Post by Mike Schwab
On Tue, May 15, 2018 at 8:33 AM, John McKown
Post by John McKown Post by Kurt Quackenbush Post by Mike Fulton
I know in ISPF under 3.2, if I am allocating a dataset and it is
SMS-managed, I can specify the space in bytes (e.g. megabytes)
instead of tracks or cylinders or blocks. Is there a similar way to
do this from the TSO ALLOC command? Looking at the manual, I can't
see it, but I would have thought if it could be done in ISPF, it
could be done in TSO...
It sure would be simpler if you could just specify the number of bytes,
but that's not how it works. Instead you specify the length of an
Post by John McKown Post by Kurt Quackenbush
record (AVBLOCK) in bytes and the multiplier (AVREC).
o you simply set AVBLOCK to 1 and AVREC to number of bytes?
It would include a lot of inter record gap space.
Even AVGREC=80 and using blocksize 8000 would add a lot of inter
record gap space.
I was thinking of doing something like:
SPACE=(1,(1048576,1048576)),RECFM=U,BLKSIZE=27998,AVGREC=U to get about
1MiB of disk space. When I ran a job to do that, the dataset was allocated
19 tracks. 19*56K=1064K or about 1MiB. If I change the BLKSIZE= to 8000
instead, I get 22 tracks. So DADSM is properly calculating the inter-block
gap to give me enough space to store about 1MiB of user data in my dataset,
assuming that every block is BLKSIZE length. My first example is 1/2 track
blocking, which is what I use a lot when I put the contents of a UNIX file
into a z/OS dataset; especially the output from PAX. In the past, I used
RECFM=FB,LRECL=1,BLKSIZE=27998 for PAX output because I was thinking that
was a good way to indicate that the data was a "byte stream".
Post by Mike Schwab
Mike A Schwab, Springfield IL USA
Where do Forest Rangers go to get away from it all?
We all have skeletons in our closet.
Mine are so old, they have osteoporosis.
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to ***@listserv.ua.edu with the message: INFO IBM-MAIN