It seems to me that tuning is usually like shampoo instructions; lather, rinse, repeat. Find the bottle neck, fix it, find the new bottle neck. You need the tools to find the bottlenecks in the critical paths. There may be some jobs that run a long time but aren't in a critical path, so in a sense, their goal is being met. In fact making them faster, if it steals resources from something more important, could be counter productive. Long ago, the best thing that happened is that a printer went down and a "critical" report didn't get printed. Guess what, nobody complained. We asked around and after some digging, nobody used the report. Elimination is the ultimate tuning. Our last problem was that the production support people would not build concurrancy where they could. If 3 jobs depended on 1 job, they would run those 3 jobs as a series. We didn't fix it, we're "moving off the mainframe" and that was about the time our usage actually started declining.
---------------------------------------------------------------------- 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
