Understand compatibility of DB2 database partition and apposition concept correc

Seeing the thing of the DPF respect of a few DB2 recently, two concepts have a few to do do not understand.

It is ” of “ partition compatibility and “ apposition ” two concepts.

Searched a few relevant data, content is as follows:

Divisional compatibility

The basic data type of the corresponding row that can bolt to partition undertakes comparative, can state them to be is divisional and compatible (Partition Compatible) . Divisional and compatible data type has following property: Have same value but two variable meetings that have different kind press identical divisional and algorithmic map to come same a divisional date. Divisional compatibility has following feature:

Basic data type and another same basic data kind are compatible.

In-house format is used at DATE, TIME and TIMESTAMP data type. They each other are incompatible, and as not compatible as CHAR.

Divisional compatibility does not suffer the effect that contains NOT NULL or the kind that FOR BIT DATA defines.

The processing that is worth to the NULL of compatible data type is identical; The processing that is worth to the NULL of incompatible data type may be not identical.

The basic data type of the kind that the user defines is used at analysing divisional compatibility.

The decimal processing of same value in bolting to partition is identical, although their scale is mixed,precision differs.

In string (CHAR, VARCHAR GRAPHIC or VARGRAPHIC) caudal blank space can be come loose to list algorithmic oversight.

BIGINT, SMALLINT and INTEGER are compatible data type.

REAL and FLOAT are compatible data type.

The CHAR of different length and VARCHAR are compatible data type.

GRAPHIC and VARGRAPHIC are compatible data type.

Divisional compatibility does not apply to LONG VARCHAR, LONG VARGRAPHIC, CLOB, DBCLOB and BLOB data type, because they cannot serve as divisional key.

Apposition

Apposition (Collocation) it is to find a place for same a travel in the diverse list that relevant data includes in database partition. The watch of apposition makes DB2 can use join strategy effectively.

You may discover, serve as the response of pair of specific inquiry, two or many watches provide data often. Here the circumstance falls, you can hope such position that expresses medium relevant data is stood by as far as possible. Be differentiated to be two physically in the database or in the environment of many databases partition, the relevant fragment that must a kind of method can express those who differentiate is stood by as far as possible. The function that completes this process calls watch apposition.

Should access be used at join or child when the many watches of inquiry, DB2(R) general database (DB2 UDB) can identify want connective data to whether be located in on identical database partition. Then DB2 can be on the database partition of memory data executive join or child inquiry, and need not move between database partition data. This kind of parcel carries out join or child the ability of inquiry has distinct performance advantage.
1、分页标题#e#<