|
View previous topic :: View next topic
|
| Author |
Message |
Mistermind
New User
Joined: 08 Feb 2008
Posts: 46
Location: Dublin
|
|
|
|
Anybody used this before:
If processing large HDAM or DEDB databases
then sorting the driving input into the same
physical sequence as the DB will significantly
improve performance. This process is often
referred to as a RAPSORT and is simply an inhouse
utility that reads the input file, gets the
DB key and invokes the appropriate IMS
randomiser to get the physical offset in the
DB, and then creates an output file with each
record prefixed by the physical offset value.
This file is then sorted in the order of the
physical offset, striping the offset. The new
input file is now in the same order physically
as the DB. |
|
| Back to top |
|
 |
Sandy Zimmer
Active Member
Joined: 13 Jun 2007
Posts: 826
Location: Wilmington, DE
|
|
|
|
| Quote: |
| RAPSORT and is simply an inhouse utility |
We would not have experience with this unless we were there. If you try to LOAD a keyed database out of sequence, the LOAD will not work. HDAM does not have a separate index, but it is keyed. I always just used syncsort - any external sort will work - sorted the input file in key sequence and then ran the load. |
|
| Back to top |
|
|
|
|
