On Thu, 19 Mar 2009 16:25:12 -0500, Mark Zelden wrote: > >It's a generational / platform problem. It seems that even the younger >(relative term) mainframe people I work with just can't grasp the concept >that 3390-9 isn't "huge". We were just going through an internal battle >trying to talk our storage management folks - aka DASD freaks :-) - >into 3390-54 instead of 3390-27 for a migration that is starting. >Because these folks have micro-managed mainframe storage for so >long they are afraid to do it any other way. > And some of them were likely carrying twice the storage of a mod-54 in their shirt pockets.
However, SMP/E's design is short of ideal. The scheme is to unpack all the SMPPTFIN files to HFS; dynalloc all concatenated to SMPPTFIN and do the RECEIVE. In a test, I have even exceeded the OS limit on DD concatenations. A more economical approach would be to allocate SMPPTFIN to the output of a POSIX pipe and in a concurrent process unpack each SMPPTFIN archive, write it to the input of the pipe, and delete before unpacking the next. Yah, I know, this would create new error scenarios where an unpack failed midway in the process. It can be dealt with. -- gil ---------------------------------------------------------------------- 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
