hql语句mysql sysdate_HQL中count(*),count(1),count(column)的区别

1、执行计划：

count(*)

explain select count(*) from dpods_SMS_Queue limit 100;

Explain

ABSTRACT SYNTAX TREE:

(TOK_QUERY (TOK_FROM (TOK_TABREF (TOK_TABNAME dpods_SMS_Queue))) (TOK_INSERT (TOK_DESTINATION (TOK_DIR TOK_TMP_FILE)) (TOK_SELECT (TOK_SELEXPR (TOK_FUNCTIONSTAR count))) (TOK_LIMIT 100)))

STAGE DEPENDENCIES:

Stage-1 is a root stage

Stage-0 is a root stage

STAGE PLANS:

Stage: Stage-1

Map Reduce

Alias -> Map Operator Tree:

dpods_sms_queue

TableScan

alias: dpods_sms_queue

Select Operator

Group By Operator

aggregations:

expr: count()    // 此处不同

bucketGroup: false

mode: hash

outputColumnNames: _col0

Reduce Output Operator

sort order:

tag: -1

value expressions:

expr: _col0

type: bigint

Reduce Operator Tree:

Group By Operator

aggregations:

expr: count(VALUE._col0)

bucketGroup: false

mode: mergepartial

outputColumnNames: _col0

Select Operator

expressions:

expr: _col0

type: bigint

outputColumnNames: _col0

Limit

File Output Operator

compressed: false

GlobalTableId: 0

table:

input format: org.apache.hadoop.mapred.TextInputFormat

output format: org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat

Stage: Stage-0

Fetch Operator

limit: 100

count(1)

explain select count(1) from dpods_SMS_Queue limit 100;

Explain

ABSTRACT SYNTAX TREE:

(TOK_QUERY (TOK_FROM (TOK_TABREF (TOK_TABNAME dpods_SMS_Queue))) (TOK_INSERT (TOK_DESTINATION (TOK_DIR TOK_TMP_FILE)) (TOK_SELECT (TOK_SELEXPR (TOK_FUNCTION count 1))) (TOK_LIMIT 100)))

STAGE DEPENDENCIES:

Stage-1 is a root stage

Stage-0 is a root stage

STAGE PLANS:

Stage: Stage-1

Map Reduce

Alias -> Map Operator Tree:

dpods_sms_queue

TableScan

alias: dpods_sms_queue

Select Operator

Group By Operator

aggregations:

expr: count(1)    // 此处不同

bucketGroup: false

mode: hash

outputColumnNames: _col0

Reduce Output Operator

sort order:

tag: -1

value expressions:

expr: _col0

type: bigint

Reduce Operator Tree:

Group By Operator

aggregations:

expr: count(VALUE._col0)

bucketGroup: false

mode: mergepartial

outputColumnNames: _col0

Select Operator

expressions:

expr: _col0

type: bigint

outputColumnNames: _col0

Limit

File Output Operator

compressed: false

GlobalTableId: 0

table:

input format: org.apache.hadoop.mapred.TextInputFormat

output format: org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat

Stage: Stage-0

Fetch Operator

limit: 100

count(column)

explain select count(msgid) from dpods_SMS_Queue limit 100;

Explain

ABSTRACT SYNTAX TREE:

(TOK_QUERY (TOK_FROM (TOK_TABREF (TOK_TABNAME dpods_SMS_Queue))) (TOK_INSERT (TOK_DESTINATION (TOK_DIR TOK_TMP_FILE)) (TOK_SELECT (TOK_SELEXPR (TOK_FUNCTION count (TOK_TABLE_OR_COL msgid)))) (TOK_LIMIT 100)))

STAGE DEPENDENCIES:

Stage-1 is a root stage

Stage-0 is a root stage

STAGE PLANS:

Stage: Stage-1

Map Reduce

Alias -> Map Operator Tree:

dpods_sms_queue

TableScan

alias: dpods_sms_queue

Select Operator    // 此处多了一个Select Operator

expressions:

expr: msgid

type: int

outputColumnNames: msgid

Group By Operator

aggregations:

expr: count(msgid)    // 此处不同

bucketGroup: false

mode: hash

outputColumnNames: _col0

Reduce Output Operator

sort order:

tag: -1

value expressions:

expr: _col0

type: bigint

Reduce Operator Tree:

Group By Operator

aggregations:

expr: count(VALUE._col0)

bucketGroup: false

mode: mergepartial

outputColumnNames: _col0

Select Operator

expressions:

expr: _col0

type: bigint

outputColumnNames: _col0

Limit

File Output Operator

compressed: false

GlobalTableId: 0

table:

input format: org.apache.hadoop.mapred.TextInputFormat

output format: org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat

Stage: Stage-0

Fetch Operator

limit: 100

从上面三个执行计划来看，只有两处不同：

count(column)的时候多了一个Select operator；

Group By Operator 的 aggregations过程略有不同；

2、结果：

count(*)：所有行进行统计，包括NULL行

count(1)：所有行进行统计，包括NULL行

count(column)：对column中非Null进行统计

count(distinct column)：对column中非Null进行去重统计