|
View previous topic :: View next topic
|
| Author |
Message |
magesh23586
Active User
Joined: 06 Jul 2009
Posts: 213
Location: Chennai
|
|
|
|
Hi All,
I am planning to add AMP=('ACCBIAS=SYSTEM') to all the job using VSAM files.
1. May i add amp parameter into the DD statement directly , without checking anything.
2. If at all I add AMP parameter in DD statements, Do i need to change anything to the definition/properties of VSAM datasets or do i need add/update any specific parameter in the delete define vsam ? |
|
| Back to top |
|
 |
expat
Global Moderator
Joined: 14 Mar 2007
Posts: 8797
Location: Welsh Wales
|
|
|
|
| What happened when you tested the planned changes before implementation ?? |
|
| Back to top |
|
|
colin777
New User
Joined: 06 Jun 2013
Posts: 19
Location: Singapore
|
|
|
|
Hi,
ACCBIAS is used for System Managed Buffering - SMB. Firstly the VSAM dataset must be SMS managed and must be defined as an Extended Format dataset in the appropriate Data Class.
ACCBIAS = SYSTEM really leaves it up to VSAM to decide on Buffering with no undertanding if the VSAM processing of the dataset needs more Index Buffers or Data Buffers.
This is not the best use of SMB. If you know that the dataset is being accessed more in Sequential mode then more Data Buffers are needed. So specify ACCBIAS=SO
If in Randon or Direct mode then more Index buffers. So specify ACCBIAS=DO.
In any case any Repro that has VSAM as an output dataset should be specifed as ACCBIAS=CO - CREATE OPTIMIZED. This will work as long as the Cluster is defined with SPEED. If the Cluster is defined with RECOVERY, then VSAM internally will use CR - CREATE RECOVERY OPTIMIZED.
How do you know what SMB option VSAM used then add this JCL DD keyword MSG=SMBBIAS and VSAM will tell you on the Joblog.
HTH
Colin777 |
|
| Back to top |
|
|
|
|
