Kbase P25106: Why dump/load improves performance?
| Autor |
Progress Software Corporation - Progress |
| Acesso |
Público |
| Publicação |
23/03/2011 |
|
Status: Verified
GOAL:
Why dump/load improves performance?
GOAL:
How to replace (*) a dump and load done for performance reasons?
FACT(s) (Environment):
All Supported Operating Systems
Progress 9.x
OpenEdge 10.x
FIX:
After the creation of a new database, a "fresh" definition and data load, you have the optimal storage of the data.
The blocks are almost full, there is the minimum of blocks on the RM chain
All records will be of the last version of the table definition and there will never be an overhead due to schema versioning when accessing them.
(*) It is to be noted that an update of each record (without modifying its value) will also make all records to be of the last version of the table definition.
You have the lowest scatter factor /* in case of non concurrent data loads */ and that results in quicker access of many records from the same table.
(*) It is to be noted that moving a table to its own area (version 9 or later) will also reduce its scatter factor.
The indexes are compacted as their optimum and therefore less index blocks need to be read for records access.
(*) It is to be noted that an "idxbuild" or "idxcompact" will achieve the same result.