On Sat, 12 Jul 2014 15:51:38 -0700, Hardee, Chuck wrote:

>Does anyone know if one can issue WRITE or PUT requests against the DCB that 
>has been opened and being used by SNAP or SNAPX requests?
>
>According to the manual, the SNAP DCB is defined as MACRF=(W) and is variable 
>length LRECL=121.
> 
I'm with JWG here: LRECL-125 is much more common for RECFM=VBA.  And if 
MACRF=(W), you want
to use WRITE, not PUT.

>What I want to do is write 0 to n "explanation" lines prior to the SNAPs of 
>one or more blocks of memory.
>I already have the SNAP working and have no problems issuing requests for 1 or 
>more storage areas with an associated "header" line.
>Now I want to write additional lines preceding the snaps.
> 
I've had indifferent results trying this sort of thing.  A lot of it depends on 
whether
the Other Guy has a buffer partially filled.  And when you do a CHECK, will you 
be
CHECKing your WRITE or the oldest unchecked WRITE issued by the Other Guy?
And it may work differently for SYSOUT as opposed to DSORG=PS data sets.

This sort of thing is much more predictable with UNIX unbuffered I/O -- data 
goes
to the file in the order in which write() was issued.  But in your case:

o SNAP(X) may refuse entirely to deal with UNIX files.

o SNAP(X) is using BSAM and doing its own buffering.  That would feed an
  emulation interface which may further use UNIX buffered I/O.

Don't expect what you expect.

-- gil

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to