________________________________ From: Hunkeler Peter (KIUP 4) <[email protected]> To: [email protected] Sent: Mon, May 2, 2011 1:32:56 AM Subject: Re: assembler batch calling unix daemon
Michael, I dare to comment on your "vast overhead of creating an ASID". While I agree that creating an address space in not a cheap thing in z/OS, it is by far not as often needed as your statement might imply. Only if there is no idle BPXAS UNIX services initiator available, is it that a new address space has to be built. Peter: Interesting comment. About 25 years ago I was talking with an IBM SE out in our NY office. She was proposing that we go IMS out there (I was surprised as they really didn't need it IMO). I asked her about address creation as she was going with the then IMS concept of every transaction was going to a new ASID. Ishook my head and said to her that I thought ASID creation would be a killer in overhead. She assured me it wasn't all that bad. I am not an IMS person for sure but it sounded a little bit off but I could not challenge her on my knowledge of IMS just on ASID creation. I asked her to double check with her sources as (at least then) ASID creation was costly. She was self assured that it was minor. I got a hold of some IBM people I knew from GUIDE and they suggested that she read some manual (Long time forgotten) and I did. She backed down on the idea though. They never went IMS so it was a mute point. Ed ---------------------------------------------------------------------- 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
