On Thu, 18 Sep 2008 08:54:07 -0500, Tom Marchant <m42tom- [EMAIL PROTECTED]> wrote:
... >At my last shop, a full refresh has been known to cause problems for the IMS >folks. I never understood why, and they claim that it was not because they >had prestaged updates to LLA managed data sets. Knowing what I do about the >change management process there, I tend to believe them. I do know that >there were times that they came looking for me because I had performed a >full refresh and they had to cycle some IMS regions. > >I don't know what else they might have been managing with LLA. Remember >that it's not just for linklist. It still goes on here. There are performance, as well as CM implications. Our LLA MAXVIRT is a little more than 16M, but a little less than 300M (varies by system), so a full refresh has a detectable impact, even outside of IMS. The IMS team does a fair bit of pre-staging. I don't know a hill of beans about IMS (I can barely spell it), but if PSB, ACB, DBD, etc., changes require a region cycle, and these are LLA-managed, that's where the trouble lies. "Us" z/OS folk are gradually doing more pre-staging these days, so a full refresh runs similar risks "our" products. Bottom line - we try to stick to selective refresh, even on our sandbox (which have IMS, DB2, and CICS, too), for both performance and availability reasons. Regards, Art Gutowski Ford Motor Company ITI ---------------------------------------------------------------------- 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