인덱스종류
FP : Fast Projection
-> create table YY_백승희_TEST1 ( AAA char(13) UNIQUE );
-> 특정칼럼에 UNIQUE 옵션줄때 자동생성됨
-> Distinct 값이 256 이하이면 1Byte FP인덱스생성됨
-> Distinct 값이 256 이상이면 2Byte FP인덱스생성됨(1Byte보다 Perfomance좋음)
LF : Low Fast (Low Cadinality Data)
-> Unique Value가 1500 미만일때 적당
-> Where 절의 조건칼럼, Min, Max함수사용칼럼, Group By 구문에 사용시 적절
-> Column Example : Male/Female, Yes/No, Product Name, State or Province,
Date (Fewer Than 1500 Distinct Value)
-> Allow 10000 Distinct value But Recommended 1500 Distinct Value CufOff.
HNG : High NonGroup
-> Bit-Wise Index Optimized for Range Searches and Aggregate Functions
-> Float, Real, Double DataType Not Supported
-> Ranges, Between
-> Sum() and AVG() function 에 적당
-> Example : Date, Amount, Quantity
HG : High Group
-> Used for High Cadinality Data
-> Enhanced B-Tree Index to Process Equality and GROUP BY Operations on
High-Cardinality Data.
-> Needs for Columns with a high Number of unique values ( > 1500 )
-> Used with : - Search arguments in WHERE clause
- JOINS(Regardless of Cardinality)
- GROUP BY
- ORDER BY
-> HG Index created as UNIQUE or as a PRIMARY KEY will not have a G-Array Structure
and will use less space
-> Example : Airport Code, Product Code, Emp_ID
CMP : Compare
WD : Word