http://git-wip-us.apache.org/repos/asf/hive/blob/55887646/ql/src/test/results/clientpositive/perf/tez/query11.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/tez/query11.q.out b/ql/src/test/results/clientpositive/perf/tez/query11.q.out index 2f453f3..da1c349 100644 --- a/ql/src/test/results/clientpositive/perf/tez/query11.q.out +++ b/ql/src/test/results/clientpositive/perf/tez/query11.q.out @@ -189,249 +189,241 @@ Stage-0 limit:100 Stage-1 Reducer 8 vectorized - File Output Operator [FS_358] - Limit [LIM_357] (rows=100 width=85) + File Output Operator [FS_354] + Limit [LIM_353] (rows=100 width=85) Number of rows:100 - Select Operator [SEL_356] (rows=12248093 width=85) + Select Operator [SEL_352] (rows=12248093 width=85) Output:["_col0"] <-Reducer 7 [SIMPLE_EDGE] - SHUFFLE [RS_97] - Select Operator [SEL_96] (rows=12248093 width=85) + SHUFFLE [RS_93] + Select Operator [SEL_92] (rows=12248093 width=85) Output:["_col0"] - Filter Operator [FIL_95] (rows=12248093 width=533) - predicate:CASE WHEN (_col3 is not null) THEN (CASE WHEN (_col5 is not null) THEN (((_col1 / _col5) > (_col8 / _col3))) ELSE ((null > (_col8 / _col3))) END) ELSE (CASE WHEN (_col5 is not null) THEN (((_col1 / _col5) > null)) ELSE (null) END) END - Merge Join Operator [MERGEJOIN_291] (rows=24496186 width=533) - Conds:RS_92._col2=RS_355._col0(Inner),Output:["_col1","_col3","_col5","_col7","_col8"] + Filter Operator [FIL_91] (rows=12248093 width=537) + predicate:CASE WHEN (_col3 is not null) THEN (CASE WHEN (_col6) THEN (((_col1 / _col5) > (_col9 / _col3))) ELSE ((null > (_col9 / _col3))) END) ELSE (CASE WHEN (_col6) THEN (((_col1 / _col5) > null)) ELSE (null) END) END + Merge Join Operator [MERGEJOIN_287] (rows=24496186 width=537) + Conds:RS_88._col2=RS_351._col0(Inner),Output:["_col1","_col3","_col5","_col6","_col8","_col9"] <-Reducer 20 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_355] + SHUFFLE [RS_351] PartitionCols:_col0 - Select Operator [SEL_354] (rows=80000000 width=297) + Select Operator [SEL_350] (rows=80000000 width=297) Output:["_col0","_col1","_col2"] - Group By Operator [GBY_353] (rows=80000000 width=764) + Group By Operator [GBY_349] (rows=80000000 width=764) Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"],aggregations:["sum(VALUE._col0)"],keys:KEY._col0, KEY._col1, KEY._col2, KEY._col3, KEY._col4, KEY._col5, KEY._col6 <-Reducer 19 [SIMPLE_EDGE] - SHUFFLE [RS_83] + SHUFFLE [RS_79] PartitionCols:_col0, _col1, _col2, _col3, _col4, _col5, _col6 - Group By Operator [GBY_82] (rows=80000000 width=764) - Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"],aggregations:["sum(_col7)"],keys:_col0, _col1, _col2, _col3, _col4, _col5, _col6 - Select Operator [SEL_80] (rows=187573258 width=847) - Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"] - Merge Join Operator [MERGEJOIN_288] (rows=187573258 width=847) - Conds:RS_77._col1=RS_321._col0(Inner),Output:["_col2","_col3","_col7","_col8","_col9","_col10","_col11","_col12","_col13"] - <-Map 26 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_321] - PartitionCols:_col0 - Select Operator [SEL_320] (rows=80000000 width=656) - Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"] - Filter Operator [FIL_319] (rows=80000000 width=656) - predicate:(c_customer_id is not null and c_customer_sk is not null) - TableScan [TS_71] (rows=80000000 width=656) - default@customer,customer,Tbl:COMPLETE,Col:COMPLETE,Output:["c_customer_sk","c_customer_id","c_first_name","c_last_name","c_preferred_cust_flag","c_birth_country","c_login","c_email_address"] - <-Reducer 18 [SIMPLE_EDGE] - SHUFFLE [RS_77] - PartitionCols:_col1 - Merge Join Operator [MERGEJOIN_287] (rows=187573258 width=199) - Conds:RS_352._col0=RS_298._col0(Inner),Output:["_col1","_col2","_col3"] - <-Map 21 [SIMPLE_EDGE] vectorized - PARTITION_ONLY_SHUFFLE [RS_298] - PartitionCols:_col0 - Select Operator [SEL_295] (rows=652 width=8) - Output:["_col0"] - Filter Operator [FIL_292] (rows=652 width=8) - predicate:((d_year = 2002) and d_date_sk is not null) - TableScan [TS_68] (rows=73049 width=8) - default@date_dim,date_dim,Tbl:COMPLETE,Col:COMPLETE,Output:["d_date_sk","d_year"] - <-Map 17 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_352] - PartitionCols:_col0 - Select Operator [SEL_351] (rows=525327388 width=221) - Output:["_col0","_col1","_col2","_col3"] - Filter Operator [FIL_350] (rows=525327388 width=221) - predicate:((ss_sold_date_sk BETWEEN DynamicValue(RS_75_date_dim_d_date_sk_min) AND DynamicValue(RS_75_date_dim_d_date_sk_max) and in_bloom_filter(ss_sold_date_sk, DynamicValue(RS_75_date_dim_d_date_sk_bloom_filter))) and ss_customer_sk is not null and ss_sold_date_sk is not null) - TableScan [TS_65] (rows=575995635 width=221) - default@store_sales,store_sales,Tbl:COMPLETE,Col:COMPLETE,Output:["ss_sold_date_sk","ss_customer_sk","ss_ext_discount_amt","ss_ext_list_price"] - <-Reducer 22 [BROADCAST_EDGE] vectorized - BROADCAST [RS_349] - Group By Operator [GBY_348] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] - <-Map 21 [CUSTOM_SIMPLE_EDGE] vectorized - PARTITION_ONLY_SHUFFLE [RS_310] - Group By Operator [GBY_306] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] - Select Operator [SEL_299] (rows=652 width=4) - Output:["_col0"] - Please refer to the previous Select Operator [SEL_295] + Group By Operator [GBY_78] (rows=80000000 width=764) + Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"],aggregations:["sum(_col2)"],keys:_col5, _col6, _col7, _col8, _col9, _col10, _col11 + Merge Join Operator [MERGEJOIN_284] (rows=187573258 width=764) + Conds:RS_74._col1=RS_317._col0(Inner),Output:["_col2","_col5","_col6","_col7","_col8","_col9","_col10","_col11"] + <-Map 26 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_317] + PartitionCols:_col0 + Select Operator [SEL_316] (rows=80000000 width=656) + Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"] + Filter Operator [FIL_315] (rows=80000000 width=656) + predicate:(c_customer_id is not null and c_customer_sk is not null) + TableScan [TS_68] (rows=80000000 width=656) + default@customer,customer,Tbl:COMPLETE,Col:COMPLETE,Output:["c_customer_sk","c_customer_id","c_first_name","c_last_name","c_preferred_cust_flag","c_birth_country","c_login","c_email_address"] + <-Reducer 18 [SIMPLE_EDGE] + SHUFFLE [RS_74] + PartitionCols:_col1 + Merge Join Operator [MERGEJOIN_283] (rows=187573258 width=115) + Conds:RS_348._col0=RS_294._col0(Inner),Output:["_col1","_col2"] + <-Map 21 [SIMPLE_EDGE] vectorized + PARTITION_ONLY_SHUFFLE [RS_294] + PartitionCols:_col0 + Select Operator [SEL_291] (rows=652 width=4) + Output:["_col0"] + Filter Operator [FIL_288] (rows=652 width=8) + predicate:((d_year = 2002) and d_date_sk is not null) + TableScan [TS_65] (rows=73049 width=8) + default@date_dim,date_dim,Tbl:COMPLETE,Col:COMPLETE,Output:["d_date_sk","d_year"] + <-Map 17 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_348] + PartitionCols:_col0 + Select Operator [SEL_347] (rows=525327388 width=119) + Output:["_col0","_col1","_col2"] + Filter Operator [FIL_346] (rows=525327388 width=221) + predicate:((ss_sold_date_sk BETWEEN DynamicValue(RS_72_date_dim_d_date_sk_min) AND DynamicValue(RS_72_date_dim_d_date_sk_max) and in_bloom_filter(ss_sold_date_sk, DynamicValue(RS_72_date_dim_d_date_sk_bloom_filter))) and ss_customer_sk is not null and ss_sold_date_sk is not null) + TableScan [TS_62] (rows=575995635 width=221) + default@store_sales,store_sales,Tbl:COMPLETE,Col:COMPLETE,Output:["ss_sold_date_sk","ss_customer_sk","ss_ext_discount_amt","ss_ext_list_price"] + <-Reducer 22 [BROADCAST_EDGE] vectorized + BROADCAST [RS_345] + Group By Operator [GBY_344] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] + <-Map 21 [CUSTOM_SIMPLE_EDGE] vectorized + PARTITION_ONLY_SHUFFLE [RS_306] + Group By Operator [GBY_302] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] + Select Operator [SEL_295] (rows=652 width=4) + Output:["_col0"] + Please refer to the previous Select Operator [SEL_291] <-Reducer 6 [ONE_TO_ONE_EDGE] - FORWARD [RS_92] + FORWARD [RS_88] PartitionCols:_col2 - Merge Join Operator [MERGEJOIN_290] (rows=20485011 width=436) - Conds:RS_89._col2=RS_347._col0(Inner),Output:["_col1","_col2","_col3","_col5"] + Merge Join Operator [MERGEJOIN_286] (rows=20485011 width=440) + Conds:RS_85._col2=RS_343._col0(Inner),Output:["_col1","_col2","_col3","_col5","_col6"] <-Reducer 16 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_347] + SHUFFLE [RS_343] PartitionCols:_col0 - Select Operator [SEL_346] (rows=17130654 width=212) - Output:["_col0","_col1"] - Filter Operator [FIL_345] (rows=17130654 width=212) + Select Operator [SEL_342] (rows=17130654 width=216) + Output:["_col0","_col1","_col2"] + Filter Operator [FIL_341] (rows=17130654 width=212) predicate:(_col7 > 0) - Select Operator [SEL_344] (rows=51391963 width=212) + Select Operator [SEL_340] (rows=51391963 width=212) Output:["_col0","_col7"] - Group By Operator [GBY_343] (rows=51391963 width=764) + Group By Operator [GBY_339] (rows=51391963 width=764) Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"],aggregations:["sum(VALUE._col0)"],keys:KEY._col0, KEY._col1, KEY._col2, KEY._col3, KEY._col4, KEY._col5, KEY._col6 <-Reducer 15 [SIMPLE_EDGE] - SHUFFLE [RS_61] + SHUFFLE [RS_58] PartitionCols:_col0, _col1, _col2, _col3, _col4, _col5, _col6 - Group By Operator [GBY_60] (rows=51391963 width=764) - Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"],aggregations:["sum(_col7)"],keys:_col0, _col1, _col2, _col3, _col4, _col5, _col6 - Select Operator [SEL_58] (rows=51391963 width=875) - Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"] - Merge Join Operator [MERGEJOIN_286] (rows=51391963 width=875) - Conds:RS_55._col1=RS_324._col0(Inner),Output:["_col2","_col3","_col7","_col8","_col9","_col10","_col11","_col12","_col13"] - <-Map 26 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_324] - PartitionCols:_col0 - Please refer to the previous Select Operator [SEL_320] - <-Reducer 14 [SIMPLE_EDGE] - SHUFFLE [RS_55] - PartitionCols:_col1 - Merge Join Operator [MERGEJOIN_285] (rows=51391963 width=227) - Conds:RS_342._col0=RS_304._col0(Inner),Output:["_col1","_col2","_col3"] - <-Map 21 [SIMPLE_EDGE] vectorized - PARTITION_ONLY_SHUFFLE [RS_304] - PartitionCols:_col0 - Select Operator [SEL_297] (rows=652 width=8) - Output:["_col0"] - Filter Operator [FIL_294] (rows=652 width=8) - predicate:((d_year = 2001) and d_date_sk is not null) - Please refer to the previous TableScan [TS_68] - <-Map 13 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_342] - PartitionCols:_col0 - Select Operator [SEL_341] (rows=143930993 width=231) - Output:["_col0","_col1","_col2","_col3"] - Filter Operator [FIL_340] (rows=143930993 width=231) - predicate:((ws_sold_date_sk BETWEEN DynamicValue(RS_53_date_dim_d_date_sk_min) AND DynamicValue(RS_53_date_dim_d_date_sk_max) and in_bloom_filter(ws_sold_date_sk, DynamicValue(RS_53_date_dim_d_date_sk_bloom_filter))) and ws_bill_customer_sk is not null and ws_sold_date_sk is not null) - TableScan [TS_43] (rows=144002668 width=231) - default@web_sales,web_sales,Tbl:COMPLETE,Col:COMPLETE,Output:["ws_sold_date_sk","ws_bill_customer_sk","ws_ext_discount_amt","ws_ext_list_price"] - <-Reducer 25 [BROADCAST_EDGE] vectorized - BROADCAST [RS_339] - Group By Operator [GBY_338] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] - <-Map 21 [CUSTOM_SIMPLE_EDGE] vectorized - PARTITION_ONLY_SHUFFLE [RS_313] - Group By Operator [GBY_309] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] - Select Operator [SEL_305] (rows=652 width=4) - Output:["_col0"] - Please refer to the previous Select Operator [SEL_297] + Group By Operator [GBY_57] (rows=51391963 width=764) + Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"],aggregations:["sum(_col2)"],keys:_col5, _col6, _col7, _col8, _col9, _col10, _col11 + Merge Join Operator [MERGEJOIN_282] (rows=51391963 width=764) + Conds:RS_53._col1=RS_320._col0(Inner),Output:["_col2","_col5","_col6","_col7","_col8","_col9","_col10","_col11"] + <-Map 26 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_320] + PartitionCols:_col0 + Please refer to the previous Select Operator [SEL_316] + <-Reducer 14 [SIMPLE_EDGE] + SHUFFLE [RS_53] + PartitionCols:_col1 + Merge Join Operator [MERGEJOIN_281] (rows=51391963 width=115) + Conds:RS_338._col0=RS_300._col0(Inner),Output:["_col1","_col2"] + <-Map 21 [SIMPLE_EDGE] vectorized + PARTITION_ONLY_SHUFFLE [RS_300] + PartitionCols:_col0 + Select Operator [SEL_293] (rows=652 width=4) + Output:["_col0"] + Filter Operator [FIL_290] (rows=652 width=8) + predicate:((d_year = 2001) and d_date_sk is not null) + Please refer to the previous TableScan [TS_65] + <-Map 13 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_338] + PartitionCols:_col0 + Select Operator [SEL_337] (rows=143930993 width=119) + Output:["_col0","_col1","_col2"] + Filter Operator [FIL_336] (rows=143930993 width=231) + predicate:((ws_sold_date_sk BETWEEN DynamicValue(RS_51_date_dim_d_date_sk_min) AND DynamicValue(RS_51_date_dim_d_date_sk_max) and in_bloom_filter(ws_sold_date_sk, DynamicValue(RS_51_date_dim_d_date_sk_bloom_filter))) and ws_bill_customer_sk is not null and ws_sold_date_sk is not null) + TableScan [TS_41] (rows=144002668 width=231) + default@web_sales,web_sales,Tbl:COMPLETE,Col:COMPLETE,Output:["ws_sold_date_sk","ws_bill_customer_sk","ws_ext_discount_amt","ws_ext_list_price"] + <-Reducer 25 [BROADCAST_EDGE] vectorized + BROADCAST [RS_335] + Group By Operator [GBY_334] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] + <-Map 21 [CUSTOM_SIMPLE_EDGE] vectorized + PARTITION_ONLY_SHUFFLE [RS_309] + Group By Operator [GBY_305] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] + Select Operator [SEL_301] (rows=652 width=4) + Output:["_col0"] + Please refer to the previous Select Operator [SEL_293] <-Reducer 5 [ONE_TO_ONE_EDGE] - FORWARD [RS_89] + FORWARD [RS_85] PartitionCols:_col2 - Merge Join Operator [MERGEJOIN_289] (rows=31888273 width=324) - Conds:RS_327._col0=RS_337._col0(Inner),Output:["_col1","_col2","_col3"] + Merge Join Operator [MERGEJOIN_285] (rows=31888273 width=324) + Conds:RS_323._col0=RS_333._col0(Inner),Output:["_col1","_col2","_col3"] <-Reducer 12 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_337] + SHUFFLE [RS_333] PartitionCols:_col0 - Select Operator [SEL_336] (rows=26666666 width=212) + Select Operator [SEL_332] (rows=26666666 width=212) Output:["_col0","_col1"] - Filter Operator [FIL_335] (rows=26666666 width=212) + Filter Operator [FIL_331] (rows=26666666 width=212) predicate:(_col7 > 0) - Select Operator [SEL_334] (rows=80000000 width=212) + Select Operator [SEL_330] (rows=80000000 width=212) Output:["_col0","_col7"] - Group By Operator [GBY_333] (rows=80000000 width=764) + Group By Operator [GBY_329] (rows=80000000 width=764) Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"],aggregations:["sum(VALUE._col0)"],keys:KEY._col0, KEY._col1, KEY._col2, KEY._col3, KEY._col4, KEY._col5, KEY._col6 <-Reducer 11 [SIMPLE_EDGE] - SHUFFLE [RS_39] + SHUFFLE [RS_37] PartitionCols:_col0, _col1, _col2, _col3, _col4, _col5, _col6 - Group By Operator [GBY_38] (rows=80000000 width=764) - Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"],aggregations:["sum(_col7)"],keys:_col0, _col1, _col2, _col3, _col4, _col5, _col6 - Select Operator [SEL_36] (rows=187573258 width=847) - Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"] - Merge Join Operator [MERGEJOIN_284] (rows=187573258 width=847) - Conds:RS_33._col1=RS_323._col0(Inner),Output:["_col2","_col3","_col7","_col8","_col9","_col10","_col11","_col12","_col13"] - <-Map 26 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_323] - PartitionCols:_col0 - Please refer to the previous Select Operator [SEL_320] - <-Reducer 10 [SIMPLE_EDGE] - SHUFFLE [RS_33] - PartitionCols:_col1 - Merge Join Operator [MERGEJOIN_283] (rows=187573258 width=199) - Conds:RS_332._col0=RS_302._col0(Inner),Output:["_col1","_col2","_col3"] - <-Map 21 [SIMPLE_EDGE] vectorized - PARTITION_ONLY_SHUFFLE [RS_302] - PartitionCols:_col0 - Select Operator [SEL_296] (rows=652 width=8) - Output:["_col0"] - Filter Operator [FIL_293] (rows=652 width=8) - predicate:((d_year = 2001) and d_date_sk is not null) - Please refer to the previous TableScan [TS_68] - <-Map 9 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_332] - PartitionCols:_col0 - Select Operator [SEL_331] (rows=525327388 width=221) - Output:["_col0","_col1","_col2","_col3"] - Filter Operator [FIL_330] (rows=525327388 width=221) - predicate:((ss_sold_date_sk BETWEEN DynamicValue(RS_31_date_dim_d_date_sk_min) AND DynamicValue(RS_31_date_dim_d_date_sk_max) and in_bloom_filter(ss_sold_date_sk, DynamicValue(RS_31_date_dim_d_date_sk_bloom_filter))) and ss_customer_sk is not null and ss_sold_date_sk is not null) - TableScan [TS_21] (rows=575995635 width=221) - default@store_sales,store_sales,Tbl:COMPLETE,Col:COMPLETE,Output:["ss_sold_date_sk","ss_customer_sk","ss_ext_discount_amt","ss_ext_list_price"] - <-Reducer 24 [BROADCAST_EDGE] vectorized - BROADCAST [RS_329] - Group By Operator [GBY_328] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] - <-Map 21 [CUSTOM_SIMPLE_EDGE] vectorized - PARTITION_ONLY_SHUFFLE [RS_312] - Group By Operator [GBY_308] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] - Select Operator [SEL_303] (rows=652 width=4) - Output:["_col0"] - Please refer to the previous Select Operator [SEL_296] + Group By Operator [GBY_36] (rows=80000000 width=764) + Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"],aggregations:["sum(_col2)"],keys:_col5, _col6, _col7, _col8, _col9, _col10, _col11 + Merge Join Operator [MERGEJOIN_280] (rows=187573258 width=764) + Conds:RS_32._col1=RS_319._col0(Inner),Output:["_col2","_col5","_col6","_col7","_col8","_col9","_col10","_col11"] + <-Map 26 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_319] + PartitionCols:_col0 + Please refer to the previous Select Operator [SEL_316] + <-Reducer 10 [SIMPLE_EDGE] + SHUFFLE [RS_32] + PartitionCols:_col1 + Merge Join Operator [MERGEJOIN_279] (rows=187573258 width=115) + Conds:RS_328._col0=RS_298._col0(Inner),Output:["_col1","_col2"] + <-Map 21 [SIMPLE_EDGE] vectorized + PARTITION_ONLY_SHUFFLE [RS_298] + PartitionCols:_col0 + Select Operator [SEL_292] (rows=652 width=4) + Output:["_col0"] + Filter Operator [FIL_289] (rows=652 width=8) + predicate:((d_year = 2001) and d_date_sk is not null) + Please refer to the previous TableScan [TS_65] + <-Map 9 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_328] + PartitionCols:_col0 + Select Operator [SEL_327] (rows=525327388 width=119) + Output:["_col0","_col1","_col2"] + Filter Operator [FIL_326] (rows=525327388 width=221) + predicate:((ss_sold_date_sk BETWEEN DynamicValue(RS_30_date_dim_d_date_sk_min) AND DynamicValue(RS_30_date_dim_d_date_sk_max) and in_bloom_filter(ss_sold_date_sk, DynamicValue(RS_30_date_dim_d_date_sk_bloom_filter))) and ss_customer_sk is not null and ss_sold_date_sk is not null) + TableScan [TS_20] (rows=575995635 width=221) + default@store_sales,store_sales,Tbl:COMPLETE,Col:COMPLETE,Output:["ss_sold_date_sk","ss_customer_sk","ss_ext_discount_amt","ss_ext_list_price"] + <-Reducer 24 [BROADCAST_EDGE] vectorized + BROADCAST [RS_325] + Group By Operator [GBY_324] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] + <-Map 21 [CUSTOM_SIMPLE_EDGE] vectorized + PARTITION_ONLY_SHUFFLE [RS_308] + Group By Operator [GBY_304] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] + Select Operator [SEL_299] (rows=652 width=4) + Output:["_col0"] + Please refer to the previous Select Operator [SEL_292] <-Reducer 4 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_327] + SHUFFLE [RS_323] PartitionCols:_col0 - Select Operator [SEL_326] (rows=51391963 width=212) + Select Operator [SEL_322] (rows=51391963 width=212) Output:["_col0","_col1"] - Group By Operator [GBY_325] (rows=51391963 width=764) + Group By Operator [GBY_321] (rows=51391963 width=764) Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"],aggregations:["sum(VALUE._col0)"],keys:KEY._col0, KEY._col1, KEY._col2, KEY._col3, KEY._col4, KEY._col5, KEY._col6 <-Reducer 3 [SIMPLE_EDGE] - SHUFFLE [RS_18] + SHUFFLE [RS_17] PartitionCols:_col0, _col1, _col2, _col3, _col4, _col5, _col6 - Group By Operator [GBY_17] (rows=51391963 width=764) - Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"],aggregations:["sum(_col7)"],keys:_col0, _col1, _col2, _col3, _col4, _col5, _col6 - Select Operator [SEL_15] (rows=51391963 width=875) - Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"] - Merge Join Operator [MERGEJOIN_282] (rows=51391963 width=875) - Conds:RS_12._col1=RS_322._col0(Inner),Output:["_col2","_col3","_col7","_col8","_col9","_col10","_col11","_col12","_col13"] - <-Map 26 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_322] - PartitionCols:_col0 - Please refer to the previous Select Operator [SEL_320] - <-Reducer 2 [SIMPLE_EDGE] - SHUFFLE [RS_12] - PartitionCols:_col1 - Merge Join Operator [MERGEJOIN_281] (rows=51391963 width=227) - Conds:RS_318._col0=RS_300._col0(Inner),Output:["_col1","_col2","_col3"] - <-Map 21 [SIMPLE_EDGE] vectorized - PARTITION_ONLY_SHUFFLE [RS_300] - PartitionCols:_col0 - Please refer to the previous Select Operator [SEL_295] - <-Map 1 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_318] - PartitionCols:_col0 - Select Operator [SEL_317] (rows=143930993 width=231) - Output:["_col0","_col1","_col2","_col3"] - Filter Operator [FIL_316] (rows=143930993 width=231) - predicate:((ws_sold_date_sk BETWEEN DynamicValue(RS_10_date_dim_d_date_sk_min) AND DynamicValue(RS_10_date_dim_d_date_sk_max) and in_bloom_filter(ws_sold_date_sk, DynamicValue(RS_10_date_dim_d_date_sk_bloom_filter))) and ws_bill_customer_sk is not null and ws_sold_date_sk is not null) - TableScan [TS_0] (rows=144002668 width=231) - default@web_sales,web_sales,Tbl:COMPLETE,Col:COMPLETE,Output:["ws_sold_date_sk","ws_bill_customer_sk","ws_ext_discount_amt","ws_ext_list_price"] - <-Reducer 23 [BROADCAST_EDGE] vectorized - BROADCAST [RS_315] - Group By Operator [GBY_314] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] - <-Map 21 [CUSTOM_SIMPLE_EDGE] vectorized - PARTITION_ONLY_SHUFFLE [RS_311] - Group By Operator [GBY_307] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] - Select Operator [SEL_301] (rows=652 width=4) - Output:["_col0"] - Please refer to the previous Select Operator [SEL_295] + Group By Operator [GBY_16] (rows=51391963 width=764) + Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7"],aggregations:["sum(_col2)"],keys:_col5, _col6, _col7, _col8, _col9, _col10, _col11 + Merge Join Operator [MERGEJOIN_278] (rows=51391963 width=764) + Conds:RS_12._col1=RS_318._col0(Inner),Output:["_col2","_col5","_col6","_col7","_col8","_col9","_col10","_col11"] + <-Map 26 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_318] + PartitionCols:_col0 + Please refer to the previous Select Operator [SEL_316] + <-Reducer 2 [SIMPLE_EDGE] + SHUFFLE [RS_12] + PartitionCols:_col1 + Merge Join Operator [MERGEJOIN_277] (rows=51391963 width=115) + Conds:RS_314._col0=RS_296._col0(Inner),Output:["_col1","_col2"] + <-Map 21 [SIMPLE_EDGE] vectorized + PARTITION_ONLY_SHUFFLE [RS_296] + PartitionCols:_col0 + Please refer to the previous Select Operator [SEL_291] + <-Map 1 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_314] + PartitionCols:_col0 + Select Operator [SEL_313] (rows=143930993 width=119) + Output:["_col0","_col1","_col2"] + Filter Operator [FIL_312] (rows=143930993 width=231) + predicate:((ws_sold_date_sk BETWEEN DynamicValue(RS_10_date_dim_d_date_sk_min) AND DynamicValue(RS_10_date_dim_d_date_sk_max) and in_bloom_filter(ws_sold_date_sk, DynamicValue(RS_10_date_dim_d_date_sk_bloom_filter))) and ws_bill_customer_sk is not null and ws_sold_date_sk is not null) + TableScan [TS_0] (rows=144002668 width=231) + default@web_sales,web_sales,Tbl:COMPLETE,Col:COMPLETE,Output:["ws_sold_date_sk","ws_bill_customer_sk","ws_ext_discount_amt","ws_ext_list_price"] + <-Reducer 23 [BROADCAST_EDGE] vectorized + BROADCAST [RS_311] + Group By Operator [GBY_310] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] + <-Map 21 [CUSTOM_SIMPLE_EDGE] vectorized + PARTITION_ONLY_SHUFFLE [RS_307] + Group By Operator [GBY_303] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] + Select Operator [SEL_297] (rows=652 width=4) + Output:["_col0"] + Please refer to the previous Select Operator [SEL_291]
http://git-wip-us.apache.org/repos/asf/hive/blob/55887646/ql/src/test/results/clientpositive/perf/tez/query12.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/tez/query12.q.out b/ql/src/test/results/clientpositive/perf/tez/query12.q.out index 75156be..b6e4999 100644 --- a/ql/src/test/results/clientpositive/perf/tez/query12.q.out +++ b/ql/src/test/results/clientpositive/perf/tez/query12.q.out @@ -109,9 +109,9 @@ Stage-0 SHUFFLE [RS_17] PartitionCols:_col0, _col1, _col2, _col3, _col4 Group By Operator [GBY_16] (rows=138600 width=689) - Output:["_col0","_col1","_col2","_col3","_col4","_col5"],aggregations:["sum(_col2)"],keys:_col10, _col9, _col6, _col7, _col8 + Output:["_col0","_col1","_col2","_col3","_col4","_col5"],aggregations:["sum(_col2)"],keys:_col9, _col8, _col5, _col6, _col7 Merge Join Operator [MERGEJOIN_58] (rows=4798568 width=689) - Conds:RS_12._col1=RS_69._col0(Inner),Output:["_col2","_col6","_col7","_col8","_col9","_col10"] + Conds:RS_12._col1=RS_69._col0(Inner),Output:["_col2","_col5","_col6","_col7","_col8","_col9"] <-Map 9 [SIMPLE_EDGE] vectorized SHUFFLE [RS_69] PartitionCols:_col0 @@ -129,7 +129,7 @@ Stage-0 <-Map 7 [SIMPLE_EDGE] vectorized PARTITION_ONLY_SHUFFLE [RS_61] PartitionCols:_col0 - Select Operator [SEL_60] (rows=8116 width=98) + Select Operator [SEL_60] (rows=8116 width=4) Output:["_col0"] Filter Operator [FIL_59] (rows=8116 width=98) predicate:(CAST( d_date AS TIMESTAMP) BETWEEN TIMESTAMP'2001-01-12 00:00:00' AND TIMESTAMP'2001-02-11 00:00:00' and d_date_sk is not null) http://git-wip-us.apache.org/repos/asf/hive/blob/55887646/ql/src/test/results/clientpositive/perf/tez/query13.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/tez/query13.q.out b/ql/src/test/results/clientpositive/perf/tez/query13.q.out index 0e02e9a..d6e86c4 100644 --- a/ql/src/test/results/clientpositive/perf/tez/query13.q.out +++ b/ql/src/test/results/clientpositive/perf/tez/query13.q.out @@ -115,17 +115,17 @@ POSTHOOK: Output: hdfs://### HDFS PATH ### Plan optimized by CBO. Vertex dependency in root stage -Map 8 <- Reducer 10 (BROADCAST_EDGE), Reducer 12 (BROADCAST_EDGE), Reducer 14 (BROADCAST_EDGE), Reducer 16 (BROADCAST_EDGE) -Reducer 10 <- Map 9 (CUSTOM_SIMPLE_EDGE) +Map 9 <- Reducer 12 (BROADCAST_EDGE), Reducer 14 (BROADCAST_EDGE), Reducer 16 (BROADCAST_EDGE), Reducer 8 (BROADCAST_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), Map 8 (SIMPLE_EDGE) -Reducer 3 <- Map 9 (SIMPLE_EDGE), Reducer 2 (SIMPLE_EDGE) +Reducer 2 <- Map 1 (SIMPLE_EDGE), Map 9 (SIMPLE_EDGE) +Reducer 3 <- Map 10 (SIMPLE_EDGE), Reducer 2 (SIMPLE_EDGE) Reducer 4 <- Map 11 (SIMPLE_EDGE), Reducer 3 (SIMPLE_EDGE) Reducer 5 <- Map 13 (SIMPLE_EDGE), Reducer 4 (SIMPLE_EDGE) Reducer 6 <- Map 15 (SIMPLE_EDGE), Reducer 5 (SIMPLE_EDGE) Reducer 7 <- Reducer 6 (CUSTOM_SIMPLE_EDGE) +Reducer 8 <- Map 1 (CUSTOM_SIMPLE_EDGE) Stage-0 Fetch Operator @@ -140,131 +140,131 @@ Stage-0 <-Reducer 6 [CUSTOM_SIMPLE_EDGE] PARTITION_ONLY_SHUFFLE [RS_37] Group By Operator [GBY_36] (rows=1 width=256) - Output:["_col0","_col1","_col2","_col3","_col4","_col5"],aggregations:["sum(_col6)","count(_col6)","sum(_col8)","count(_col8)","sum(_col9)","count(_col9)"] - Select Operator [SEL_35] (rows=4851 width=183) - Output:["_col6","_col8","_col9"] - Filter Operator [FIL_34] (rows=4851 width=183) - predicate:(((_col19 = 'D') and (_col20 = 'Primary') and _col7 BETWEEN 50 AND 100 and (_col14 = 1)) or ((_col19 = 'M') and (_col20 = '4 yr Degree') and _col7 BETWEEN 100 AND 150 and (_col14 = 3)) or ((_col19 = 'U') and (_col20 = 'Advanced Degree') and _col7 BETWEEN 150 AND 200 and (_col14 = 1))) - Merge Join Operator [MERGEJOIN_121] (rows=58239 width=183) - Conds:RS_31._col2=RS_151._col0(Inner),Output:["_col6","_col7","_col8","_col9","_col14","_col19","_col20"] + Output:["_col0","_col1","_col2","_col3","_col4","_col5"],aggregations:["sum(_col6)","count(_col6)","sum(_col7)","count(_col7)","sum(_col8)","count(_col8)"] + Select Operator [SEL_35] (rows=40950 width=44) + Output:["_col6","_col7","_col8"] + Filter Operator [FIL_34] (rows=40950 width=44) + predicate:((_col24 and _col25 and _col12 and _col17) or (_col26 and _col27 and _col13 and _col18) or (_col28 and _col29 and _col14 and _col18)) + Merge Join Operator [MERGEJOIN_121] (rows=218403 width=44) + Conds:RS_31._col2=RS_148._col0(Inner),Output:["_col6","_col7","_col8","_col12","_col13","_col14","_col17","_col18","_col24","_col25","_col26","_col27","_col28","_col29"] <-Map 15 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_151] + SHUFFLE [RS_148] PartitionCols:_col0 - Select Operator [SEL_150] (rows=265971 width=183) - Output:["_col0","_col1","_col2"] - Filter Operator [FIL_149] (rows=265971 width=183) + Select Operator [SEL_147] (rows=265971 width=28) + Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6"] + Filter Operator [FIL_146] (rows=265971 width=183) predicate:((cd_education_status) IN ('4 yr Degree', 'Primary', 'Advanced Degree') and (cd_marital_status) IN ('M', 'D', 'U') and cd_demo_sk is not null) TableScan [TS_15] (rows=1861800 width=183) default@customer_demographics,customer_demographics,Tbl:COMPLETE,Col:COMPLETE,Output:["cd_demo_sk","cd_marital_status","cd_education_status"] <-Reducer 5 [SIMPLE_EDGE] SHUFFLE [RS_31] PartitionCols:_col2 - Filter Operator [FIL_30] (rows=58239 width=90) - predicate:(((_col16) IN ('KY', 'GA', 'NM') and _col10 BETWEEN 100 AND 200) or ((_col16) IN ('MT', 'OR', 'IN') and _col10 BETWEEN 150 AND 300) or ((_col16) IN ('WI', 'MO', 'WV') and _col10 BETWEEN 50 AND 250)) - Merge Join Operator [MERGEJOIN_120] (rows=291204 width=90) - Conds:RS_27._col4=RS_143._col0(Inner),Output:["_col2","_col6","_col7","_col8","_col9","_col10","_col14","_col16"] + Filter Operator [FIL_30] (rows=218403 width=44) + predicate:((_col20 and _col9) or (_col21 and _col10) or (_col22 and _col11)) + Merge Join Operator [MERGEJOIN_120] (rows=291204 width=44) + Conds:RS_27._col4=RS_140._col0(Inner),Output:["_col2","_col6","_col7","_col8","_col9","_col10","_col11","_col12","_col13","_col14","_col17","_col18","_col20","_col21","_col22"] <-Map 13 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_143] + SHUFFLE [RS_140] PartitionCols:_col0 - Select Operator [SEL_142] (rows=3529412 width=187) - Output:["_col0","_col1"] - Filter Operator [FIL_141] (rows=3529412 width=187) + Select Operator [SEL_139] (rows=3529412 width=16) + Output:["_col0","_col1","_col2","_col3"] + Filter Operator [FIL_138] (rows=3529412 width=187) predicate:((ca_country = 'United States') and (ca_state) IN ('KY', 'GA', 'NM', 'MT', 'OR', 'IN', 'WI', 'MO', 'WV') and ca_address_sk is not null) TableScan [TS_12] (rows=40000000 width=187) default@customer_address,customer_address,Tbl:COMPLETE,Col:COMPLETE,Output:["ca_address_sk","ca_state","ca_country"] <-Reducer 4 [SIMPLE_EDGE] SHUFFLE [RS_27] PartitionCols:_col4 - Merge Join Operator [MERGEJOIN_119] (rows=3300311 width=145) - Conds:RS_24._col3=RS_135._col0(Inner),Output:["_col2","_col4","_col6","_col7","_col8","_col9","_col10","_col14"] + Merge Join Operator [MERGEJOIN_119] (rows=3300311 width=104) + Conds:RS_24._col3=RS_132._col0(Inner),Output:["_col2","_col4","_col6","_col7","_col8","_col9","_col10","_col11","_col12","_col13","_col14","_col17","_col18"] <-Map 11 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_135] + SHUFFLE [RS_132] PartitionCols:_col0 - Select Operator [SEL_134] (rows=1309 width=8) - Output:["_col0","_col1"] - Filter Operator [FIL_133] (rows=1309 width=8) + Select Operator [SEL_131] (rows=1309 width=12) + Output:["_col0","_col1","_col2"] + Filter Operator [FIL_130] (rows=1309 width=8) predicate:((hd_dep_count) IN (3, 1) and hd_demo_sk is not null) TableScan [TS_9] (rows=7200 width=8) default@household_demographics,household_demographics,Tbl:COMPLETE,Col:COMPLETE,Output:["hd_demo_sk","hd_dep_count"] <-Reducer 3 [SIMPLE_EDGE] SHUFFLE [RS_24] PartitionCols:_col3 - Merge Join Operator [MERGEJOIN_118] (rows=18152968 width=405) - Conds:RS_21._col1=RS_127._col0(Inner),Output:["_col2","_col3","_col4","_col6","_col7","_col8","_col9","_col10"] - <-Map 9 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_127] + Merge Join Operator [MERGEJOIN_118] (rows=18152968 width=233) + Conds:RS_21._col5=RS_159._col0(Inner),Output:["_col2","_col3","_col4","_col6","_col7","_col8","_col9","_col10","_col11","_col12","_col13","_col14"] + <-Map 10 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_159] PartitionCols:_col0 - Select Operator [SEL_126] (rows=652 width=8) + Select Operator [SEL_158] (rows=1704 width=4) Output:["_col0"] - Filter Operator [FIL_125] (rows=652 width=8) - predicate:((d_year = 2001) and d_date_sk is not null) - TableScan [TS_6] (rows=73049 width=8) - default@date_dim,date_dim,Tbl:COMPLETE,Col:COMPLETE,Output:["d_date_sk","d_year"] + Filter Operator [FIL_157] (rows=1704 width=4) + predicate:s_store_sk is not null + TableScan [TS_6] (rows=1704 width=4) + default@store,store,Tbl:COMPLETE,Col:COMPLETE,Output:["s_store_sk"] <-Reducer 2 [SIMPLE_EDGE] SHUFFLE [RS_21] - PartitionCols:_col1 - Merge Join Operator [MERGEJOIN_117] (rows=50840141 width=446) - Conds:RS_124._col0=RS_159._col4(Inner),Output:["_col1","_col2","_col3","_col4","_col6","_col7","_col8","_col9","_col10"] + PartitionCols:_col5 + Merge Join Operator [MERGEJOIN_117] (rows=18152968 width=237) + Conds:RS_124._col0=RS_156._col0(Inner),Output:["_col2","_col3","_col4","_col5","_col6","_col7","_col8","_col9","_col10","_col11","_col12","_col13","_col14"] <-Map 1 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_124] + PARTITION_ONLY_SHUFFLE [RS_124] PartitionCols:_col0 - Select Operator [SEL_123] (rows=1704 width=4) + Select Operator [SEL_123] (rows=652 width=4) Output:["_col0"] - Filter Operator [FIL_122] (rows=1704 width=4) - predicate:s_store_sk is not null - TableScan [TS_0] (rows=1704 width=4) - default@store,store,Tbl:COMPLETE,Col:COMPLETE,Output:["s_store_sk"] - <-Map 8 [SIMPLE_EDGE] vectorized - SHUFFLE [RS_159] - PartitionCols:_col4 - Select Operator [SEL_158] (rows=50840141 width=450) - Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7","_col8","_col9"] - Filter Operator [FIL_157] (rows=50840141 width=450) - predicate:((ss_addr_sk BETWEEN DynamicValue(RS_28_customer_address_ca_address_sk_min) AND DynamicValue(RS_28_customer_address_ca_address_sk_max) and in_bloom_filter(ss_addr_sk, DynamicValue(RS_28_customer_address_ca_address_sk_bloom_filter))) and (ss_cdemo_sk BETWEEN DynamicValue(RS_32_customer_demographics_cd_demo_sk_min) AND DynamicValue(RS_32_customer_demographics_cd_demo_sk_max) and in_bloom_filter(ss_cdemo_sk, DynamicValue(RS_32_customer_demographics_cd_demo_sk_bloom_filter))) and (ss_hdemo_sk BETWEEN DynamicValue(RS_25_household_demographics_hd_demo_sk_min) AND DynamicValue(RS_25_household_demographics_hd_demo_sk_max) and in_bloom_filter(ss_hdemo_sk, DynamicValue(RS_25_household_demographics_hd_demo_sk_bloom_filter))) and (ss_net_profit BETWEEN 100 AND 200 or ss_net_profit BETWEEN 150 AND 300 or ss_net_profit BETWEEN 50 AND 250) and (ss_sales_price BETWEEN 100 AND 150 or ss_sales_price BETWEEN 50 AND 100 or ss_sales_price BETWEEN 1 50 AND 200) and (ss_sold_date_sk BETWEEN DynamicValue(RS_22_date_dim_d_date_sk_min) AND DynamicValue(RS_22_date_dim_d_date_sk_max) and in_bloom_filter(ss_sold_date_sk, DynamicValue(RS_22_date_dim_d_date_sk_bloom_filter))) and ss_addr_sk is not null and ss_cdemo_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) + Filter Operator [FIL_122] (rows=652 width=8) + predicate:((d_year = 2001) and d_date_sk is not null) + TableScan [TS_0] (rows=73049 width=8) + default@date_dim,date_dim,Tbl:COMPLETE,Col:COMPLETE,Output:["d_date_sk","d_year"] + <-Map 9 [SIMPLE_EDGE] vectorized + SHUFFLE [RS_156] + PartitionCols:_col0 + Select Operator [SEL_155] (rows=50840141 width=260) + Output:["_col0","_col1","_col2","_col3","_col4","_col5","_col6","_col7","_col8","_col9","_col10","_col11","_col12","_col13"] + Filter Operator [FIL_154] (rows=50840141 width=450) + predicate:((ss_addr_sk BETWEEN DynamicValue(RS_28_customer_address_ca_address_sk_min) AND DynamicValue(RS_28_customer_address_ca_address_sk_max) and in_bloom_filter(ss_addr_sk, DynamicValue(RS_28_customer_address_ca_address_sk_bloom_filter))) and (ss_cdemo_sk BETWEEN DynamicValue(RS_32_customer_demographics_cd_demo_sk_min) AND DynamicValue(RS_32_customer_demographics_cd_demo_sk_max) and in_bloom_filter(ss_cdemo_sk, DynamicValue(RS_32_customer_demographics_cd_demo_sk_bloom_filter))) and (ss_hdemo_sk BETWEEN DynamicValue(RS_25_household_demographics_hd_demo_sk_min) AND DynamicValue(RS_25_household_demographics_hd_demo_sk_max) and in_bloom_filter(ss_hdemo_sk, DynamicValue(RS_25_household_demographics_hd_demo_sk_bloom_filter))) and (ss_net_profit BETWEEN 100 AND 200 or ss_net_profit BETWEEN 150 AND 300 or ss_net_profit BETWEEN 50 AND 250) and (ss_sales_price BETWEEN 100 AND 150 or ss_sales_price BETWEEN 50 AND 100 or ss_sales_price BETWEEN 1 50 AND 200) 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_addr_sk is not null and ss_cdemo_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=450) default@store_sales,store_sales,Tbl:COMPLETE,Col:COMPLETE,Output:["ss_sold_date_sk","ss_cdemo_sk","ss_hdemo_sk","ss_addr_sk","ss_store_sk","ss_quantity","ss_sales_price","ss_ext_sales_price","ss_ext_wholesale_cost","ss_net_profit"] - <-Reducer 10 [BROADCAST_EDGE] vectorized - BROADCAST [RS_132] - Group By Operator [GBY_131] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] - <-Map 9 [CUSTOM_SIMPLE_EDGE] vectorized - SHUFFLE [RS_130] - Group By Operator [GBY_129] (rows=1 width=12) - Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] - Select Operator [SEL_128] (rows=652 width=4) - Output:["_col0"] - Please refer to the previous Select Operator [SEL_126] <-Reducer 12 [BROADCAST_EDGE] vectorized - BROADCAST [RS_140] - Group By Operator [GBY_139] (rows=1 width=12) + BROADCAST [RS_137] + Group By Operator [GBY_136] (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_138] - Group By Operator [GBY_137] (rows=1 width=12) + SHUFFLE [RS_135] + Group By Operator [GBY_134] (rows=1 width=12) Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] - Select Operator [SEL_136] (rows=1309 width=4) + Select Operator [SEL_133] (rows=1309 width=4) Output:["_col0"] - Please refer to the previous Select Operator [SEL_134] + Please refer to the previous Select Operator [SEL_131] <-Reducer 14 [BROADCAST_EDGE] vectorized - BROADCAST [RS_148] - Group By Operator [GBY_147] (rows=1 width=12) + BROADCAST [RS_145] + Group By Operator [GBY_144] (rows=1 width=12) Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=3529412)"] <-Map 13 [CUSTOM_SIMPLE_EDGE] vectorized - SHUFFLE [RS_146] - Group By Operator [GBY_145] (rows=1 width=12) + SHUFFLE [RS_143] + Group By Operator [GBY_142] (rows=1 width=12) Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=3529412)"] - Select Operator [SEL_144] (rows=3529412 width=4) + Select Operator [SEL_141] (rows=3529412 width=4) Output:["_col0"] - Please refer to the previous Select Operator [SEL_142] + Please refer to the previous Select Operator [SEL_139] <-Reducer 16 [BROADCAST_EDGE] vectorized - BROADCAST [RS_156] - Group By Operator [GBY_155] (rows=1 width=12) + BROADCAST [RS_153] + Group By Operator [GBY_152] (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_154] - Group By Operator [GBY_153] (rows=1 width=12) + SHUFFLE [RS_151] + Group By Operator [GBY_150] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] + Select Operator [SEL_149] (rows=265971 width=4) + Output:["_col0"] + Please refer to the previous Select Operator [SEL_147] + <-Reducer 8 [BROADCAST_EDGE] vectorized + BROADCAST [RS_129] + Group By Operator [GBY_128] (rows=1 width=12) + Output:["_col0","_col1","_col2"],aggregations:["min(VALUE._col0)","max(VALUE._col1)","bloom_filter(VALUE._col2, expectedEntries=1000000)"] + <-Map 1 [CUSTOM_SIMPLE_EDGE] vectorized + PARTITION_ONLY_SHUFFLE [RS_127] + Group By Operator [GBY_126] (rows=1 width=12) Output:["_col0","_col1","_col2"],aggregations:["min(_col0)","max(_col0)","bloom_filter(_col0, expectedEntries=1000000)"] - Select Operator [SEL_152] (rows=265971 width=4) + Select Operator [SEL_125] (rows=652 width=4) Output:["_col0"] - Please refer to the previous Select Operator [SEL_150] + Please refer to the previous Select Operator [SEL_123] http://git-wip-us.apache.org/repos/asf/hive/blob/55887646/ql/src/test/results/clientpositive/perf/tez/query14.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/tez/query14.q.out b/ql/src/test/results/clientpositive/perf/tez/query14.q.out index 2c28023..2e99363 100644 --- a/ql/src/test/results/clientpositive/perf/tez/query14.q.out +++ b/ql/src/test/results/clientpositive/perf/tez/query14.q.out @@ -376,7 +376,7 @@ Stage-0 <-Map 96 [SIMPLE_EDGE] vectorized PARTITION_ONLY_SHUFFLE [RS_1623] PartitionCols:_col0 - Select Operator [SEL_1612] (rows=1957 width=8) + Select Operator [SEL_1612] (rows=1957 width=4) Output:["_col0"] Filter Operator [FIL_1611] (rows=1957 width=8) predicate:(d_date_sk is not null and d_year BETWEEN 1999 AND 2001) @@ -415,7 +415,7 @@ Stage-0 <-Map 24 [SIMPLE_EDGE] vectorized SHUFFLE [RS_1763] PartitionCols:_col0 - Select Operator [SEL_1758] (rows=1957 width=8) + Select Operator [SEL_1758] (rows=1957 width=4) Output:["_col0"] Filter Operator [FIL_1757] (rows=1957 width=8) predicate:(d_date_sk is not null and d_year BETWEEN 1998 AND 2000) @@ -454,7 +454,7 @@ Stage-0 <-Map 40 [SIMPLE_EDGE] vectorized SHUFFLE [RS_1791] PartitionCols:_col0 - Select Operator [SEL_1786] (rows=1957 width=8) + Select Operator [SEL_1786] (rows=1957 width=4) Output:["_col0"] Filter Operator [FIL_1785] (rows=1957 width=8) predicate:(d_date_sk is not null and d_year BETWEEN 1998 AND 2000) @@ -601,7 +601,7 @@ Stage-0 Select Operator [SEL_368] (rows=1 width=128) Output:["_col0","_col1","_col2","_col3"] Merge Join Operator [MERGEJOIN_1438] (rows=1 width=128) - Conds:RS_365._col1=RS_366._col0(Left Semi),Output:["_col2","_col3","_col8","_col9","_col10"] + Conds:RS_365._col1=RS_366._col0(Left Semi),Output:["_col2","_col3","_col6","_col7","_col8"] <-Reducer 79 [SIMPLE_EDGE] PARTITION_ONLY_SHUFFLE [RS_366] PartitionCols:_col0 @@ -639,9 +639,9 @@ Stage-0 SHUFFLE [RS_302] PartitionCols:_col0, _col1, _col2 Group By Operator [GBY_110] (rows=121728 width=19) - Output:["_col0","_col1","_col2","_col3"],aggregations:["count()"],keys:_col5, _col6, _col7 + Output:["_col0","_col1","_col2","_col3"],aggregations:["count()"],keys:_col4, _col5, _col6 Merge Join Operator [MERGEJOIN_1401] (rows=14628613 width=11) - Conds:RS_106._col1=RS_1689._col0(Inner),Output:["_col5","_col6","_col7"] + Conds:RS_106._col1=RS_1689._col0(Inner),Output:["_col4","_col5","_col6"] <-Map 69 [SIMPLE_EDGE] vectorized SHUFFLE [RS_1689] PartitionCols:_col0 @@ -690,9 +690,9 @@ Stage-0 SHUFFLE [RS_322] PartitionCols:_col0, _col1, _col2 Group By Operator [GBY_130] (rows=121728 width=19) - Output:["_col0","_col1","_col2","_col3"],aggregations:["count()"],keys:_col5, _col6, _col7 + Output:["_col0","_col1","_col2","_col3"],aggregations:["count()"],keys:_col4, _col5, _col6 Merge Join Operator [MERGEJOIN_1403] (rows=7620440 width=11) - Conds:RS_126._col1=RS_1690._col0(Inner),Output:["_col5","_col6","_col7"] + Conds:RS_126._col1=RS_1690._col0(Inner),Output:["_col4","_col5","_col6"] <-Map 69 [SIMPLE_EDGE] vectorized SHUFFLE [RS_1690] PartitionCols:_col0 @@ -741,9 +741,9 @@ Stage-0 SHUFFLE [RS_343] PartitionCols:_col0, _col1, _col2 Group By Operator [GBY_151] (rows=121728 width=19) - Output:["_col0","_col1","_col2","_col3"],aggregations:["count()"],keys:_col5, _col6, _col7 + Output:["_col0","_col1","_col2","_col3"],aggregations:["count()"],keys:_col4, _col5, _col6 Merge Join Operator [MERGEJOIN_1405] (rows=3828623 width=11) - Conds:RS_147._col1=RS_1691._col0(Inner),Output:["_col5","_col6","_col7"] + Conds:RS_147._col1=RS_1691._col0(Inner),Output:["_col4","_col5","_col6"] <-Map 69 [SIMPLE_EDGE] vectorized SHUFFLE [RS_1691] PartitionCols:_col0 @@ -785,7 +785,7 @@ Stage-0 FORWARD [RS_365] PartitionCols:_col1 Merge Join Operator [MERGEJOIN_1414] (rows=7790806 width=110) - Conds:RS_360._col1=RS_1692._col0(Inner),Output:["_col1","_col2","_col3","_col8","_col9","_col10"] + Conds:RS_360._col1=RS_1692._col0(Inner),Output:["_col1","_col2","_col3","_col6","_col7","_col8"] <-Map 69 [SIMPLE_EDGE] vectorized SHUFFLE [RS_1692] PartitionCols:_col0 @@ -802,7 +802,7 @@ Stage-0 <-Map 57 [SIMPLE_EDGE] vectorized PARTITION_ONLY_SHUFFLE [RS_1657] PartitionCols:_col0 - Select Operator [SEL_1654] (rows=50 width=12) + Select Operator [SEL_1654] (rows=50 width=4) Output:["_col0"] Filter Operator [FIL_1653] (rows=50 width=12) predicate:((d_moy = 11) and (d_year = 2000) and d_date_sk is not null) @@ -994,7 +994,7 @@ Stage-0 Select Operator [SEL_560] (rows=1 width=128) Output:["_col0","_col1","_col2","_col3"] Merge Join Operator [MERGEJOIN_1439] (rows=1 width=128) - Conds:RS_557._col1=RS_558._col0(Left Semi),Output:["_col2","_col3","_col8","_col9","_col10"] + Conds:RS_557._col1=RS_558._col0(Left Semi),Output:["_col2","_col3","_col6","_col7","_col8"] <-Reducer 84 [SIMPLE_EDGE] PARTITION_ONLY_SHUFFLE [RS_558] PartitionCols:_col0 @@ -1057,7 +1057,7 @@ Stage-0 FORWARD [RS_557] PartitionCols:_col1 Merge Join Operator [MERGEJOIN_1429] (rows=3942084 width=130) - Conds:RS_552._col1=RS_1694._col0(Inner),Output:["_col1","_col2","_col3","_col8","_col9","_col10"] + Conds:RS_552._col1=RS_1694._col0(Inner),Output:["_col1","_col2","_col3","_col6","_col7","_col8"] <-Map 69 [SIMPLE_EDGE] vectorized SHUFFLE [RS_1694] PartitionCols:_col0 @@ -1261,7 +1261,7 @@ Stage-0 Select Operator [SEL_177] (rows=1 width=128) Output:["_col0","_col1","_col2","_col3"] Merge Join Operator [MERGEJOIN_1437] (rows=1 width=128) - Conds:RS_174._col1=RS_175._col0(Left Semi),Output:["_col2","_col3","_col8","_col9","_col10"] + Conds:RS_174._col1=RS_175._col0(Left Semi),Output:["_col2","_col3","_col6","_col7","_col8"] <-Reducer 70 [SIMPLE_EDGE] PARTITION_ONLY_SHUFFLE [RS_175] PartitionCols:_col0 @@ -1324,7 +1324,7 @@ Stage-0 FORWARD [RS_174] PartitionCols:_col1 Merge Join Operator [MERGEJOIN_1399] (rows=15062131 width=15) - Conds:RS_169._col1=RS_1687._col0(Inner),Output:["_col1","_col2","_col3","_col8","_col9","_col10"] + Conds:RS_169._col1=RS_1687._col0(Inner),Output:["_col1","_col2","_col3","_col6","_col7","_col8"] <-Map 69 [SIMPLE_EDGE] vectorized SHUFFLE [RS_1687] PartitionCols:_col0 http://git-wip-us.apache.org/repos/asf/hive/blob/55887646/ql/src/test/results/clientpositive/perf/tez/query15.q.out ---------------------------------------------------------------------- diff --git a/ql/src/test/results/clientpositive/perf/tez/query15.q.out b/ql/src/test/results/clientpositive/perf/tez/query15.q.out index 565f101..c3c08a1 100644 --- a/ql/src/test/results/clientpositive/perf/tez/query15.q.out +++ b/ql/src/test/results/clientpositive/perf/tez/query15.q.out @@ -74,21 +74,21 @@ Stage-0 <-Reducer 3 [SIMPLE_EDGE] SHUFFLE [RS_25] PartitionCols:_col0 - Group By Operator [GBY_24] (rows=56210 width=201) - Output:["_col0","_col1"],aggregations:["sum(_col7)"],keys:_col4 - Top N Key Operator [TNK_44] (rows=20154874 width=279) - keys:_col4,sort order:+,top n:100 - Select Operator [SEL_23] (rows=20154874 width=279) - Output:["_col4","_col7"] - Filter Operator [FIL_22] (rows=20154874 width=279) - predicate:((_col3) IN ('CA', 'WA', 'GA') or (_col7 > 500) or (substr(_col4, 1, 5)) IN ('85669', '86197', '88274', '83405', '86475', '85392', '85460', '80348', '81792')) - Merge Join Operator [MERGEJOIN_77] (rows=20154874 width=279) - Conds:RS_19._col0=RS_20._col1(Inner),Output:["_col3","_col4","_col7"] + Group By Operator [GBY_24] (rows=43435 width=201) + Output:["_col0","_col1"],aggregations:["sum(_col8)"],keys:_col3 + Top N Key Operator [TNK_44] (rows=20154874 width=205) + keys:_col3,sort order:+,top n:100 + Select Operator [SEL_23] (rows=20154874 width=205) + Output:["_col3","_col8"] + Filter Operator [FIL_22] (rows=20154874 width=205) + predicate:(_col4 or _col5 or _col9) + Merge Join Operator [MERGEJOIN_77] (rows=20154874 width=205) + Conds:RS_19._col0=RS_20._col1(Inner),Output:["_col3","_col4","_col5","_col8","_col9"] <-Reducer 2 [SIMPLE_EDGE] SHUFFLE [RS_19] PartitionCols:_col0 - Merge Join Operator [MERGEJOIN_75] (rows=80000000 width=179) - Conds:RS_80._col1=RS_83._col0(Inner),Output:["_col0","_col3","_col4"] + Merge Join Operator [MERGEJOIN_75] (rows=80000000 width=101) + Conds:RS_80._col1=RS_83._col0(Inner),Output:["_col0","_col3","_col4","_col5"] <-Map 1 [SIMPLE_EDGE] vectorized SHUFFLE [RS_80] PartitionCols:_col1 @@ -101,8 +101,8 @@ Stage-0 <-Map 6 [SIMPLE_EDGE] vectorized SHUFFLE [RS_83] PartitionCols:_col0 - Select Operator [SEL_82] (rows=40000000 width=179) - Output:["_col0","_col1","_col2"] + Select Operator [SEL_82] (rows=40000000 width=101) + Output:["_col0","_col1","_col2","_col3"] Filter Operator [FIL_81] (rows=40000000 width=179) predicate:ca_address_sk is not null TableScan [TS_3] (rows=40000000 width=179) @@ -110,12 +110,12 @@ Stage-0 <-Reducer 8 [SIMPLE_EDGE] SHUFFLE [RS_20] PartitionCols:_col1 - Merge Join Operator [MERGEJOIN_76] (rows=20154874 width=107) - Conds:RS_94._col0=RS_86._col0(Inner),Output:["_col1","_col2"] + Merge Join Operator [MERGEJOIN_76] (rows=20154874 width=111) + Conds:RS_94._col0=RS_86._col0(Inner),Output:["_col1","_col2","_col3"] <-Map 9 [SIMPLE_EDGE] vectorized SHUFFLE [RS_86] PartitionCols:_col0 - Select Operator [SEL_85] (rows=130 width=12) + Select Operator [SEL_85] (rows=130 width=4) Output:["_col0"] Filter Operator [FIL_84] (rows=130 width=12) predicate:((d_qoy = 2) and (d_year = 2000) and d_date_sk is not null) @@ -124,8 +124,8 @@ Stage-0 <-Map 7 [SIMPLE_EDGE] vectorized SHUFFLE [RS_94] PartitionCols:_col0 - Select Operator [SEL_93] (rows=285117831 width=119) - Output:["_col0","_col1","_col2"] + Select Operator [SEL_93] (rows=285117831 width=123) + Output:["_col0","_col1","_col2","_col3"] Filter Operator [FIL_92] (rows=285117831 width=119) predicate:((cs_sold_date_sk BETWEEN DynamicValue(RS_13_date_dim_d_date_sk_min) AND DynamicValue(RS_13_date_dim_d_date_sk_max) and in_bloom_filter(cs_sold_date_sk, DynamicValue(RS_13_date_dim_d_date_sk_bloom_filter))) and cs_bill_customer_sk is not null and cs_sold_date_sk is not null) TableScan [TS_6] (rows=287989836 width=119)