Re: Output record blocking

[Tony Carolla <carolla@xxxxxxxxx>]

Oops.  I pasted in the same f-spec in the email.  The second f-spec is:

     FIRvp1     IF   E             Disk    ExtFile('DMME/RVP1')
     F                                     Rename(Rrvp1:Rrvp2)
     F                                     Block(*Yes)
     F                                     UsrOpn

Of course, the pgm wouldn't have compiled with the same file, same
name, etc.  This is the 'other' instance of the same file.

TIA!


On Fri, 5 Nov 2004 10:40:03 -0800, Tony Carolla <carolla@xxxxxxxxx> wrote:
> I have written a program that is doing output to a file.  The output
> file is declared as such:
> 
>      FRvp1      O    E             Disk    ExtFile('DMME/RVP1')
>      F                                     Block(*Yes)
>      F                                     UsrOpn
> 
> When writing to this file, if I look at the 'Display Open Files'
> display for the job, the I/O Count and Relative Record number for this
> file stay the same.  As I understand, this means that the program is
> performing one write operation for each WRITE op that is executed in
> the program (not using blocking).
> 
> I have no overrides to the file, and I am simply usign the WRITE op.
> The only two things that might be causing it are:
> 
> 1)  I am writing from a DS, i.e.
> Write Rrvp1 RvpOut;
> The DS was defined:
>      D RvpOut          DS                  LikeRec(Rrvp1:*Output)
> 
> 2)  I have the same file opened for input with a different F-spec:
>      FRvp1      O    E             Disk    ExtFile('DMME/RVP1')
>      F                                     Block(*Yes)
>      F                                     UsrOpn
> 
> 3)  I have two logicals over this file opened also, with blocking turned off.
> 
> When I look in the job log, I get the message:
> Open of member RVP1 was changed to SEQONLY(*NO).
> I have never understood that message, or what causes it.  I have the
> file declared with no 'K', so I don't understand why this message says
> that I opened it in some 'seuqential' way... But it might not be
> related.
> 
> I like things to run fast, and without record blocking on this
> program, I know it could go faster.  Has anybody else had this
> problem?
> 
> 
> --
> "Enter any 11-digit prime number to continue..."
> 


-- 
"Enter any 11-digit prime number to continue..."