http://git-wip-us.apache.org/repos/asf/hive/blob/5553c59e/ql/src/test/results/clientpositive/perf/tez/constraints/query67.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/tez/constraints/query67.q.out b/ql/src/test/results/clientpositive/perf/tez/constraints/query67.q.out index 7abc959..f629767 100644 --- a/ql/src/test/results/clientpositive/perf/tez/constraints/query67.q.out +++ b/ql/src/test/results/clientpositive/perf/tez/constraints/query67.q.out @@ -123,12 +123,12 @@ Stage-0 Filter Operator [FIL_102] (rows=273593580 width=613) predicate:(rank_window_0 <= 100) PTF Operator [PTF_101] (rows=820780740 width=613) - Function definitions:[{},{"name:":"windowingtablefunction","order by:":"_col8 DESC NULLS LAST","partition by:":"_col2"}] + Function definitions:[{},{"name:":"windowingtablefunction","order by:":"_col8 DESC NULLS LAST","partition by:":"_col6"}] Select Operator [SEL_100] (rows=820780740 width=613) Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7","_col8"] <-Reducer 5 [SIMPLE_EDGE] vectorized SHUFFLE [RS_99] - PartitionCols:_col2 + PartitionCols:_col6 Select Operator [SEL_98] (rows=820780740 width=613) Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7","_col8"] Group By Operator [GBY_97] (rows=820780740 width=621) @@ -137,7 +137,7 @@ Stage-0 SHUFFLE [RS_21] PartitionCols:_col0, _col1, _col2, _col3, _col4, _col5, _col6, _col7, _col8 Group By Operator [GBY_20] (rows=820780740 width=621) - Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7","_col8","_col9"],aggregations:["sum(_col3)"],keys:_col11, _col12, _col13, _col14, _col5, _col6, _col7, _col9, 0L + Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7","_col8","_col9"],aggregations:["sum(_col3)"],keys:_col5, _col6, _col7, _col9, _col11, _col12, _col13, _col14, 0L Merge Join Operator [MERGEJOIN_81] (rows=91197860 width=613) Conds:RS_16._col1=RS_96._col0(Inner),Output:["_col3","_col5","_col6","_col7","_col9","_col11","_col12","_col13","_col14"] <-Map 11 [SIMPLE_EDGE] vectorized
http://git-wip-us.apache.org/repos/asf/hive/blob/5553c59e/ql/src/test/results/clientpositive/perf/tez/constraints/query68.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/tez/constraints/query68.q.out b/ql/src/test/results/clientpositive/perf/tez/constraints/query68.q.out index 2ce7059..582c7da 100644 --- a/ql/src/test/results/clientpositive/perf/tez/constraints/query68.q.out +++ b/ql/src/test/results/clientpositive/perf/tez/constraints/query68.q.out @@ -97,158 +97,158 @@ POSTHOOK: Output: hdfs://### HDFS PATH ### Plan optimized by CBO. Vertex dependency in root stage -Map 8 <- Reducer 13 (BROADCAST_EDGE), Reducer 15 (BROADCAST_EDGE), Reducer 17 (BROADCAST_EDGE) -Reducer 10 <- Map 14 (SIMPLE_EDGE), Reducer 9 (SIMPLE_EDGE) -Reducer 11 <- Map 16 (SIMPLE_EDGE), Reducer 10 (SIMPLE_EDGE) -Reducer 13 <- Map 12 (CUSTOM_SIMPLE_EDGE) -Reducer 15 <- Map 14 (CUSTOM_SIMPLE_EDGE) -Reducer 17 <- Map 16 (CUSTOM_SIMPLE_EDGE) -Reducer 2 <- Map 1 (SIMPLE_EDGE), Map 5 (SIMPLE_EDGE) -Reducer 3 <- Reducer 2 (SIMPLE_EDGE), Reducer 7 (SIMPLE_EDGE) +Map 5 <- Reducer 12 (BROADCAST_EDGE), Reducer 14 (BROADCAST_EDGE), Reducer 16 (BROADCAST_EDGE) +Reducer 10 <- Reducer 9 (SIMPLE_EDGE) +Reducer 12 <- Map 11 (CUSTOM_SIMPLE_EDGE) +Reducer 14 <- Map 13 (CUSTOM_SIMPLE_EDGE) +Reducer 16 <- Map 15 (CUSTOM_SIMPLE_EDGE) +Reducer 2 <- Map 1 (SIMPLE_EDGE), Reducer 10 (SIMPLE_EDGE) +Reducer 3 <- Map 17 (SIMPLE_EDGE), Reducer 2 (SIMPLE_EDGE) Reducer 4 <- Reducer 3 (SIMPLE_EDGE) -Reducer 6 <- Map 5 (SIMPLE_EDGE), Reducer 11 (SIMPLE_EDGE) -Reducer 7 <- Reducer 6 (SIMPLE_EDGE) -Reducer 9 <- Map 12 (SIMPLE_EDGE), Map 8 (SIMPLE_EDGE) +Reducer 6 <- Map 11 (SIMPLE_EDGE), Map 5 (SIMPLE_EDGE) +Reducer 7 <- Map 13 (SIMPLE_EDGE), Reducer 6 (SIMPLE_EDGE) +Reducer 8 <- Map 15 (SIMPLE_EDGE), Reducer 7 (SIMPLE_EDGE) +Reducer 9 <- Map 17 (SIMPLE_EDGE), Reducer 8 (SIMPLE_EDGE) Stage-0 Fetch Operator limit:100 Stage-1 Reducer 4 vectorized - File Output Operator [FS_182] - Limit [LIM_181] (rows=100 width=706) + File Output Operator [FS_186] + Limit [LIM_185] (rows=100 width=706) Number of rows:100 - Select Operator [SEL_180] (rows=727776 width=706) + Select Operator [SEL_184] (rows=727776 width=706) Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"] <-Reducer 3 [SIMPLE_EDGE] SHUFFLE [RS_44] Select Operator [SEL_43] (rows=727776 width=706) Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"] Filter Operator [FIL_42] (rows=727776 width=706) - predicate:(_col5 <> _col8) - Merge Join Operator [MERGEJOIN_143] (rows=727776 width=706) - Conds:RS_39._col0=RS_179._col1(Inner),Output:["_col2","_col3","_col5","_col6","_col8","_col9","_col10","_col11"] + predicate:(_col11 <> _col6) + Merge Join Operator [MERGEJOIN_147] (rows=727776 width=706) + Conds:RS_39._col1=RS_179._col0(Inner),Output:["_col2","_col3","_col4","_col6","_col7","_col8","_col9","_col11"] + <-Map 17 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_179] + PartitionCols:_col0 + Select Operator [SEL_178] (rows=40000000 width=97) + Output:["_col0","_col1"] + TableScan [TS_34] (rows=40000000 width=97) + default@customer_address,current_addr,Tbl:COMPLETE,Col:COMPLETE,Output:["ca_address_sk","ca_city"] <-Reducer 2 [SIMPLE_EDGE] SHUFFLE [RS_39] - PartitionCols:_col0 - Merge Join Operator [MERGEJOIN_138] (rows=80000000 width=277) - Conds:RS_146._col1=RS_148._col0(Inner),Output:["_col0","_col2","_col3","_col5"] - <-Map 5 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_148] - PartitionCols:_col0 - Select Operator [SEL_147] (rows=40000000 width=97) - Output:["_col0","_col1"] - TableScan [TS_3] (rows=40000000 width=97) - default@customer_address,current_addr,Tbl:COMPLETE,Col:COMPLETE,Output:["ca_address_sk","ca_city"] + PartitionCols:_col1 + Merge Join Operator [MERGEJOIN_146] (rows=727776 width=617) + Conds:RS_150._col0=RS_183._col1(Inner),Output:["_col1","_col2","_col3","_col4","_col6","_col7","_col8","_col9"] <-Map 1 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_146] - PartitionCols:_col1 - Select Operator [SEL_145] (rows=80000000 width=188) + SHUFFLE [RS_150] + PartitionCols:_col0 + Select Operator [SEL_149] (rows=80000000 width=188) Output:["_col0","_col1","_col2","_col3"] - Filter Operator [FIL_144] (rows=80000000 width=188) + Filter Operator [FIL_148] (rows=80000000 width=188) predicate:c_current_addr_sk is not null TableScan [TS_0] (rows=80000000 width=188) default@customer,customer,Tbl:COMPLETE,Col:COMPLETE,Output:["c_customer_sk","c_current_addr_sk","c_first_name","c_last_name"] - <-Reducer 7 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_179] - PartitionCols:_col1 - Select Operator [SEL_178] (rows=727776 width=433) - Output:["_col0","_col1","_col2","_col3","_col4","_col5"] - Group By Operator [GBY_177] (rows=727776 width=433) - Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6"],aggregations:["sum(VALUE._col0)","sum(VALUE._col1)","sum(VALUE._col2)"],keys:KEY._col0, KEY._col1, KEY._col2, KEY._col3 - <-Reducer 6 [SIMPLE_EDGE] - SHUFFLE [RS_33] - PartitionCols:_col0, _col1, _col2, _col3 - Group By Operator [GBY_32] (rows=727776 width=433) - Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6"],aggregations:["sum(_col6)","sum(_col7)","sum(_col8)"],keys:_col1, _col13, _col3, _col5 - Merge Join Operator [MERGEJOIN_142] (rows=727776 width=97) - Conds:RS_28._col3=RS_149._col0(Inner),Output:["_col1","_col3","_col5","_col6","_col7","_col8","_col13"] - <-Map 5 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_149] - PartitionCols:_col0 - Please refer to the previous Select Operator [SEL_147] - <-Reducer 11 [SIMPLE_EDGE] - SHUFFLE [RS_28] - PartitionCols:_col3 - Merge Join Operator [MERGEJOIN_141] (rows=727776 width=4) - Conds:RS_25._col2=RS_168._col0(Inner),Output:["_col1","_col3","_col5","_col6","_col7","_col8"] - <-Map 16 [SIMPLE_EDGE] vectorized - PARTITION_ONLY_SHUFFLE [RS_168] + <-Reducer 10 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_183] + PartitionCols:_col1 + Select Operator [SEL_182] (rows=727776 width=433) + Output:["_col0","_col1","_col2","_col3","_col4","_col5"] + Group By Operator [GBY_181] (rows=727776 width=433) + Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6"],aggregations:["sum(VALUE._col0)","sum(VALUE._col1)","sum(VALUE._col2)"],keys:KEY._col0, KEY._col1, KEY._col2, KEY._col3 + <-Reducer 9 [SIMPLE_EDGE] + SHUFFLE [RS_31] + PartitionCols:_col0, _col1, _col2, _col3 + Group By Operator [GBY_30] (rows=727776 width=433) + Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6"],aggregations:["sum(_col6)","sum(_col7)","sum(_col8)"],keys:_col1, _col13, _col3, _col5 + Merge Join Operator [MERGEJOIN_145] (rows=727776 width=97) + Conds:RS_26._col3=RS_180._col0(Inner),Output:["_col1","_col3","_col5","_col6","_col7","_col8","_col13"] + <-Map 17 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_180] PartitionCols:_col0 - Select Operator [SEL_167] (rows=1855 width=4) - Output:["_col0"] - Filter Operator [FIL_166] (rows=1855 width=12) - predicate:((hd_dep_count = 2) or (hd_vehicle_count = 1)) - TableScan [TS_14] (rows=7200 width=12) - default@household_demographics,household_demographics,Tbl:COMPLETE,Col:COMPLETE,Output:["hd_demo_sk","hd_dep_count","hd_vehicle_count"] - <-Reducer 10 [SIMPLE_EDGE] - SHUFFLE [RS_25] - PartitionCols:_col2 - Merge Join Operator [MERGEJOIN_140] (rows=2824787 width=4) - Conds:RS_22._col4=RS_160._col0(Inner),Output:["_col1","_col2","_col3","_col5","_col6","_col7","_col8"] - <-Map 14 [SIMPLE_EDGE] vectorized - PARTITION_ONLY_SHUFFLE [RS_160] + Please refer to the previous Select Operator [SEL_178] + <-Reducer 8 [SIMPLE_EDGE] + SHUFFLE [RS_26] + PartitionCols:_col3 + Merge Join Operator [MERGEJOIN_144] (rows=727776 width=4) + Conds:RS_23._col2=RS_169._col0(Inner),Output:["_col1","_col3","_col5","_col6","_col7","_col8"] + <-Map 15 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_169] PartitionCols:_col0 - Select Operator [SEL_159] (rows=14 width=4) + Select Operator [SEL_168] (rows=1855 width=4) Output:["_col0"] - Filter Operator [FIL_158] (rows=14 width=97) - predicate:(s_city) IN ('Cedar Grove', 'Wildwood') - TableScan [TS_11] (rows=1704 width=97) - default@store,store,Tbl:COMPLETE,Col:COMPLETE,Output:["s_store_sk","s_city"] - <-Reducer 9 [SIMPLE_EDGE] - SHUFFLE [RS_22] - PartitionCols:_col4 - Merge Join Operator [MERGEJOIN_139] (rows=42598570 width=185) - Conds:RS_176._col0=RS_152._col0(Inner),Output:["_col1","_col2","_col3","_col4","_col5","_col6","_col7","_col8"] - <-Map 12 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_152] + Filter Operator [FIL_167] (rows=1855 width=12) + predicate:((hd_dep_count = 2) or (hd_vehicle_count = 1)) + TableScan [TS_12] (rows=7200 width=12) + default@household_demographics,household_demographics,Tbl:COMPLETE,Col:COMPLETE,Output:["hd_demo_sk","hd_dep_count","hd_vehicle_count"] + <-Reducer 7 [SIMPLE_EDGE] + SHUFFLE [RS_23] + PartitionCols:_col2 + Merge Join Operator [MERGEJOIN_143] (rows=2824787 width=4) + Conds:RS_20._col4=RS_161._col0(Inner),Output:["_col1","_col2","_col3","_col5","_col6","_col7","_col8"] + <-Map 13 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_161] PartitionCols:_col0 - Select Operator [SEL_151] (rows=170 width=4) + Select Operator [SEL_160] (rows=14 width=4) Output:["_col0"] - Filter Operator [FIL_150] (rows=170 width=12) - predicate:((d_year) IN (1998, 1999, 2000) and d_dom BETWEEN 1 AND 2) - TableScan [TS_8] (rows=73049 width=12) - default@date_dim,date_dim,Tbl:COMPLETE,Col:COMPLETE,Output:["d_date_sk","d_year","d_dom"] - <-Map 8 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_176] - PartitionCols:_col0 - Select Operator [SEL_175] (rows=457565061 width=343) - Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7","_col8"] - Filter Operator [FIL_174] (rows=457565061 width=343) - predicate:((ss_hdemo_sk BETWEEN DynamicValue(RS_26_household_demographics_hd_demo_sk_min) AND DynamicValue(RS_26_household_demographics_hd_demo_sk_max) and in_bloom_filter(ss_hdemo_sk, DynamicValue(RS_26_household_demographics_hd_demo_sk_bloom_filter))) and (ss_sold_date_sk BETWEEN DynamicValue(RS_20_date_dim_d_date_sk_min) AND DynamicValue(RS_20_date_dim_d_date_sk_max) and in_bloom_filter(ss_sold_date_sk, DynamicValue(RS_20_date_dim_d_date_sk_bloom_filter))) and (ss_store_sk BETWEEN DynamicValue(RS_23_store_s_store_sk_min) AND DynamicValue(RS_23_store_s_store_sk_max) and in_bloom_filter(ss_store_sk, DynamicValue(RS_23_store_s_store_sk_bloom_filter))) and ss_addr_sk is not null and ss_customer_sk is not null and ss_hdemo_sk is not null and ss_sold_date_sk is not null and ss_store_sk is not null) - TableScan [TS_5] (rows=575995635 width=343) - default@store_sales,store_sales,Tbl:COMPLETE,Col:COMPLETE,Output:["ss_sold_date_sk","ss_customer_sk","ss_hdemo_sk","ss_addr_sk","ss_store_sk","ss_ticket_number","ss_ext_sales_price","ss_ext_list_price","ss_ext_tax"] - <-Reducer 13 [BROADCAST_EDGE] vectorized - BROADCAST [RS_157] - Group By Operator [GBY_156] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] - <-Map 12 [CUSTOM_SIMPLE_EDGE] vectorized - SHUFFLE [RS_155] - Group By Operator [GBY_154] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] - Select Operator [SEL_153] (rows=170 width=4) - Output:["_col0"] - Please refer to the previous Select Operator [SEL_151] - <-Reducer 15 [BROADCAST_EDGE] vectorized - BROADCAST [RS_165] - Group By Operator [GBY_164] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] - <-Map 14 [CUSTOM_SIMPLE_EDGE] vectorized - PARTITION_ONLY_SHUFFLE [RS_163] - Group By Operator [GBY_162] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] - Select Operator [SEL_161] (rows=14 width=4) - Output:["_col0"] - Please refer to the previous Select Operator [SEL_159] - <-Reducer 17 [BROADCAST_EDGE] vectorized - BROADCAST [RS_173] - Group By Operator [GBY_172] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] - <-Map 16 [CUSTOM_SIMPLE_EDGE] vectorized - PARTITION_ONLY_SHUFFLE [RS_171] - Group By Operator [GBY_170] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] - Select Operator [SEL_169] (rows=1855 width=4) - Output:["_col0"] - Please refer to the previous Select Operator [SEL_167] + Filter Operator [FIL_159] (rows=14 width=97) + predicate:(s_city) IN ('Cedar Grove', 'Wildwood') + TableScan [TS_9] (rows=1704 width=97) + default@store,store,Tbl:COMPLETE,Col:COMPLETE,Output:["s_store_sk","s_city"] + <-Reducer 6 [SIMPLE_EDGE] + SHUFFLE [RS_20] + PartitionCols:_col4 + Merge Join Operator [MERGEJOIN_142] (rows=42598570 width=185) + Conds:RS_177._col0=RS_153._col0(Inner),Output:["_col1","_col2","_col3","_col4","_col5","_col6","_col7","_col8"] + <-Map 11 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_153] + PartitionCols:_col0 + Select Operator [SEL_152] (rows=170 width=4) + Output:["_col0"] + Filter Operator [FIL_151] (rows=170 width=12) + predicate:((d_year) IN (1998, 1999, 2000) and d_dom BETWEEN 1 AND 2) + TableScan [TS_6] (rows=73049 width=12) + default@date_dim,date_dim,Tbl:COMPLETE,Col:COMPLETE,Output:["d_date_sk","d_year","d_dom"] + <-Map 5 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_177] + PartitionCols:_col0 + Select Operator [SEL_176] (rows=457565061 width=343) + Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7","_col8"] + Filter Operator [FIL_175] (rows=457565061 width=343) + predicate:((ss_hdemo_sk BETWEEN DynamicValue(RS_24_household_demographics_hd_demo_sk_min) AND DynamicValue(RS_24_household_demographics_hd_demo_sk_max) and in_bloom_filter(ss_hdemo_sk, DynamicValue(RS_24_household_demographics_hd_demo_sk_bloom_filter))) and (ss_sold_date_sk BETWEEN DynamicValue(RS_18_date_dim_d_date_sk_min) AND DynamicValue(RS_18_date_dim_d_date_sk_max) and in_bloom_filter(ss_sold_date_sk, DynamicValue(RS_18_date_dim_d_date_sk_bloom_filter))) and (ss_store_sk BETWEEN DynamicValue(RS_21_store_s_store_sk_min) AND DynamicValue(RS_21_store_s_store_sk_max) and in_bloom_filter(ss_store_sk, DynamicValue(RS_21_store_s_store_sk_bloom_filter))) and ss_addr_sk is not null and ss_customer_sk is not null and ss_hdemo_sk is not null and ss_sold_date_sk is not null and ss_store_sk is not null) + TableScan [TS_3] (rows=575995635 width=343) + default@store_sales,store_sales,Tbl:COMPLETE,Col:COMPLETE,Output:["ss_sold_date_sk","ss_customer_sk","ss_hdemo_sk","ss_addr_sk","ss_store_sk","ss_ticket_number","ss_ext_sales_price","ss_ext_list_price","ss_ext_tax"] + <-Reducer 12 [BROADCAST_EDGE] vectorized + BROADCAST [RS_158] + Group By Operator [GBY_157] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] + <-Map 11 [CUSTOM_SIMPLE_EDGE] vectorized + SHUFFLE [RS_156] + Group By Operator [GBY_155] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] + Select Operator [SEL_154] (rows=170 width=4) + Output:["_col0"] + Please refer to the previous Select Operator [SEL_152] + <-Reducer 14 [BROADCAST_EDGE] vectorized + BROADCAST [RS_166] + Group By Operator [GBY_165] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] + <-Map 13 [CUSTOM_SIMPLE_EDGE] vectorized + SHUFFLE [RS_164] + Group By Operator [GBY_163] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] + Select Operator [SEL_162] (rows=14 width=4) + Output:["_col0"] + Please refer to the previous Select Operator [SEL_160] + <-Reducer 16 [BROADCAST_EDGE] vectorized + BROADCAST [RS_174] + Group By Operator [GBY_173] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] + <-Map 15 [CUSTOM_SIMPLE_EDGE] vectorized + SHUFFLE [RS_172] + Group By Operator [GBY_171] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] + Select Operator [SEL_170] (rows=1855 width=4) + Output:["_col0"] + Please refer to the previous Select Operator [SEL_168] http://git-wip-us.apache.org/repos/asf/hive/blob/5553c59e/ql/src/test/results/clientpositive/perf/tez/constraints/query70.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/tez/constraints/query70.q.out b/ql/src/test/results/clientpositive/perf/tez/constraints/query70.q.out index 2df373a..2d83c98 100644 --- a/ql/src/test/results/clientpositive/perf/tez/constraints/query70.q.out +++ b/ql/src/test/results/clientpositive/perf/tez/constraints/query70.q.out @@ -83,134 +83,174 @@ POSTHOOK: Output: hdfs://### HDFS PATH ### Plan optimized by CBO. Vertex dependency in root stage -Map 1 <- Reducer 12 (BROADCAST_EDGE) -Reducer 10 <- Reducer 9 (SIMPLE_EDGE) -Reducer 12 <- Map 11 (CUSTOM_SIMPLE_EDGE) -Reducer 2 <- Map 1 (SIMPLE_EDGE), Map 11 (SIMPLE_EDGE) -Reducer 3 <- Map 13 (SIMPLE_EDGE), Reducer 2 (SIMPLE_EDGE) -Reducer 4 <- Reducer 10 (SIMPLE_EDGE), Reducer 3 (SIMPLE_EDGE) +Map 1 <- Reducer 14 (BROADCAST_EDGE), Reducer 8 (BROADCAST_EDGE) +Map 17 <- Reducer 15 (BROADCAST_EDGE) +Reducer 10 <- Map 18 (SIMPLE_EDGE), Reducer 9 (SIMPLE_EDGE) +Reducer 11 <- Reducer 10 (SIMPLE_EDGE) +Reducer 12 <- Reducer 11 (SIMPLE_EDGE) +Reducer 13 <- Map 16 (SIMPLE_EDGE), Reducer 12 (SIMPLE_EDGE) +Reducer 14 <- Reducer 13 (CUSTOM_SIMPLE_EDGE) +Reducer 15 <- Map 7 (CUSTOM_SIMPLE_EDGE) +Reducer 2 <- Map 1 (SIMPLE_EDGE), Map 7 (SIMPLE_EDGE) +Reducer 3 <- Reducer 13 (SIMPLE_EDGE), Reducer 2 (SIMPLE_EDGE) +Reducer 4 <- Reducer 3 (SIMPLE_EDGE) Reducer 5 <- Reducer 4 (SIMPLE_EDGE) Reducer 6 <- Reducer 5 (SIMPLE_EDGE) -Reducer 7 <- Reducer 6 (SIMPLE_EDGE) -Reducer 8 <- Map 14 (SIMPLE_EDGE), Reducer 2 (SIMPLE_EDGE) -Reducer 9 <- Reducer 8 (SIMPLE_EDGE) +Reducer 8 <- Map 7 (CUSTOM_SIMPLE_EDGE) +Reducer 9 <- Map 17 (SIMPLE_EDGE), Map 7 (SIMPLE_EDGE) Stage-0 Fetch Operator limit:-1 Stage-1 - Reducer 7 vectorized - File Output Operator [FS_168] - Limit [LIM_167] (rows=100 width=492) + Reducer 6 vectorized + File Output Operator [FS_181] + Limit [LIM_180] (rows=1 width=492) Number of rows:100 - Select Operator [SEL_166] (rows=720 width=492) + Select Operator [SEL_179] (rows=1 width=492) Output:["_col0","_col1","_col2","_col3","_col4"] - <-Reducer 6 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_165] - Select Operator [SEL_164] (rows=720 width=492) + <-Reducer 5 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_178] + Select Operator [SEL_177] (rows=1 width=492) Output:["_col0","_col1","_col2","_col3","_col4","_col5"] - PTF Operator [PTF_163] (rows=720 width=304) + PTF Operator [PTF_176] (rows=1 width=304) Function definitions:[{},{"name:":"windowingtablefunction","order by:":"_col2 DESC NULLS LAST","partition by:":"(grouping(_col3, 1) + grouping(_col3, 0)), CASE WHEN ((grouping(_col3, 0) = 0)) THEN (_col0) ELSE (CAST( null AS STRING)) END"}] - Select Operator [SEL_162] (rows=720 width=304) + Select Operator [SEL_175] (rows=1 width=304) Output:["_col0","_col1","_col2","_col3"] - <-Reducer 5 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_161] + <-Reducer 4 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_174] PartitionCols:(grouping(_col3, 1) + grouping(_col3, 0)), CASE WHEN ((grouping(_col3, 0) = 0)) THEN (_col0) ELSE (CAST( null AS STRING)) END - Select Operator [SEL_160] (rows=720 width=304) + Select Operator [SEL_173] (rows=1 width=304) Output:["_col0","_col1","_col2","_col3"] - Group By Operator [GBY_159] (rows=720 width=304) + Group By Operator [GBY_172] (rows=1 width=304) Output:["_col0","_col1","_col2","_col3"],aggregations:["sum(VALUE._col0)"],keys:KEY._col0, KEY._col1, KEY._col2 - <-Reducer 4 [SIMPLE_EDGE] - SHUFFLE [RS_49] + <-Reducer 3 [SIMPLE_EDGE] + SHUFFLE [RS_50] PartitionCols:_col0, _col1, _col2 - Group By Operator [GBY_48] (rows=18000 width=304) + Group By Operator [GBY_49] (rows=1 width=304) Output:["_col0","_col1","_col2","_col3"],aggregations:["sum(_col2)"],keys:_col0, _col1, 0L - Select Operator [SEL_46] (rows=29778893 width=207) + Select Operator [SEL_47] (rows=1 width=296) Output:["_col0","_col1","_col2"] - Merge Join Operator [MERGEJOIN_134] (rows=29778893 width=207) - Conds:RS_43._col7=RS_158._col0(Inner),Output:["_col2","_col6","_col7"] - <-Reducer 10 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_158] + Merge Join Operator [MERGEJOIN_136] (rows=1 width=296) + Conds:RS_44._col1=RS_45._col0(Inner),Output:["_col2","_col6","_col7"] + <-Reducer 13 [SIMPLE_EDGE] + SHUFFLE [RS_45] PartitionCols:_col0 - Select Operator [SEL_157] (rows=16 width=86) - Output:["_col0"] - Filter Operator [FIL_156] (rows=16 width=198) - predicate:(rank_window_0 <= 5) - PTF Operator [PTF_155] (rows=49 width=198) - Function definitions:[{},{"name:":"windowingtablefunction","order by:":"_col1 DESC NULLS LAST","partition by:":"_col0"}] - Select Operator [SEL_154] (rows=49 width=198) - Output:["_col0","_col1"] - <-Reducer 9 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_153] - PartitionCols:_col0 - Group By Operator [GBY_152] (rows=49 width=198) - Output:["_col0","_col1"],aggregations:["sum(VALUE._col0)"],keys:KEY._col0 - <-Reducer 8 [SIMPLE_EDGE] - SHUFFLE [RS_26] - PartitionCols:_col0 - Group By Operator [GBY_25] (rows=2989 width=198) - Output:["_col0","_col1"],aggregations:["sum(_col2)"],keys:_col5 - Merge Join Operator [MERGEJOIN_133] (rows=91197860 width=168) - Conds:RS_21._col1=RS_151._col0(Inner),Output:["_col2","_col5"] - <-Reducer 2 [SIMPLE_EDGE] - SHUFFLE [RS_21] - PartitionCols:_col1 - Merge Join Operator [MERGEJOIN_130] (rows=91197860 width=85) - Conds:RS_145._col0=RS_137._col0(Inner),Output:["_col1","_col2"] - <-Map 11 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_137] - PartitionCols:_col0 - Select Operator [SEL_136] (rows=317 width=8) - Output:["_col0"] - Filter Operator [FIL_135] (rows=317 width=8) - predicate:d_month_seq BETWEEN 1212 AND 1223 - TableScan [TS_3] (rows=73049 width=8) - default@date_dim,d1,Tbl:COMPLETE,Col:COMPLETE,Output:["d_date_sk","d_month_seq"] - <-Map 1 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_145] - PartitionCols:_col0 - Select Operator [SEL_144] (rows=525329897 width=114) - Output:["_col0","_col1","_col2"] - Filter Operator [FIL_143] (rows=525329897 width=114) - predicate:((ss_sold_date_sk BETWEEN DynamicValue(RS_38_d1_d_date_sk_min) AND DynamicValue(RS_38_d1_d_date_sk_max) and in_bloom_filter(ss_sold_date_sk, DynamicValue(RS_38_d1_d_date_sk_bloom_filter))) and ss_sold_date_sk is not null and ss_store_sk is not null) - TableScan [TS_0] (rows=575995635 width=114) - default@store_sales,store_sales,Tbl:COMPLETE,Col:COMPLETE,Output:["ss_sold_date_sk","ss_store_sk","ss_net_profit"] - <-Reducer 12 [BROADCAST_EDGE] vectorized - BROADCAST [RS_142] - Group By Operator [GBY_141] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] - <-Map 11 [CUSTOM_SIMPLE_EDGE] vectorized - SHUFFLE [RS_140] - Group By Operator [GBY_139] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] - Select Operator [SEL_138] (rows=317 width=4) - Output:["_col0"] - Please refer to the previous Select Operator [SEL_136] - <-Map 14 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_151] - PartitionCols:_col0 - Select Operator [SEL_150] (rows=1704 width=90) - Output:["_col0","_col1"] - Filter Operator [FIL_149] (rows=1704 width=90) - predicate:s_state is not null - TableScan [TS_15] (rows=1704 width=90) - default@store,store,Tbl:COMPLETE,Col:COMPLETE,Output:["s_store_sk","s_state"] - <-Reducer 3 [SIMPLE_EDGE] - SHUFFLE [RS_43] - PartitionCols:_col7 - Merge Join Operator [MERGEJOIN_131] (rows=91197860 width=266) - Conds:RS_40._col1=RS_148._col0(Inner),Output:["_col2","_col6","_col7"] - <-Reducer 2 [SIMPLE_EDGE] - SHUFFLE [RS_40] - PartitionCols:_col1 - Please refer to the previous Merge Join Operator [MERGEJOIN_130] - <-Map 13 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_148] - PartitionCols:_col0 - Select Operator [SEL_147] (rows=1704 width=188) + Merge Join Operator [MERGEJOIN_135] (rows=556 width=188) + Conds:RS_166._col2=RS_163._col0(Inner),Output:["_col0","_col1","_col2"] + <-Map 16 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_166] + PartitionCols:_col2 + Select Operator [SEL_165] (rows=1704 width=188) Output:["_col0","_col1","_col2"] - Filter Operator [FIL_146] (rows=1704 width=188) + Filter Operator [FIL_164] (rows=1704 width=188) predicate:s_state is not null TableScan [TS_6] (rows=1704 width=188) default@store,store,Tbl:COMPLETE,Col:COMPLETE,Output:["s_store_sk","s_county","s_state"] + <-Reducer 12 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_163] + PartitionCols:_col0 + Select Operator [SEL_162] (rows=16 width=86) + Output:["_col0"] + Filter Operator [FIL_161] (rows=16 width=198) + predicate:(rank_window_0 <= 5) + PTF Operator [PTF_160] (rows=49 width=198) + Function definitions:[{},{"name:":"windowingtablefunction","order by:":"_col1 DESC NULLS LAST","partition by:":"_col0"}] + Select Operator [SEL_159] (rows=49 width=198) + Output:["_col0","_col1"] + <-Reducer 11 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_158] + PartitionCols:_col0 + Group By Operator [GBY_157] (rows=49 width=198) + Output:["_col0","_col1"],aggregations:["sum(VALUE._col0)"],keys:KEY._col0 + <-Reducer 10 [SIMPLE_EDGE] + SHUFFLE [RS_26] + PartitionCols:_col0 + Group By Operator [GBY_25] (rows=2989 width=198) + Output:["_col0","_col1"],aggregations:["sum(_col2)"],keys:_col5 + Merge Join Operator [MERGEJOIN_134] (rows=91197860 width=168) + Conds:RS_21._col1=RS_156._col0(Inner),Output:["_col2","_col5"] + <-Map 18 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_156] + PartitionCols:_col0 + Select Operator [SEL_155] (rows=1704 width=90) + Output:["_col0","_col1"] + Filter Operator [FIL_154] (rows=1704 width=90) + predicate:s_state is not null + TableScan [TS_15] (rows=1704 width=90) + default@store,store,Tbl:COMPLETE,Col:COMPLETE,Output:["s_store_sk","s_state"] + <-Reducer 9 [SIMPLE_EDGE] + SHUFFLE [RS_21] + PartitionCols:_col1 + Merge Join Operator [MERGEJOIN_133] (rows=91197860 width=85) + Conds:RS_153._col0=RS_141._col0(Inner),Output:["_col1","_col2"] + <-Map 7 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_141] + PartitionCols:_col0 + Select Operator [SEL_138] (rows=317 width=8) + Output:["_col0"] + Filter Operator [FIL_137] (rows=317 width=8) + predicate:d_month_seq BETWEEN 1212 AND 1223 + TableScan [TS_3] (rows=73049 width=8) + default@date_dim,d1,Tbl:COMPLETE,Col:COMPLETE,Output:["d_date_sk","d_month_seq"] + <-Map 17 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_153] + PartitionCols:_col0 + Select Operator [SEL_152] (rows=525329897 width=114) + Output:["_col0","_col1","_col2"] + Filter Operator [FIL_151] (rows=525329897 width=114) + predicate:((ss_sold_date_sk BETWEEN DynamicValue(RS_19_date_dim_d_date_sk_min) AND DynamicValue(RS_19_date_dim_d_date_sk_max) and in_bloom_filter(ss_sold_date_sk, DynamicValue(RS_19_date_dim_d_date_sk_bloom_filter))) and ss_sold_date_sk is not null and ss_store_sk is not null) + TableScan [TS_9] (rows=575995635 width=114) + default@store_sales,store_sales,Tbl:COMPLETE,Col:COMPLETE,Output:["ss_sold_date_sk","ss_store_sk","ss_net_profit"] + <-Reducer 15 [BROADCAST_EDGE] vectorized + BROADCAST [RS_150] + Group By Operator [GBY_149] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] + <-Map 7 [CUSTOM_SIMPLE_EDGE] vectorized + SHUFFLE [RS_146] + Group By Operator [GBY_144] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] + Select Operator [SEL_142] (rows=317 width=4) + Output:["_col0"] + Please refer to the previous Select Operator [SEL_138] + <-Reducer 2 [SIMPLE_EDGE] + SHUFFLE [RS_44] + PartitionCols:_col1 + Merge Join Operator [MERGEJOIN_132] (rows=91197860 width=85) + Conds:RS_171._col0=RS_139._col0(Inner),Output:["_col1","_col2"] + <-Map 7 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_139] + PartitionCols:_col0 + Please refer to the previous Select Operator [SEL_138] + <-Map 1 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_171] + PartitionCols:_col0 + Select Operator [SEL_170] (rows=525329897 width=114) + Output:["_col0","_col1","_col2"] + Filter Operator [FIL_169] (rows=525329897 width=114) + predicate:((ss_sold_date_sk BETWEEN DynamicValue(RS_42_d1_d_date_sk_min) AND DynamicValue(RS_42_d1_d_date_sk_max) and in_bloom_filter(ss_sold_date_sk, DynamicValue(RS_42_d1_d_date_sk_bloom_filter))) and (ss_store_sk BETWEEN DynamicValue(RS_45_store_s_store_sk_min) AND DynamicValue(RS_45_store_s_store_sk_max) and in_bloom_filter(ss_store_sk, DynamicValue(RS_45_store_s_store_sk_bloom_filter))) and ss_sold_date_sk is not null and ss_store_sk is not null) + TableScan [TS_0] (rows=575995635 width=114) + default@store_sales,store_sales,Tbl:COMPLETE,Col:COMPLETE,Output:["ss_sold_date_sk","ss_store_sk","ss_net_profit"] + <-Reducer 14 [BROADCAST_EDGE] vectorized + BROADCAST [RS_168] + Group By Operator [GBY_167] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] + <-Reducer 13 [CUSTOM_SIMPLE_EDGE] + SHUFFLE [RS_94] + Group By Operator [GBY_93] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] + Select Operator [SEL_92] (rows=556 width=4) + Output:["_col0"] + Please refer to the previous Merge Join Operator [MERGEJOIN_135] + <-Reducer 8 [BROADCAST_EDGE] vectorized + BROADCAST [RS_148] + Group By Operator [GBY_147] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] + <-Map 7 [CUSTOM_SIMPLE_EDGE] vectorized + SHUFFLE [RS_145] + Group By Operator [GBY_143] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] + Select Operator [SEL_140] (rows=317 width=4) + Output:["_col0"] + Please refer to the previous Select Operator [SEL_138]