>...and you get what vendor wants. BTDT. Again, I disagree.
>The performance strongly depends on workload type. Absolutely. >How do you define "my workload must be performed well and I must be happy of >that"? The easiest is to give them RMF data from your real Production workload. >I saw (even did for some vendors) how much the numbers depend on type of workload. Yes, but. >And *I* prepared the workload for test scenarios - just to prove assumed performance numbers. Who said anything about test scenarios? Give them the RMF data and make them run their models (they all have them). Make sure the performance guarantees are there before the contract is signed, not after. I don't believe in benchmarketting; the only TRUE test is Production. So, agree on what RMF data is applicable, before and after. Use your own data and make them pay attention. Testing scenarios are meaningless. - I'm a SuperHero with neither powers, nor motivation! Kimota! ---------------------------------------------------------------------- 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
