jprdonnelly/impala
1
0
Fork 0
mirror of https://github.com/apache/impala.git synced 2026-01-02 03:00:32 -05:00
Code Issues Projects Releases Wiki Activity
Files
7767d300a3f018c8c8b32fa72abe5c126900a2be
impala/testdata/workloads/functional-query/queries/QueryTest/compute-stats.test
Alex Behm a41710a0c8 Use unique_database fixture in test_compute_stats.py. 
This patch makes it a little easier to use the unique_database fixture
with .test files. The RESULTS section can now contain $DATABASE which
is replaced with the current database by the test framework.

Testing:
- ran the test locally on exhaustive
- ran the test on hdfs and the local filesystem on Jenkins

Change-Id: I8655eb769003f88c0e1ec1b254118e4ec3353b48
Reviewed-on: http://gerrit.cloudera.org:8080/2947
Reviewed-by: Alex Behm <alex.behm@cloudera.com>
Tested-by: Internal Jenkins
2016-05-12 14:17:50 -07:00

907 lines
29 KiB
Plaintext
Raw Blame History

====
---- QUERY
# test computing stats on a partitioned text table with all types
create table alltypes like functional.alltypes;
insert into alltypes partition(year, month)
select * from functional.alltypes;
====
---- QUERY
compute stats alltypes
---- RESULTS
'Updated 24 partition(s) and 11 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats alltypes
---- LABELS
YEAR, MONTH, #ROWS, #FILES, SIZE, BYTES CACHED, CACHE REPLICATION, FORMAT, INCREMENTAL STATS, LOCATION
---- RESULTS
'2009','1',310,1,'24.56KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','2',280,1,'22.27KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','3',310,1,'24.67KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','4',300,1,'24.06KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','5',310,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','6',300,1,'24.16KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','7',310,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','8',310,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','9',300,1,'24.16KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','10',310,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','11',300,1,'24.16KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','12',310,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','1',310,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','2',280,1,'22.54KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','3',310,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','4',300,1,'24.16KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','5',310,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','6',300,1,'24.16KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','7',310,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','8',310,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','9',300,1,'24.16KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','10',310,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','11',300,1,'24.16KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','12',310,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'Total','',7300,24,'586.84KB','0B','','','',''
---- TYPES
STRING, STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
show column stats alltypes
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',7505,-1,4,4
'bool_col','BOOLEAN',2,-1,1,1
'tinyint_col','TINYINT',10,-1,1,1
'smallint_col','SMALLINT',10,-1,2,2
'int_col','INT',10,-1,4,4
'bigint_col','BIGINT',10,-1,8,8
'float_col','FLOAT',10,-1,4,4
'double_col','DOUBLE',10,-1,8,8
'date_string_col','STRING',736,-1,8,8
'string_col','STRING',10,-1,1,1
'timestamp_col','TIMESTAMP',7554,-1,16,16
'year','INT',2,0,4,4
'month','INT',12,0,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# drop stats from this table
drop stats alltypes
====
---- QUERY
show table stats alltypes
---- LABELS
YEAR, MONTH, #ROWS, #FILES, SIZE, BYTES CACHED, CACHE REPLICATION, FORMAT, INCREMENTAL STATS, LOCATION
---- RESULTS
'2009','1',-1,1,'24.56KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','10',-1,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','11',-1,1,'24.16KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','12',-1,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','2',-1,1,'22.27KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','3',-1,1,'24.67KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','4',-1,1,'24.06KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','5',-1,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','6',-1,1,'24.16KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','7',-1,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','8',-1,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2009','9',-1,1,'24.16KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','1',-1,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','10',-1,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','11',-1,1,'24.16KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','12',-1,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','2',-1,1,'22.54KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','3',-1,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','4',-1,1,'24.16KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','5',-1,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','6',-1,1,'24.16KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','7',-1,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','8',-1,1,'24.97KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2010','9',-1,1,'24.16KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'Total','',-1,24,'586.84KB','0B','','','',''
---- TYPES
STRING, STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
# Note - the NDV for partition columns is read from the table metadata.
show column stats alltypes
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',-1,-1,4,4
'bool_col','BOOLEAN',-1,-1,1,1
'tinyint_col','TINYINT',-1,-1,1,1
'smallint_col','SMALLINT',-1,-1,2,2
'int_col','INT',-1,-1,4,4
'bigint_col','BIGINT',-1,-1,8,8
'float_col','FLOAT',-1,-1,4,4
'double_col','DOUBLE',-1,-1,8,8
'date_string_col','STRING',-1,-1,-1,-1
'string_col','STRING',-1,-1,-1,-1
'timestamp_col','TIMESTAMP',-1,-1,16,16
'year','INT',2,0,4,4
'month','INT',12,0,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# Add partitions with NULL values and check for stats.
alter table alltypes add partition (year=NULL, month=NULL)
---- RESULTS
====
---- QUERY
show column stats alltypes
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',-1,-1,4,4
'bool_col','BOOLEAN',-1,-1,1,1
'tinyint_col','TINYINT',-1,-1,1,1
'smallint_col','SMALLINT',-1,-1,2,2
'int_col','INT',-1,-1,4,4
'bigint_col','BIGINT',-1,-1,8,8
'float_col','FLOAT',-1,-1,4,4
'double_col','DOUBLE',-1,-1,8,8
'date_string_col','STRING',-1,-1,-1,-1
'string_col','STRING',-1,-1,-1,-1
'timestamp_col','TIMESTAMP',-1,-1,16,16
'year','INT',3,1,4,4
'month','INT',13,1,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
alter table alltypes add partition (year=2011, month=NULL)
---- RESULTS
====
---- QUERY
show column stats alltypes
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',-1,-1,4,4
'bool_col','BOOLEAN',-1,-1,1,1
'tinyint_col','TINYINT',-1,-1,1,1
'smallint_col','SMALLINT',-1,-1,2,2
'int_col','INT',-1,-1,4,4
'bigint_col','BIGINT',-1,-1,8,8
'float_col','FLOAT',-1,-1,4,4
'double_col','DOUBLE',-1,-1,8,8
'date_string_col','STRING',-1,-1,-1,-1
'string_col','STRING',-1,-1,-1,-1
'timestamp_col','TIMESTAMP',-1,-1,16,16
'year','INT',4,1,4,4
'month','INT',13,2,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# Drop the partitions with NULL values and check for stats.
alter table alltypes drop partition (year=NULL, month=NULL)
---- RESULTS
====
---- QUERY
show column stats alltypes
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',-1,-1,4,4
'bool_col','BOOLEAN',-1,-1,1,1
'tinyint_col','TINYINT',-1,-1,1,1
'smallint_col','SMALLINT',-1,-1,2,2
'int_col','INT',-1,-1,4,4
'bigint_col','BIGINT',-1,-1,8,8
'float_col','FLOAT',-1,-1,4,4
'double_col','DOUBLE',-1,-1,8,8
'date_string_col','STRING',-1,-1,-1,-1
'string_col','STRING',-1,-1,-1,-1
'timestamp_col','TIMESTAMP',-1,-1,16,16
'year','INT',3,0,4,4
'month','INT',13,1,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
alter table alltypes drop partition (year=2011, month=NULL)
---- RESULTS
====
---- QUERY
show column stats alltypes
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',-1,-1,4,4
'bool_col','BOOLEAN',-1,-1,1,1
'tinyint_col','TINYINT',-1,-1,1,1
'smallint_col','SMALLINT',-1,-1,2,2
'int_col','INT',-1,-1,4,4
'bigint_col','BIGINT',-1,-1,8,8
'float_col','FLOAT',-1,-1,4,4
'double_col','DOUBLE',-1,-1,8,8
'date_string_col','STRING',-1,-1,-1,-1
'string_col','STRING',-1,-1,-1,-1
'timestamp_col','TIMESTAMP',-1,-1,16,16
'year','INT',2,0,4,4
'month','INT',12,0,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# drop stats from this table a second time, should not throw an error.
drop stats alltypes
====
---- QUERY
# test computing stats on an partitioned text table with all types
create table alltypesnopart like functional.alltypesnopart;
insert into alltypesnopart
select id, bool_col, tinyint_col, smallint_col, int_col, bigint_col, float_col,
double_col, date_string_col, string_col, timestamp_col
from functional.alltypessmall;
====
---- QUERY
compute stats alltypesnopart
---- RESULTS
'Updated 1 partition(s) and 11 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats alltypesnopart
---- LABELS
#ROWS, #FILES, SIZE, BYTES CACHED, CACHE REPLICATION, FORMAT, INCREMENTAL STATS, LOCATION
---- RESULTS
100,3,'7.73KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
---- TYPES
BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
show column stats alltypesnopart
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',99,-1,4,4
'bool_col','BOOLEAN',2,-1,1,1
'tinyint_col','TINYINT',10,-1,1,1
'smallint_col','SMALLINT',10,-1,2,2
'int_col','INT',10,-1,4,4
'bigint_col','BIGINT',10,-1,8,8
'float_col','FLOAT',10,-1,4,4
'double_col','DOUBLE',10,-1,8,8
'date_string_col','STRING',12,-1,8,8
'string_col','STRING',10,-1,1,1
'timestamp_col','TIMESTAMP',101,-1,16,16
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# test computing stats on a partitioned parquet table with all types
create table alltypes_parquet
like functional_parquet.alltypes;
insert into alltypes_parquet partition(year, month)
select * from functional.alltypes;
====
---- QUERY
compute stats alltypes_parquet
---- RESULTS
'Updated 24 partition(s) and 11 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats alltypes_parquet
---- LABELS
YEAR, MONTH, #ROWS, #FILES, SIZE, BYTES CACHED, CACHE REPLICATION, FORMAT, INCREMENTAL STATS, LOCATION
---- RESULTS
'2009','1',310,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2009','2',280,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2009','3',310,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2009','4',300,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2009','5',310,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2009','6',300,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2009','7',310,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2009','8',310,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2009','9',300,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2009','10',310,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2009','11',300,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2009','12',310,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2010','1',310,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2010','2',280,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2010','3',310,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2010','4',300,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2010','5',310,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2010','6',300,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2010','7',310,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2010','8',310,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2010','9',300,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2010','10',310,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2010','11',300,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'2010','12',310,1,regex:.+KB,'NOT CACHED','NOT CACHED','PARQUET','false',regex:.*
'Total','',7300,24,regex:.+KB,'0B','','','',''
---- TYPES
STRING, STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
show column stats alltypes_parquet
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',7505,-1,4,4
'bool_col','BOOLEAN',2,-1,1,1
'tinyint_col','TINYINT',10,-1,1,1
'smallint_col','SMALLINT',10,-1,2,2
'int_col','INT',10,-1,4,4
'bigint_col','BIGINT',10,-1,8,8
'float_col','FLOAT',10,-1,4,4
'double_col','DOUBLE',10,-1,8,8
'date_string_col','STRING',736,-1,8,8
'string_col','STRING',10,-1,1,1
'timestamp_col','TIMESTAMP',7554,-1,16,16
'year','INT',2,0,4,4
'month','INT',12,0,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# test computing stats on an empty table
create table alltypes_empty like functional_rc_snap.alltypes
====
---- QUERY
compute stats alltypes_empty
---- RESULTS
'Updated 0 partition(s) and 11 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats alltypes_empty
---- LABELS
YEAR, MONTH, #ROWS, #FILES, SIZE, BYTES CACHED, CACHE REPLICATION, FORMAT, INCREMENTAL STATS, LOCATION
---- RESULTS
'Total','',0,0,'0B','0B','','','',''
---- TYPES
STRING, STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
show column stats alltypes_empty
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',0,-1,4,4
'bool_col','BOOLEAN',2,-1,1,1
'tinyint_col','TINYINT',0,-1,1,1
'smallint_col','SMALLINT',0,-1,2,2
'int_col','INT',0,-1,4,4
'bigint_col','BIGINT',0,-1,8,8
'float_col','FLOAT',0,-1,4,4
'double_col','DOUBLE',0,-1,8,8
'date_string_col','STRING',0,-1,0,0
'string_col','STRING',0,-1,0,0
'timestamp_col','TIMESTAMP',0,-1,16,16
'year','INT',0,0,4,4
'month','INT',0,0,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# IMPALA-867: Test computing stats on Avro tables created by Hive with
# matching/mismatched column definitions and Avro schema.
# Clone the used tables here.
create table avro_hive_alltypes
like functional_avro_snap.alltypes;
create table avro_hive_alltypes_extra_coldef
like functional_avro_snap.alltypes_extra_coldef;
create table avro_hive_alltypes_missing_coldef
like functional_avro_snap.alltypes_missing_coldef;
create table avro_hive_alltypes_type_mismatch
like functional_avro_snap.alltypes_type_mismatch;
create table avro_hive_no_avro_schema
like functional_avro_snap.no_avro_schema;
====
---- QUERY
# Avro table with matching column definitions and Avro schema
compute stats avro_hive_alltypes
---- RESULTS
'Updated 0 partition(s) and 11 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats avro_hive_alltypes
---- LABELS
YEAR, MONTH, #ROWS, #FILES, SIZE, BYTES CACHED, CACHE REPLICATION, FORMAT, INCREMENTAL STATS, LOCATION
---- RESULTS
'Total','',0,0,'0B','0B','','','',''
---- TYPES
STRING, STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
show column stats avro_hive_alltypes
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',0,-1,4,4
'bool_col','BOOLEAN',2,-1,1,1
'tinyint_col','INT',0,-1,4,4
'smallint_col','INT',0,-1,4,4
'int_col','INT',0,-1,4,4
'bigint_col','BIGINT',0,-1,8,8
'float_col','FLOAT',0,-1,4,4
'double_col','DOUBLE',0,-1,8,8
'date_string_col','STRING',0,-1,0,0
'string_col','STRING',0,-1,0,0
'timestamp_col','STRING',0,-1,0,0
'year','INT',0,0,4,4
'month','INT',0,0,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# Avro table with an extra column definition.
compute stats avro_hive_alltypes_extra_coldef
---- RESULTS
'Updated 0 partition(s) and 12 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats avro_hive_alltypes_extra_coldef
---- LABELS
YEAR, MONTH, #ROWS, #FILES, SIZE, BYTES CACHED, CACHE REPLICATION, FORMAT, INCREMENTAL STATS, LOCATION
---- RESULTS
'Total','',0,0,'0B','0B','','','',''
---- TYPES
STRING, STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
show column stats avro_hive_alltypes_extra_coldef
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',0,-1,4,4
'bool_col','BOOLEAN',2,-1,1,1
'tinyint_col','TINYINT',0,-1,1,1
'smallint_col','SMALLINT',0,-1,2,2
'int_col','INT',0,-1,4,4
'bigint_col','BIGINT',0,-1,8,8
'float_col','FLOAT',0,-1,4,4
'double_col','DOUBLE',0,-1,8,8
'date_string_col','STRING',0,-1,0,0
'string_col','STRING',0,-1,0,0
'timestamp_col','TIMESTAMP',0,-1,16,16
'extra_col','STRING',0,-1,0,0
'year','INT',0,0,4,4
'month','INT',0,0,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# Avro table with missing two column definitions.
compute stats avro_hive_alltypes_missing_coldef
---- RESULTS
'Updated 0 partition(s) and 9 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats avro_hive_alltypes_missing_coldef
---- LABELS
YEAR, MONTH, #ROWS, #FILES, SIZE, BYTES CACHED, CACHE REPLICATION, FORMAT, INCREMENTAL STATS, LOCATION
---- RESULTS
'Total','',0,0,'0B','0B','','','',''
---- TYPES
STRING, STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
show column stats avro_hive_alltypes_missing_coldef
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',0,-1,4,4
'bool_col','BOOLEAN',2,-1,1,1
'smallint_col','SMALLINT',0,-1,2,2
'int_col','INT',0,-1,4,4
'bigint_col','BIGINT',0,-1,8,8
'float_col','FLOAT',0,-1,4,4
'double_col','DOUBLE',0,-1,8,8
'date_string_col','STRING',0,-1,0,0
'string_col','STRING',0,-1,0,0
'year','INT',0,0,4,4
'month','INT',0,0,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# Avro table with one column definition having a different
# type than the Avro schema (bigint_col is a string).
compute stats avro_hive_alltypes_type_mismatch
---- RESULTS
'Updated 0 partition(s) and 11 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats avro_hive_alltypes_type_mismatch
---- LABELS
YEAR, MONTH, #ROWS, #FILES, SIZE, BYTES CACHED, CACHE REPLICATION, FORMAT, INCREMENTAL STATS, LOCATION
---- RESULTS
'Total','',0,0,'0B','0B','','','',''
---- TYPES
STRING, STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
show column stats avro_hive_alltypes_type_mismatch
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',0,-1,4,4
'bool_col','BOOLEAN',2,-1,1,1
'tinyint_col','TINYINT',0,-1,1,1
'smallint_col','SMALLINT',0,-1,2,2
'int_col','INT',0,-1,4,4
'bigint_col','STRING',0,-1,0,0
'float_col','FLOAT',0,-1,4,4
'double_col','DOUBLE',0,-1,8,8
'date_string_col','STRING',0,-1,0,0
'string_col','STRING',0,-1,0,0
'timestamp_col','TIMESTAMP',0,-1,16,16
'year','INT',0,0,4,4
'month','INT',0,0,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# Avro table without an Avro schema created by Hive.
# The Avro schema is inferred from the column definitions,
compute stats avro_hive_no_avro_schema
---- RESULTS
'Updated 0 partition(s) and 11 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats avro_hive_no_avro_schema
---- LABELS
YEAR, MONTH, #ROWS, #FILES, SIZE, BYTES CACHED, CACHE REPLICATION, FORMAT, INCREMENTAL STATS, LOCATION
---- RESULTS
'Total','',0,0,'0B','0B','','','',''
---- TYPES
STRING, STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
show column stats avro_hive_no_avro_schema
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',0,-1,4,4
'bool_col','BOOLEAN',2,-1,1,1
'tinyint_col','INT',0,-1,4,4
'smallint_col','INT',0,-1,4,4
'int_col','INT',0,-1,4,4
'bigint_col','BIGINT',0,-1,8,8
'float_col','FLOAT',0,-1,4,4
'double_col','DOUBLE',0,-1,8,8
'date_string_col','STRING',0,-1,0,0
'string_col','STRING',0,-1,0,0
'timestamp_col','STRING',0,-1,0,0
'year','INT',0,0,4,4
'month','INT',0,0,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# Test Avro table created without any column definitions.
create table avro_impala_alltypes_no_coldefs
partitioned by (year int, month int)
with serdeproperties
('avro.schema.url'='$FILESYSTEM_PREFIX/test-warehouse/avro_schemas/functional/alltypes.json')
stored as avro;
====
---- QUERY
compute stats avro_impala_alltypes_no_coldefs
---- RESULTS
'Updated 0 partition(s) and 11 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats avro_impala_alltypes_no_coldefs
---- LABELS
YEAR, MONTH, #ROWS, #FILES, SIZE, BYTES CACHED, CACHE REPLICATION, FORMAT, INCREMENTAL STATS, LOCATION
---- RESULTS
'Total','',0,0,'0B','0B','','','',''
---- TYPES
STRING, STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
show column stats avro_impala_alltypes_no_coldefs
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',0,-1,4,4
'bool_col','BOOLEAN',2,-1,1,1
'tinyint_col','INT',0,-1,4,4
'smallint_col','INT',0,-1,4,4
'int_col','INT',0,-1,4,4
'bigint_col','BIGINT',0,-1,8,8
'float_col','FLOAT',0,-1,4,4
'double_col','DOUBLE',0,-1,8,8
'date_string_col','STRING',0,-1,0,0
'string_col','STRING',0,-1,0,0
'timestamp_col','STRING',0,-1,0,0
'year','INT',0,0,4,4
'month','INT',0,0,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# IMPALA-1104: Test computing stats on Avro tables created by Impala
# with mismatched column definitions and Avro schema. Mismatched column name.
create table avro_impala_alltypes_bad_colname
(id int, bool_col boolean, tinyint_col int, smallint_col int, bad_int_col int,
bigint_col bigint, float_col float, double_col double, date_string_col string,
string_col string, timestamp_col timestamp)
partitioned by (year int, month int)
with serdeproperties
('avro.schema.url'='$FILESYSTEM_PREFIX/test-warehouse/avro_schemas/functional/alltypes.json')
stored as avro;
====
---- QUERY
compute stats avro_impala_alltypes_bad_colname
---- RESULTS
'Updated 0 partition(s) and 11 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats avro_impala_alltypes_bad_colname
---- LABELS
YEAR, MONTH, #ROWS, #FILES, SIZE, BYTES CACHED, CACHE REPLICATION, FORMAT, INCREMENTAL STATS, LOCATION
---- RESULTS
'Total','',0,0,'0B','0B','','','',''
---- TYPES
STRING, STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
show column stats avro_impala_alltypes_bad_colname
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',0,-1,4,4
'bool_col','BOOLEAN',2,-1,1,1
'tinyint_col','INT',0,-1,4,4
'smallint_col','INT',0,-1,4,4
'int_col','INT',0,-1,4,4
'bigint_col','BIGINT',0,-1,8,8
'float_col','FLOAT',0,-1,4,4
'double_col','DOUBLE',0,-1,8,8
'date_string_col','STRING',0,-1,0,0
'string_col','STRING',0,-1,0,0
'timestamp_col','STRING',0,-1,0,0
'year','INT',0,0,4,4
'month','INT',0,0,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# IMPALA-1104: Test computing stats on Avro tables created by Impala
# with mismatched column definitions and Avro schema. Mismatched column type.
create table avro_impala_alltypes_bad_coltype
(id int, bool_col boolean, tinyint_col int, smallint_col int, int_col int,
bigint_col bigint, float_col float, double_col bigint, date_string_col string,
string_col string, timestamp_col timestamp)
partitioned by (year int, month int)
with serdeproperties
('avro.schema.url'='$FILESYSTEM_PREFIX/test-warehouse/avro_schemas/functional/alltypes.json')
stored as avro;
====
---- QUERY
compute stats avro_impala_alltypes_bad_coltype
---- RESULTS
'Updated 0 partition(s) and 11 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats avro_impala_alltypes_bad_coltype
---- LABELS
YEAR, MONTH, #ROWS, #FILES, SIZE, BYTES CACHED, CACHE REPLICATION, FORMAT, INCREMENTAL STATS, LOCATION
---- RESULTS
'Total','',0,0,'0B','0B','','','',''
---- TYPES
STRING, STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
show column stats avro_impala_alltypes_bad_coltype
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',0,-1,4,4
'bool_col','BOOLEAN',2,-1,1,1
'tinyint_col','INT',0,-1,4,4
'smallint_col','INT',0,-1,4,4
'int_col','INT',0,-1,4,4
'bigint_col','BIGINT',0,-1,8,8
'float_col','FLOAT',0,-1,4,4
'double_col','DOUBLE',0,-1,8,8
'date_string_col','STRING',0,-1,0,0
'string_col','STRING',0,-1,0,0
'timestamp_col','STRING',0,-1,0,0
'year','INT',0,0,4,4
'month','INT',0,0,4,4
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# IMPALA-883: Compute table stats for an empty partition.
create table empty_partitioned (i int) partitioned by (j int);
alter table empty_partitioned add partition (j=1);
====
---- QUERY
compute stats empty_partitioned
---- RESULTS
'Updated 1 partition(s) and 1 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats empty_partitioned
---- RESULTS
'1',0,0,'0B','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'Total',0,0,'0B','0B','','','',''
---- TYPES
STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
# Insert non empty partition to the table with empty partition.
insert into table empty_partitioned partition (j=2) select 1;
====
---- QUERY
# Verify partition stats work with empty and non-empty partition.
drop stats empty_partitioned;
compute stats empty_partitioned;
---- RESULTS
'Updated 2 partition(s) and 1 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats empty_partitioned
---- RESULTS
'1',0,0,'0B','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2',1,1,'2B','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'Total',1,1,'2B','0B','','','',''
---- TYPES
STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
# Verify partition stats work with empty and non-empty partition.
drop stats empty_partitioned;
compute stats empty_partitioned;
---- RESULTS
'Updated 2 partition(s) and 1 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats empty_partitioned
---- RESULTS
'1',0,0,'0B','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'2',1,1,'2B','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'Total',1,1,'2B','0B','','','',''
---- TYPES
STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
---- QUERY
# IMPALA-1614 Verify that COMPUTE STATS works on a table whose name starts with numbers.
create table `123_table` (i int, 1p int) partitioned by (2j int);
alter table `123_table` add partition (2j=1);
====
---- QUERY
compute stats `123_table`
---- RESULTS
'Updated 1 partition(s) and 2 column(s).'
---- TYPES
STRING
====
---- QUERY
show table stats `123_table`
---- RESULTS
'1',0,0,'0B','NOT CACHED','NOT CACHED','TEXT','false',regex:.*
'Total',0,0,'0B','0B','','','',''
---- TYPES
STRING, BIGINT, BIGINT, STRING, STRING, STRING, STRING, STRING, STRING
====
====
---- QUERY
# IMPALA-1629: Verify that the column stats for CHAR/VARCHAR columns are populated.
# The values of date_string_col always have exactly 8 characters. The CHAR/VARCHAR
# sizes below are chosen such that they are smaller, equal, and greater than the
# source data values, in particular, to test the CHAR padding behavior.
create table chars_tbl (
id int,
ch1 char(1),
ch2 char(8),
ch3 char(20),
ts timestamp,
vc1 varchar(1),
vc2 varchar(8),
vc3 varchar(20)
)
partitioned by (
year char(5),
day varchar(13)
);
insert overwrite chars_tbl partition(year, day)
select
id,
cast(date_string_col as char(1)),
cast(date_string_col as char(8)),
cast(date_string_col as char(20)),
timestamp_col,
cast(date_string_col as varchar(1)),
cast(date_string_col as varchar(8)),
cast(date_string_col as varchar(20)),
cast(year as char(5)),
cast(day as varchar(13))
from functional.alltypesagg
where day is null or day in (3, 7);
---- RESULTS: VERIFY_IS_EQUAL_SORTED
year=2010 /day=7/: 1000
year=2010 /day=3/: 1000
year=2010 /day=__HIVE_DEFAULT_PARTITION__/: 1000
====
---- QUERY
compute stats chars_tbl
---- RESULTS
'Updated 3 partition(s) and 8 column(s).'
---- TYPES
STRING
====
---- QUERY
show column stats chars_tbl
---- LABELS
COLUMN, TYPE, #DISTINCT VALUES, #NULLS, MAX SIZE, AVG SIZE
---- RESULTS
'id','INT',2915,-1,4,4
'ch1','CHAR(1)',1,-1,1,1
'ch2','CHAR(8)',10,-1,8,8
'ch3','CHAR(20)',10,-1,8,8
'ts','TIMESTAMP',2871,-1,16,16
'vc1','VARCHAR(1)',1,-1,1,1
'vc2','VARCHAR(8)',10,-1,8,8
'vc3','VARCHAR(20)',10,-1,8,8
'year','CHAR(5)',1,0,5,5
'day','VARCHAR(13)',3,1,-1,-1
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
---- QUERY
# Test that compute stats on a Hive-created Avro table without column defs
# works (HIVE-6308, IMPALA-867).
create table alltypes_no_coldef like functional_avro_snap.alltypes_no_coldef;
compute stats alltypes_no_coldef
---- RESULTS
'Updated 1 partition(s) and 11 column(s).'
---- TYPES
STRING
====
---- QUERY
show column stats alltypes_no_coldef
---- RESULTS
'id','INT',0,-1,4,4
'bool_col','BOOLEAN',2,-1,1,1
'tinyint_col','INT',0,-1,4,4
'smallint_col','INT',0,-1,4,4
'int_col','INT',0,-1,4,4
'bigint_col','BIGINT',0,-1,8,8
'float_col','FLOAT',0,-1,4,4
'double_col','DOUBLE',0,-1,8,8
'date_string_col','STRING',0,-1,0,0
'string_col','STRING',0,-1,0,0
'timestamp_col','STRING',0,-1,0,0
---- TYPES
STRING, STRING, BIGINT, BIGINT, INT, DOUBLE
====
Reference in New Issue View Git Blame Copy Permalink