5.2.3 Criteria for selecting single or multi-chunk tables
First, determine whether to define the table as a row store table or a column store table. Then, determine whether to define that table as a single-chunk table or a multi-chunk table (regular multi-chunk table or archivable multi-chunk table). The following table shows the key features of single-chunk tables and multi-chunk tables:
Item |
Table type |
||
---|---|---|---|
Single-chunk table |
Multi-chunk table |
||
Regular multi-chunk table |
Archivable multi-chunk table |
||
Can the background-import facility be used? |
N |
Y |
Y |
Can a primary key be defined for the table? |
Y |
N |
N |
Can a unique index be defined for the table? |
Y |
N |
N |
Can the data in a row store table be compressed? |
N |
N |
Y |
Can the data in a table be updated or deleted (can an UPDATE statement or DELETE statement be executed)? |
Y |
Y |
P#1 |
Is data retrieval faster or slower? |
Faster |
Faster |
Slower#2 |
- Legend:
-
Y: Can be executed, or has no impact on performance.
P: Some data cannot be updated or deleted.
N: Cannot be executed.
- #1
-
You cannot update or delete the data in an archived chunk.
- #2
-
Applies when retrieving data from an archived chunk.
Note that the preceding table lists only the key features of single-chunk tables and multi-chunk tables. For details about the advantages, disadvantages, and restrictions of single-chunk tables and multi-chunk table, see the following topics. Determine whether to define a table as a single-chunk table or a multi-chunk table (regular multi-chunk table or archivable multi-chunk table) after reading these topics.
- Important
-
You cannot define a column store table as an archivable multi-chunk table.