I don't know about "best practices", but we: - Make religious use of LPAR grouping - Test images in a test folder, Prod images in a Prod folder. That at least keeps the Operators from activating a Prod image when they meant to activate a test image. We still get the occasional brain check where the wrong LPAR was activated in the middle of a change window. That can be fun. (NOT)
- Lock everything. However, experience has shown that once an operator "gets used" to the extra step, that operator can, but not necessarily will, get numb to it. - Make religious use of activation profiles. We would like to use the dynamic settings approach described earlier, but we that responsibility would still fall in the hands of Operations. - Practice, practice, practice. Our Operators IPL our sandbox systems at least as often as our sysprogs do these days. - Train, coach, and most of all - communicate. Every request is confirmed with the originator and on-the-spot peer-reviewed with another Operator. It takes an extra minute or two, but I think it's worth the effort. We have an extra body in on change weekends just to run point. - Have a system of checks and balances. Only Operations is responsible for updating profiles, weights, etc., but they do it with information provided by sysprogs, performance, storage, etc., and everything is cross-checked before anything is updated. There's always one in the bunch who takes a little more patience and tolerance than the rest of the group, but that's true of sysprogs, storage managers, and performance analysts, too. Hopefully, appropriate use of PR's will address the situation if nothing else seems to work. I also had one instance where the entire group kept me up nights (literally... the all-too- frequent 3AM call - "were getting message IEAxxxI and we've never seen it before, what does it mean?"). That meant time for me to seek alternate employment. FWIW, Art Gutowski Ford Motor Company ITInfrastructure ---------------------------------------------------------------------- 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
