Charles, my response has nothing to do with track capacity, but since you mentioned an unexpected SB37-04 and previously talked about a blocksize mismatch between JCL and actual allocation ... Is the user's program, by any chance, a COBOL program? Does the JCL hardcode a blocksize value (or BLKSIZE=0), yet you still end up with an unblocked dataset? If the answer to both questions is Yes, please check the program source: In the FD of the output file in question, did the programmer explicitly code the phrase "BLOCK CONTAINS 0 RECORDS"? If that phrase is omitted, you get a compiler default of "block contains _1(one)_ record". I found that forgetting to put this clause on FD definitions for sequential output files is a frequent cause of problems.
HTH. Regards, Ulrich Krueger Mainframe Systems Services National Semiconductor Corp. Santa Clara, CA 95051 Tel: (408) 721-8071 Email: [EMAIL PROTECTED] "IBM Mainframe Discussion List" <[email protected]> wrote on 06/16/2006 12:39:05 PM: > I care because I am trying to debug a problem at a customer site. There are > about a thousand considerations that I have left out of my note - life is > more complex than the average listserve question. If I had posted the entire > universe of issues, it would have been a very long note. I am looking for > clues so as to debug how an unexpected SB37-04 might have come about. I > don't have direct to the customer machine. The customer personnel are > incredibly busy. > > Is that sufficient justification for asking this question? > > Charles ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
