Comments on: Record Blocking Factor

By: boomersooner

boomersooner — Sat, 02 Apr 2005 12:49:45 +0000

You say the file’s size has been constant but processing of the job has slowed down. So what else has changed on your system ? Any other new batch jobs or system activities running at the same time, does the problem job have lower priority than other jobs, is it in a memory pool that is now too small, and when is the last time you reorg’d the file ? Any of these could affect the job — especially if you’re no reusing records from the file, but may be deleting records. The total # of (active) records won’t change (much) but the total file size will — and deleted records can slow the process.

If a job’s processing changes and nothing has been done with/to it, look at something besides that particular job.

By: ddomic

ddomic — Fri, 25 Mar 2005 04:10:42 +0000

Check IBM Software Technical document: Blocking, Sequential Only, and the Effect on a Program
http://www-912.ibm.com/s_dir/slkbase.nsf/1ac66549a21402188625680b0002037e/d6738e1cd37e1f33862565c2007cef79?OpenDocument&Highlight=0,Block,Length

By: rchevalier

rchevalier — Wed, 23 Mar 2005 11:33:07 +0000

The 32k size was incresed to 128k several releases back. Divide it by the record size and round down.

If you want it to change with file size changes retrieve the record length and recalculate as stated previously.

By: fanbot

fanbot — Wed, 23 Mar 2005 07:18:16 +0000

P.S. Forgot to mention. In the event that a file is changed your calculation, if hard coded, will degrade performace. The best is to retrive the file definition prior to execution and use the retrived record length for the calculation.