To me, the biggest difference between MIPS and MSU is the fact that MSU are actually reported and quantified. True, they are of limited use in comparing different platforms but they are of some use in upgrading platforms.
And, they are a MEASURED, REPORTED mechanism that all contract parties can agree to if contracts are capacity based. If you like MIPS, I'd recommend developing a conversion formula of your choice to report the MIPS to management. None of this solves the fact that workloads run different on different boxes. The difference here between MIPS and MSU is that published machine MIPS ratings are adjusted for each workload while MSU is a straight reporting procedure. The superiority of MSU for use, in my opinion, is that the MIPS rating of a box must be agreed on by all interested parties while the MSU used is the same. How many MSU a box should deliver is wide open to debate; how many MSU is actually delivers is an objective number. Before anyone starts to beat me up for that statement, I am aware of the fact that z890 and z990 boxes report different MSU rates in TYPE30 SMF records as opposed to TYPE72 records (thank you IBM marketing - NOT!) but it is still an objective fact and the numbers, used consistently by not mixing SMF record types when reporting workload, are still measured and reported. When could MIPS ever make that statement? Just my three cents (inflation's a bear), Jim Horne Lowe's Companies, Inc. ---------------------------------------------------------------------- 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
