No, the request was for an LRECL of 32760, NOT A BLOCKSIZE of 32760. BLKSIZE=0 was chosen precisely for half track blocking on DASD with VBS; a BLKSIZE=32760 on DASD can waste LOTS of disk space.
So there never would be a BLOCK greater than half track; I was showing (first time in my MANY-YEAR history with him) that Bill was incorrect; physical blocks of less that BLKSIZE can be written in the interior blocks, but the (I'm pretty sure) will all be within 4 bytes of that half-track. The historic use of VBS was precisely for SMF in OS/360, when the NUC was 86K and SMF want to write 32K logical records. Using VBS with a pair of 500 byte buffers took only 1K of that 86K but permitted logical records of 32760 to be created (RECFM=VBS,LRECL=32760,BLKSIZE=500). NOT PARTICULARLY EFFICIENT, but met the requirments. Barry -----Original Message----- From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of Paul Gilmartin Sent: Monday, June 10, 2013 11:33 PM To: [email protected] Subject: Re: I/O Optimization On Mon, 10 Jun 2013 17:18:48 -0500, Barry Merrill wrote: >VBS Blocksize in dumped SMF data with RECFM=VBS,LRECL=32760,BLKSIZE=0 >does NOT produce interior same-sized blocks, as this small file from my DEV >machine shows: > > LENGTH FREQUENCY PERCENT > 15712 1 1.45 > 27995 2 2.90 > 27998 66 95.65 > >The 15712 was the last block, but other two blocks were interior blocks >that were not quite filled. I think I recall seeing similar values of >up to 4 bytes less than full blocksize when written to tape with BLKSIZE=32760. > You may have asked for 32760, but no block is even close to that; something appears to be overriding to half-track. In any case, your observation strongly contradicts DASDBILL2's (theoretical?) assertion. No I/O errors? -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
