Kbase P71611: Does -q speed up the performance when running code in a loop
| Autor |
Progress Software Corporation - Progress |
| Acesso |
Público |
| Publicação |
11/03/2004 |
|
Status: Unverified
GOAL:
Does -q speed up the performance when running code in a loop?
FACT(s) (Environment):
Progress 9.1D
FIX:
The Quick Request startup parameter -q
directs Progress to search PROPATH directories only on the first use of
a procedure.
There is not much different in performance if you run the
same procedure in a loop.
The reason for that is that the procedure is kept in the cache anyway as
long as you don't run enough other procedures between so that so that this
specific procedure is not located or found anymore in the cache.
Chapter A "R-Code Features and Functions" of the Progress
Programming Handbook gives quite good inside information about how the
cache for the r-code works.
For measuring the influence of the -q startup parameter
you have also to look at where the r-code is stored. If the r-code is stored in a library than it's the same as if you are using
-q . This is explained in more detail within the Progress version 9 Product
Update Bulletin Chapter B .
The difference between using -q or not -q for r-code which is not in a library
depends on how fast it is possible to re-read the file, how fast the file is found within the propath and also on how the Operating system is caching the files and the directory entries once read.
If the Operating system is not very efficiently handling repeatable file reads than the Progress -q startup parameter is of great value otherwise not such
great results in improving performance should be expected.