: string), _col22 (type: string), _col23 (type: int) mode: hash outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6, _col7, _col8, _col9, _col10 Statistics: Num rows: 766650239 Data size: 67634106676 Basic stats: COMPLETE Column stats: NONE @@ -397,12 +397,12 @@ STAGE PLANS: Statistics: Num rows: 852 Data size: 1628138 Basic stats: COMPLETE Column stats: NONE Select Operator expressions: s_store_sk (type: int), s_store_name (type: string), s_state (type: string), s_zip (type: string) - outputColumnNames: _col0, _col1, _col3, _col4 + outputColumnNames: _col0, _col1, _col2, _col3 Statistics: Num rows: 852 Data size: 1628138 Basic stats: COMPLETE Column stats: NONE Spark HashTable Sink Operator keys: 0 _col2 (type: string) - 1 _col4 (type: string) + 1 _col3 (type: string) Execution mode: vectorized Local Work: Map Reduce Local Work @@ -487,14 +487,14 @@ STAGE PLANS: Statistics: Num rows: 231000 Data size: 331780228 Basic stats: COMPLETE Column stats: NONE Select Operator expressions: i_item_sk (type: int), i_current_price (type: decimal(7,2)), i_size (type: string), i_units (type: string), i_manager_id (type: int) - outputColumnNames: _col0, _col1, _col2, _col4, _col5 + outputColumnNames: _col0, _col1, _col2, _col3, _col4 Statistics: Num rows: 231000 Data size: 331780228 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: 231000 Data size: 331780228 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: decimal(7,2)), _col2 (type: string), _col4 (type: string), _col5 (type: int) + value expressions: _col1 (type: decimal(7,2)), _col2 (type: string), _col3 (type: string), _col4 (type: int) Execution mode: vectorized Map 7 Map Operator Tree: @@ -506,7 +506,7 @@ STAGE PLANS: predicate: (ca_address_sk is not null and ca_zip is not null) (type: boolean) Statistics: Num rows: 40000000 Data size: 40595195284 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: ca_address_sk (type: int), ca_state (type: string), ca_zip (type: string), ca_country (type: string) + expressions: ca_address_sk (type: int), ca_state (type: string), ca_zip (type: string), upper(ca_country) (type: string) outputColumnNames: _col0, _col1, _col2, _col3 Statistics: Num rows: 40000000 Data size: 40595195284 Basic stats: COMPLETE Column stats: NONE Map Join Operator @@ -514,8 +514,8 @@ STAGE PLANS: Inner Join 0 to 1 keys: 0 _col2 (type: string) - 1 _col4 (type: string) - outputColumnNames: _col0, _col1, _col3, _col4, _col5, _col7 + 1 _col3 (type: string) + outputColumnNames: _col0, _col1, _col3, _col4, _col5, _col6 input vertices: 1 Map 9 Statistics: Num rows: 44000000 Data size: 44654715780 Basic stats: COMPLETE Column stats: NONE @@ -524,7 +524,7 @@ STAGE PLANS: sort order: + Map-reduce partition columns: _col0 (type: int) Statistics: Num rows: 44000000 Data size: 44654715780 Basic stats: COMPLETE Column stats: NONE - value expressions: _col1 (type: string), _col3 (type: string), _col4 (type: int), _col5 (type: string), _col7 (type: string) + value expressions: _col1 (type: string), _col3 (type: string), _col4 (type: int), _col5 (type: string), _col6 (type: string) Execution mode: vectorized Local Work: Map Reduce Local Work @@ -536,14 +536,14 @@ STAGE PLANS: keys: 0 _col0 (type: int) 1 _col0 (type: int) - outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col6, _col7, _col9, _col10 + outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col6, _col7, _col8, _col9 Statistics: Num rows: 633595212 Data size: 55895953508 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator key expressions: _col1 (type: int), _col2 (type: int) sort order: ++ Map-reduce partition columns: _col1 (type: int), _col2 (type: int) Statistics: Num rows: 633595212 Data size: 55895953508 Basic stats: COMPLETE Column stats: NONE - value expressions: _col0 (type: int), _col3 (type: int), _col4 (type: decimal(7,2)), _col6 (type: decimal(7,2)), _col7 (type: string), _col9 (type: string), _col10 (type: int) + value expressions: _col0 (type: int), _col3 (type: int), _col4 (type: decimal(7,2)), _col6 (type: decimal(7,2)), _col7 (type: string), _col8 (type: string), _col9 (type: int) Reducer 3 Reduce Operator Tree: Join Operator @@ -552,14 +552,14 @@ STAGE PLANS: keys: 0 _col1 (type: int), _col2 (type: int) 1 _col0 (type: int), _col9 (type: int) - outputColumnNames: _col0, _col3, _col4, _col6, _col7, _col9, _col10, _col13, _col14, _col17, _col21, _col23 + outputColumnNames: _col0, _col3, _col4, _col6, _col7, _col8, _col9, _col12, _col13, _col16, _col20, _col21 Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator key expressions: _col0 (type: int), _col3 (type: int) sort order: ++ Map-reduce partition columns: _col0 (type: int), _col3 (type: int) Statistics: Num rows: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - value expressions: _col4 (type: decimal(7,2)), _col6 (type: decimal(7,2)), _col7 (type: string), _col9 (type: string), _col10 (type: int), _col13 (type: string), _col14 (type: string), _col17 (type: string), _col21 (type: string), _col23 (type: string) + value expressions: _col4 (type: decimal(7,2)), _col6 (type: decimal(7,2)), _col7 (type: string), _col8 (type: string), _col9 (type: int), _col12 (type: string), _col13 (type: string), _col16 (type: string), _col20 (type: string), _col21 (type: string) Reducer 4 Reduce Operator Tree: Join Operator @@ -568,11 +568,11 @@ STAGE PLANS: keys: 0 _col0 (type: int), _col3 (type: int) 1 _col0 (type: int), _col1 (type: int) - outputColumnNames: _col4, _col6, _col7, _col9, _col10, _col13, _col14, _col17, _col21, _col23 + outputColumnNames: _col4, _col6, _col7, _col8, _col9, _col12, _col13, _col16, _col20, _col21 Statistics: Num rows: 766650239 Data size: 67634106676 Basic stats: COMPLETE Column stats: NONE Group By Operator aggregations: sum(_col4) - keys: _col13 (type: string), _col14 (type: string), _col21 (type: string), _col6 (type: decimal(7,2)), _col7 (type: string), _col9 (type: string), _col10 (type: int), _col17 (type: string), _col23 (type: string) + keys: _col12 (type: string), _col13 (type: string), _col20 (type: string), _col6 (type: decimal(7,2)), _col7 (type: string), _col8 (type: string), _col9 (type: int), _col16 (type: string), _col21 (type: string) mode: hash outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6, _col7, _col8, _col9 Statistics: Num rows: 766650239 Data size: 67634106676 Basic stats: COMPLETE Column stats: NONE @@ -639,21 +639,21 @@ STAGE PLANS: keys: 0 _col0 (type: int) 1 _col1 (type: int) - outputColumnNames: _col1, _col3, _col4, _col5, _col7, _col9, _col11, _col12, _col13 + outputColumnNames: _col1, _col3, _col4, _col5, _col6, _col8, _col10, _col11, _col12 Statistics: Num rows: 88000001 Data size: 75681779077 Basic stats: COMPLETE Column stats: NONE Filter Operator - predicate: (_col13 <> upper(_col3)) (type: boolean) + predicate: (_col12 <> _col3) (type: boolean) Statistics: Num rows: 88000001 Data size: 75681779077 Basic stats: COMPLETE Column stats: NONE Select Operator - expressions: _col9 (type: int), _col11 (type: string), _col12 (type: string), _col1 (type: string), _col4 (type: int), _col5 (type: string), _col7 (type: string) - outputColumnNames: _col0, _col2, _col3, _col6, _col9, _col10, _col12 + expressions: _col8 (type: int), _col10 (type: string), _col11 (type: string), _col1 (type: string), _col4 (type: int), _col5 (type: string), _col6 (type: string) + outputColumnNames: _col0, _col2, _col3, _col6, _col9, _col10, _col11 Statistics: Num rows: 88000001 Data size: 75681779077 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator key expressions: _col0 (type: int), _col9 (type: int) sort order: ++ Map-reduce partition columns: _col0 (type: int), _col9 (type: int) Statistics: Num rows: 88000001 Data size: 75681779077 Basic stats: COMPLETE Column stats: NONE - value expressions: _col2 (type: string), _col3 (type: string), _col6 (type: string), _col10 (type: string), _col12 (type: string) + value expressions: _col2 (type: string), _col3 (type: string), _col6 (type: string), _col10 (type: string), _col11 (type: string) Stage: Stage-0 Fetch Operator http://git-wip-us.apache.org/repos/asf/hive/blob/55887646/ql/src/test/results/clientpositive/perf/spark/query25.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/spark/query25.q.out b/ql/src/test/results/clientpositive/perf/spark/query25.q.out index 5bee2a3..72fe357 100644 --- a/ql/src/test/results/clientpositive/perf/spark/query25.q.out +++ b/ql/src/test/results/clientpositive/perf/spark/query25.q.out @@ -295,14 +295,14 @@ STAGE PLANS: keys: 0 _col2 (type: int), _col1 (type: int) 1 _col1 (type: int), _col2 (type: int) - outputColumnNames: _col3, _col8, _col9, _col10, _col11 + outputColumnNames: _col3, _col6, _col7, _col8, _col9 Statistics: Num rows: 348467716 Data size: 47189528877 Basic stats: COMPLETE Column stats: NONE Reduce Output Operator - key expressions: _col8 (type: int), _col9 (type: int), _col10 (type: int) + key expressions: _col6 (type: int), _col7 (type: int), _col8 (type: int) sort order: +++ - Map-reduce partition columns: _col8 (type: int), _col9 (type: int), _col10 (type: int) + Map-reduce partition columns: _col6 (type: int), _col7 (type: int), _col8 (type: int) Statistics: Num rows: 348467716 Data size: 47189528877 Basic stats: COMPLETE Column stats: NONE - value expressions: _col3 (type: decimal(7,2)), _col11 (type: decimal(7,2)) + value expressions: _col3 (type: decimal(7,2)), _col9 (type: decimal(7,2)) Reducer 13 Reduce Operator Tree: Join Operator @@ -342,15 +342,15 @@ STAGE PLANS: Inner Join 0 to 1 keys: 0 _col1 (type: int), _col2 (type: int), _col4 (type: int) - 1 _col8 (type: int), _col9 (type: int), _col10 (type: int) - outputColumnNames: _col1, _col3, _col5, _col12, _col20 + 1 _col6 (type: int), _col7 (type: int), _col8 (type: int) + outputColumnNames: _col1, _col3, _col5, _col10, _col16 Statistics: Num rows: 696954748 Data size: 61485550191 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: 696954748 Data size: 61485550191 Basic stats: COMPLETE Column stats: NONE - value expressions: _col3 (type: int), _col5 (type: decimal(7,2)), _col12 (type: decimal(7,2)), _col20 (type: decimal(7,2)) + value expressions: _col3 (type: int), _col5 (type: decimal(7,2)), _col10 (type: decimal(7,2)), _col16 (type: decimal(7,2)) Reducer 4 Local Work: Map Reduce Local Work @@ -361,7 +361,7 @@ STAGE PLANS: keys: 0 _col1 (type: int) 1 _col0 (type: int) - outputColumnNames: _col3, _col5, _col12, _col20, _col25, _col26 + outputColumnNames: _col3, _col5, _col10, _col16, _col19, _col20 Statistics: Num rows: 766650239 Data size: 67634106676 Basic stats: COMPLETE Column stats: NONE Map Join Operator condition map: @@ -369,13 +369,13 @@ STAGE PLANS: keys: 0 _col3 (type: int) 1 _col0 (type: int) - outputColumnNames: _col5, _col12, _col20, _col25, _col26, _col28, _col29 + outputColumnNames: _col5, _col10, _col16, _col19, _col20, _col22, _col23 input vertices: 1 Map 16 Statistics: Num rows: 843315281 Data size: 74397518956 Basic stats: COMPLETE Column stats: NONE Group By Operator - aggregations: sum(_col5), sum(_col20), sum(_col12) - keys: _col25 (type: string), _col26 (type: string), _col28 (type: string), _col29 (type: string) + aggregations: sum(_col5), sum(_col16), sum(_col10) + keys: _col19 (type: string), _col20 (type: string), _col22 (type: string), _col23 (type: string) mode: hash outputColumnNames: _col0, _col1, _col2, _col3, _col4, _col5, _col6 Statistics: Num rows: 843315281 Data size: 74397518956 Basic stats: COMPLETE Column stats: NONE