Kbase P99891: How to use full Table scan for Type II Areas with binary dump ?
| Autor |
Progress Software Corporation - Progress |
| Acesso |
Público |
| Publicação |
12/16/2008 |
|
Status: Verified
GOAL:
How to use full Table scan for Type II Areas with binary dump ?
GOAL:
How to take advantage of binary dump from Type II Areas?
GOAL:
How to speed up an binary dump ?
GOAL:
Is binary dump faster for Type I or Type II Storage Area architectures?
GOAL:
Is there a way to speed up a binary dump for a Type II Storage Area?
GOAL:
What is the "-index 0" option for a binary dump command?
GOAL:
What does the option "-index 0" do when used with the binary dump command?
FACT(s) (Environment):
OpenEdge 10.0B
All Supported Operating Systems
FIX:
A table scan will take an advantage of Type II Storage Areas and will perform the dump in the fastest way. In order to have binary dump for a Type II Storage Area make use of a table scan, append the "-index 0" option to the "proutil <db> -C dump" command.
Example:
$ proutil sports2000 -C dump customer . -index 0
Table scans are facilitated by the addition of 'clusters' to the structure hierarchy of Type II Areas. If a table is in a Type I Storage Area, binary dump will use the primary index instead of a table scan.
Small benchmark comparison:
OpenEdge 10.1C01
Table Records Size Min Max Mean
PUB.Customer 1000000 2433MB 27B 20063B 2551B
d "Cust_Data":9,32;8
with -index 0: 1' 23" circa. 15% faster
with logical index: 1' 31" circa. 7% faster
with default (prim index): 1' 38" BASE