PREHOOK: query: -- Based on display_colstats_tbllvl.q, output should be almost exactly the same. DROP TABLE IF EXISTS UserVisits_web_text_none PREHOOK: type: DROPTABLE POSTHOOK: query: -- Based on display_colstats_tbllvl.q, output should be almost exactly the same. DROP TABLE IF EXISTS UserVisits_web_text_none POSTHOOK: type: DROPTABLE PREHOOK: query: -- Hack, set external location because generated filename changes during test runs CREATE TEMPORARY EXTERNAL TABLE UserVisits_web_text_none ( sourceIP string, destURL string, visitDate string, adRevenue float, userAgent string, cCode string, lCode string, sKeyword string, avgTimeOnSite int) row format delimited fields terminated by '|' stored as textfile #### A masked pattern was here #### PREHOOK: type: CREATETABLE #### A masked pattern was here #### PREHOOK: Output: database:default PREHOOK: Output: default@UserVisits_web_text_none POSTHOOK: query: -- Hack, set external location because generated filename changes during test runs CREATE TEMPORARY EXTERNAL TABLE UserVisits_web_text_none ( sourceIP string, destURL string, visitDate string, adRevenue float, userAgent string, cCode string, lCode string, sKeyword string, avgTimeOnSite int) row format delimited fields terminated by '|' stored as textfile #### A masked pattern was here #### POSTHOOK: type: CREATETABLE #### A masked pattern was here #### POSTHOOK: Output: database:default POSTHOOK: Output: default@UserVisits_web_text_none PREHOOK: query: LOAD DATA LOCAL INPATH "../../data/files/UserVisits.dat" INTO TABLE UserVisits_web_text_none PREHOOK: type: LOAD #### A masked pattern was here #### PREHOOK: Output: default@uservisits_web_text_none POSTHOOK: query: LOAD DATA LOCAL INPATH "../../data/files/UserVisits.dat" INTO TABLE UserVisits_web_text_none POSTHOOK: type: LOAD #### A masked pattern was here #### POSTHOOK: Output: default@uservisits_web_text_none PREHOOK: query: desc extended UserVisits_web_text_none sourceIP PREHOOK: type: DESCTABLE PREHOOK: Input: default@uservisits_web_text_none POSTHOOK: query: desc extended UserVisits_web_text_none sourceIP POSTHOOK: type: DESCTABLE POSTHOOK: Input: default@uservisits_web_text_none sourceIP string from deserializer PREHOOK: query: desc formatted UserVisits_web_text_none sourceIP PREHOOK: type: DESCTABLE PREHOOK: Input: default@uservisits_web_text_none POSTHOOK: query: desc formatted UserVisits_web_text_none sourceIP POSTHOOK: type: DESCTABLE POSTHOOK: Input: default@uservisits_web_text_none # col_name data_type min max num_nulls distinct_count avg_col_len max_col_len num_trues num_falses comment sourceIP string from deserializer PREHOOK: query: explain analyze table UserVisits_web_text_none compute statistics for columns sourceIP, avgTimeOnSite, adRevenue PREHOOK: type: QUERY POSTHOOK: query: explain analyze table UserVisits_web_text_none compute statistics for columns sourceIP, avgTimeOnSite, adRevenue POSTHOOK: type: QUERY STAGE DEPENDENCIES: Stage-0 is a root stage Stage-1 is a root stage STAGE PLANS: Stage: Stage-0 Map Reduce Map Operator Tree: TableScan alias: uservisits_web_text_none Select Operator expressions: sourceip (type: string), avgtimeonsite (type: int), adrevenue (type: float) outputColumnNames: sourceip, avgtimeonsite, adrevenue Group By Operator aggregations: compute_stats(sourceip, 16), compute_stats(avgtimeonsite, 16), compute_stats(adrevenue, 16) mode: hash outputColumnNames: _col0, _col1, _col2 Reduce Output Operator sort order: value expressions: _col0 (type: struct), _col1 (type: struct), _col2 (type: struct) Reduce Operator Tree: Group By Operator aggregations: compute_stats(VALUE._col0), compute_stats(VALUE._col1), compute_stats(VALUE._col2) mode: mergepartial outputColumnNames: _col0, _col1, _col2 File Output Operator compressed: false table: input format: org.apache.hadoop.mapred.TextInputFormat output format: org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat serde: org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe Stage: Stage-1 Column Stats Work Column Stats Desc: Columns: sourceIP, avgTimeOnSite, adRevenue Column Types: string, int, float Table: default.uservisits_web_text_none PREHOOK: query: explain extended analyze table UserVisits_web_text_none compute statistics for columns sourceIP, avgTimeOnSite, adRevenue PREHOOK: type: QUERY POSTHOOK: query: explain extended analyze table UserVisits_web_text_none compute statistics for columns sourceIP, avgTimeOnSite, adRevenue POSTHOOK: type: QUERY ABSTRACT SYNTAX TREE: TOK_ANALYZE TOK_TAB TOK_TABNAME UserVisits_web_text_none columns TOK_TABCOLNAME sourceIP avgTimeOnSite adRevenue STAGE DEPENDENCIES: Stage-0 is a root stage Stage-1 is a root stage STAGE PLANS: Stage: Stage-0 Map Reduce Map Operator Tree: TableScan alias: uservisits_web_text_none GatherStats: false Select Operator expressions: sourceip (type: string), avgtimeonsite (type: int), adrevenue (type: float) outputColumnNames: sourceip, avgtimeonsite, adrevenue Group By Operator aggregations: compute_stats(sourceip, 16), compute_stats(avgtimeonsite, 16), compute_stats(adrevenue, 16) mode: hash outputColumnNames: _col0, _col1, _col2 Reduce Output Operator sort order: tag: -1 value expressions: _col0 (type: struct), _col1 (type: struct), _col2 (type: struct) auto parallelism: false Path -> Alias: #### A masked pattern was here #### Path -> Partition: #### A masked pattern was here #### Partition base file name: uservisits_web_text_none input format: org.apache.hadoop.mapred.TextInputFormat output format: org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat properties: COLUMN_STATS_ACCURATE true EXTERNAL TRUE bucket_count -1 columns sourceip,desturl,visitdate,adrevenue,useragent,ccode,lcode,skeyword,avgtimeonsite columns.comments columns.types string:string:string:float:string:string:string:string:int field.delim | #### A masked pattern was here #### name default.uservisits_web_text_none numFiles 1 serialization.ddl struct uservisits_web_text_none { string sourceip, string desturl, string visitdate, float adrevenue, string useragent, string ccode, string lcode, string skeyword, i32 avgtimeonsite} serialization.format | serialization.lib org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe totalSize 7060 serde: org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe input format: org.apache.hadoop.mapred.TextInputFormat output format: org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat properties: COLUMN_STATS_ACCURATE true EXTERNAL TRUE bucket_count -1 columns sourceip,desturl,visitdate,adrevenue,useragent,ccode,lcode,skeyword,avgtimeonsite columns.comments columns.types string:string:string:float:string:string:string:string:int field.delim | #### A masked pattern was here #### name default.uservisits_web_text_none numFiles 1 serialization.ddl struct uservisits_web_text_none { string sourceip, string desturl, string visitdate, float adrevenue, string useragent, string ccode, string lcode, string skeyword, i32 avgtimeonsite} serialization.format | serialization.lib org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe totalSize 7060 serde: org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe name: default.uservisits_web_text_none name: default.uservisits_web_text_none Truncated Path -> Alias: #### A masked pattern was here #### Needs Tagging: false Reduce Operator Tree: Group By Operator aggregations: compute_stats(VALUE._col0), compute_stats(VALUE._col1), compute_stats(VALUE._col2) mode: mergepartial outputColumnNames: _col0, _col1, _col2 File Output Operator compressed: false GlobalTableId: 0 #### A masked pattern was here #### NumFilesPerFileSink: 1 #### A masked pattern was here #### table: input format: org.apache.hadoop.mapred.TextInputFormat output format: org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat properties: columns _col0,_col1,_col2 columns.types struct:struct:struct escape.delim \ hive.serialization.extend.additional.nesting.levels true serialization.format 1 serialization.lib org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe serde: org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe TotalFiles: 1 GatherStats: false MultiFileSpray: false Stage: Stage-1 Column Stats Work Column Stats Desc: Columns: sourceIP, avgTimeOnSite, adRevenue Column Types: string, int, float Table: default.uservisits_web_text_none Is Table Level Stats: true PREHOOK: query: analyze table UserVisits_web_text_none compute statistics for columns sourceIP, avgTimeOnSite, adRevenue PREHOOK: type: QUERY PREHOOK: Input: default@uservisits_web_text_none #### A masked pattern was here #### POSTHOOK: query: analyze table UserVisits_web_text_none compute statistics for columns sourceIP, avgTimeOnSite, adRevenue POSTHOOK: type: QUERY POSTHOOK: Input: default@uservisits_web_text_none #### A masked pattern was here #### PREHOOK: query: desc formatted UserVisits_web_text_none sourceIP PREHOOK: type: DESCTABLE PREHOOK: Input: default@uservisits_web_text_none POSTHOOK: query: desc formatted UserVisits_web_text_none sourceIP POSTHOOK: type: DESCTABLE POSTHOOK: Input: default@uservisits_web_text_none # col_name data_type min max num_nulls distinct_count avg_col_len max_col_len num_trues num_falses comment sourceIP string 0 69 12.763636363636364 13 from deserializer PREHOOK: query: desc formatted UserVisits_web_text_none avgTimeOnSite PREHOOK: type: DESCTABLE PREHOOK: Input: default@uservisits_web_text_none POSTHOOK: query: desc formatted UserVisits_web_text_none avgTimeOnSite POSTHOOK: type: DESCTABLE POSTHOOK: Input: default@uservisits_web_text_none # col_name data_type min max num_nulls distinct_count avg_col_len max_col_len num_trues num_falses comment avgTimeOnSite int 1 9 0 11 from deserializer PREHOOK: query: desc formatted UserVisits_web_text_none adRevenue PREHOOK: type: DESCTABLE PREHOOK: Input: default@uservisits_web_text_none POSTHOOK: query: desc formatted UserVisits_web_text_none adRevenue POSTHOOK: type: DESCTABLE POSTHOOK: Input: default@uservisits_web_text_none # col_name data_type min max num_nulls distinct_count avg_col_len max_col_len num_trues num_falses comment adRevenue float 13.099044799804688 492.98870849609375 0 58 from deserializer PREHOOK: query: CREATE TEMPORARY TABLE empty_tab( a int, b double, c string, d boolean, e binary) row format delimited fields terminated by '|' stored as textfile PREHOOK: type: CREATETABLE PREHOOK: Output: database:default PREHOOK: Output: default@empty_tab POSTHOOK: query: CREATE TEMPORARY TABLE empty_tab( a int, b double, c string, d boolean, e binary) row format delimited fields terminated by '|' stored as textfile POSTHOOK: type: CREATETABLE POSTHOOK: Output: database:default POSTHOOK: Output: default@empty_tab PREHOOK: query: desc formatted empty_tab a PREHOOK: type: DESCTABLE PREHOOK: Input: default@empty_tab POSTHOOK: query: desc formatted empty_tab a POSTHOOK: type: DESCTABLE POSTHOOK: Input: default@empty_tab # col_name data_type min max num_nulls distinct_count avg_col_len max_col_len num_trues num_falses comment a int from deserializer PREHOOK: query: explain analyze table empty_tab compute statistics for columns a,b,c,d,e PREHOOK: type: QUERY POSTHOOK: query: explain analyze table empty_tab compute statistics for columns a,b,c,d,e POSTHOOK: type: QUERY STAGE DEPENDENCIES: Stage-0 is a root stage Stage-1 is a root stage STAGE PLANS: Stage: Stage-0 Map Reduce Map Operator Tree: TableScan alias: empty_tab Select Operator expressions: a (type: int), b (type: double), c (type: string), d (type: boolean), e (type: binary) outputColumnNames: a, b, c, d, e Group By Operator aggregations: compute_stats(a, 16), compute_stats(b, 16), compute_stats(c, 16), compute_stats(d, 16), compute_stats(e, 16) mode: hash outputColumnNames: _col0, _col1, _col2, _col3, _col4 Reduce Output Operator sort order: value expressions: _col0 (type: struct), _col1 (type: struct), _col2 (type: struct), _col3 (type: struct), _col4 (type: struct) Reduce Operator Tree: Group By Operator aggregations: compute_stats(VALUE._col0), compute_stats(VALUE._col1), compute_stats(VALUE._col2), compute_stats(VALUE._col3), compute_stats(VALUE._col4) mode: mergepartial outputColumnNames: _col0, _col1, _col2, _col3, _col4 File Output Operator compressed: false table: input format: org.apache.hadoop.mapred.TextInputFormat output format: org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat serde: org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe Stage: Stage-1 Column Stats Work Column Stats Desc: Columns: a, b, c, d, e Column Types: int, double, string, boolean, binary Table: default.empty_tab PREHOOK: query: analyze table empty_tab compute statistics for columns a,b,c,d,e PREHOOK: type: QUERY PREHOOK: Input: default@empty_tab #### A masked pattern was here #### POSTHOOK: query: analyze table empty_tab compute statistics for columns a,b,c,d,e POSTHOOK: type: QUERY POSTHOOK: Input: default@empty_tab #### A masked pattern was here #### PREHOOK: query: desc formatted empty_tab a PREHOOK: type: DESCTABLE PREHOOK: Input: default@empty_tab POSTHOOK: query: desc formatted empty_tab a POSTHOOK: type: DESCTABLE POSTHOOK: Input: default@empty_tab # col_name data_type min max num_nulls distinct_count avg_col_len max_col_len num_trues num_falses comment a int 0 0 0 0 from deserializer PREHOOK: query: desc formatted empty_tab b PREHOOK: type: DESCTABLE PREHOOK: Input: default@empty_tab POSTHOOK: query: desc formatted empty_tab b POSTHOOK: type: DESCTABLE POSTHOOK: Input: default@empty_tab # col_name data_type min max num_nulls distinct_count avg_col_len max_col_len num_trues num_falses comment b double 0.0 0.0 0 0 from deserializer PREHOOK: query: CREATE DATABASE test PREHOOK: type: CREATEDATABASE PREHOOK: Output: database:test POSTHOOK: query: CREATE DATABASE test POSTHOOK: type: CREATEDATABASE POSTHOOK: Output: database:test PREHOOK: query: USE test PREHOOK: type: SWITCHDATABASE PREHOOK: Input: database:test POSTHOOK: query: USE test POSTHOOK: type: SWITCHDATABASE POSTHOOK: Input: database:test PREHOOK: query: CREATE TEMPORARY TABLE UserVisits_web_text_none ( sourceIP string, destURL string, visitDate string, adRevenue float, userAgent string, cCode string, lCode string, sKeyword string, avgTimeOnSite int) row format delimited fields terminated by '|' stored as textfile PREHOOK: type: CREATETABLE PREHOOK: Output: database:test PREHOOK: Output: test@UserVisits_web_text_none POSTHOOK: query: CREATE TEMPORARY TABLE UserVisits_web_text_none ( sourceIP string, destURL string, visitDate string, adRevenue float, userAgent string, cCode string, lCode string, sKeyword string, avgTimeOnSite int) row format delimited fields terminated by '|' stored as textfile POSTHOOK: type: CREATETABLE POSTHOOK: Output: database:test POSTHOOK: Output: test@UserVisits_web_text_none PREHOOK: query: LOAD DATA LOCAL INPATH "../../data/files/UserVisits.dat" INTO TABLE UserVisits_web_text_none PREHOOK: type: LOAD #### A masked pattern was here #### PREHOOK: Output: test@uservisits_web_text_none POSTHOOK: query: LOAD DATA LOCAL INPATH "../../data/files/UserVisits.dat" INTO TABLE UserVisits_web_text_none POSTHOOK: type: LOAD #### A masked pattern was here #### POSTHOOK: Output: test@uservisits_web_text_none PREHOOK: query: desc extended UserVisits_web_text_none sourceIP PREHOOK: type: DESCTABLE PREHOOK: Input: test@uservisits_web_text_none POSTHOOK: query: desc extended UserVisits_web_text_none sourceIP POSTHOOK: type: DESCTABLE POSTHOOK: Input: test@uservisits_web_text_none sourceIP string from deserializer PREHOOK: query: desc extended test.UserVisits_web_text_none sourceIP PREHOOK: type: DESCTABLE PREHOOK: Input: test@uservisits_web_text_none POSTHOOK: query: desc extended test.UserVisits_web_text_none sourceIP POSTHOOK: type: DESCTABLE POSTHOOK: Input: test@uservisits_web_text_none sourceIP string from deserializer PREHOOK: query: desc extended default.UserVisits_web_text_none sourceIP PREHOOK: type: DESCTABLE PREHOOK: Input: default@uservisits_web_text_none POSTHOOK: query: desc extended default.UserVisits_web_text_none sourceIP POSTHOOK: type: DESCTABLE POSTHOOK: Input: default@uservisits_web_text_none sourceIP string from deserializer PREHOOK: query: desc formatted UserVisits_web_text_none sourceIP PREHOOK: type: DESCTABLE PREHOOK: Input: test@uservisits_web_text_none POSTHOOK: query: desc formatted UserVisits_web_text_none sourceIP POSTHOOK: type: DESCTABLE POSTHOOK: Input: test@uservisits_web_text_none # col_name data_type min max num_nulls distinct_count avg_col_len max_col_len num_trues num_falses comment sourceIP string from deserializer PREHOOK: query: desc formatted test.UserVisits_web_text_none sourceIP PREHOOK: type: DESCTABLE PREHOOK: Input: test@uservisits_web_text_none POSTHOOK: query: desc formatted test.UserVisits_web_text_none sourceIP POSTHOOK: type: DESCTABLE POSTHOOK: Input: test@uservisits_web_text_none # col_name data_type min max num_nulls distinct_count avg_col_len max_col_len num_trues num_falses comment sourceIP string from deserializer PREHOOK: query: desc formatted default.UserVisits_web_text_none sourceIP PREHOOK: type: DESCTABLE PREHOOK: Input: default@uservisits_web_text_none POSTHOOK: query: desc formatted default.UserVisits_web_text_none sourceIP POSTHOOK: type: DESCTABLE POSTHOOK: Input: default@uservisits_web_text_none # col_name data_type min max num_nulls distinct_count avg_col_len max_col_len num_trues num_falses comment sourceIP string 0 69 12.763636363636364 13 from deserializer PREHOOK: query: analyze table UserVisits_web_text_none compute statistics for columns sKeyword PREHOOK: type: QUERY PREHOOK: Input: test@uservisits_web_text_none #### A masked pattern was here #### POSTHOOK: query: analyze table UserVisits_web_text_none compute statistics for columns sKeyword POSTHOOK: type: QUERY POSTHOOK: Input: test@uservisits_web_text_none #### A masked pattern was here #### PREHOOK: query: desc extended UserVisits_web_text_none sKeyword PREHOOK: type: DESCTABLE PREHOOK: Input: test@uservisits_web_text_none POSTHOOK: query: desc extended UserVisits_web_text_none sKeyword POSTHOOK: type: DESCTABLE POSTHOOK: Input: test@uservisits_web_text_none sKeyword string from deserializer PREHOOK: query: desc formatted UserVisits_web_text_none sKeyword PREHOOK: type: DESCTABLE PREHOOK: Input: test@uservisits_web_text_none POSTHOOK: query: desc formatted UserVisits_web_text_none sKeyword POSTHOOK: type: DESCTABLE POSTHOOK: Input: test@uservisits_web_text_none # col_name data_type min max num_nulls distinct_count avg_col_len max_col_len num_trues num_falses comment sKeyword string 0 49 7.872727272727273 19 from deserializer PREHOOK: query: desc formatted test.UserVisits_web_text_none sKeyword PREHOOK: type: DESCTABLE PREHOOK: Input: test@uservisits_web_text_none POSTHOOK: query: desc formatted test.UserVisits_web_text_none sKeyword POSTHOOK: type: DESCTABLE POSTHOOK: Input: test@uservisits_web_text_none # col_name data_type min max num_nulls distinct_count avg_col_len max_col_len num_trues num_falses comment sKeyword string 0 49 7.872727272727273 19 from deserializer