Can anyone tell me what the IDMS prefetch feature exactly means and how it works ?What is the difference with IDMSQSAM ?
Best reagrds, Frank.
I think that quite correctly you have been pointed to the CA-IDMS manuals for a detailed explanation - I don't think the purpose of the discussions area is for one participant to paraphrase the manuals for another, or to <cut>+<patse> on behalf of the original participant. But trying to characterise the nature of the difference between IDMSQSAM and Prefetch might be useful - I'll try with my limited understanding.
That would be my thumbnail characterisation of the the nature of the IDMSQSAM vs. IDMS Prefetch - I hope that it is useful at a high level.
Take care - cheers - GaryC
The basic difference is that prefetch can be used with more than one area at a time and IDMSQSAM can be used with a single area at a time.
At least that is what I understand at this point in time (things can always, and usually do, change).
I will leave it to the CA Techs to explain further.
Charles (Chuck) Hardee
Senior Systems Engineer/Database Administration
CCG Information Technology
Thermo Fisher Scientific
300 Industry Drive | Pittsburgh, PA 15275
Phone +1 (724) 517-2633 | Mobile +1 (412) 877-2809 | FAX: +1 (412) 490-9230
Chuck.Hardee@ThermoFisher.com | www.thermofisher.com
WORLDWIDE CONFIDENTIALITY NOTE: Dissemination, distribution or copying of this e-mail or the information herein by anyone other than the intended recipient, or an employee or agent of a system responsible for delivering the message to the intended recipient, is prohibited. If you are not the intended recipient, please inform the sender and delete all copies.
Frank, As far as the basic explanation of how PREFETCH works I refer you to the DATABASE ADMINISTRATION GUIDE and the section BUFFER MANAGEMENT/ USING CHAINED READS.
If you have any specific questions let us know.
If you are looking at batch usage especially, SYSIDMS parameter BUFFERSTAT is helpful to see the results of prefetch and various tuning options. When Gary stated that IDMSQSAM uses IDMS buffers, I don't believe that is correct. IDMSQSAM seems to work more effectively when the DD statement has a DCB=BUFNO=nnn parameter coded. There are also some third party bolt-on products that do better than QSAM or prefetch for pure sweeps because they read full cylinders in each I/O. We use Megabuf-Sequential from Cogito and another product used in past lives is Fast Access. Of course most disk now is not true "cylinder" geometry and just emulates that, but it still produces more bang for each I/O to read what z/OS thinks is a cylinder.
At one time – (and this may no longer be true) when PREFETCH encountered a non-sequential i/o request in an area for which prefetch was requested, the entire buffer is flushed before obtaining the random? i/o request
This would happen most often (I think) when doing an area sweep with the need to pick up info from an owner (or child) record within the same area bit o necessarily on the same page as the “swept” record.
I know that the 3rd-party product ASG-Fast-Access does not exhibit this behavior - I have never used QSAM nor MEGABUF-SEQ – so I cannot address those features/products
Retrieving data ...