> It's not that difficult to create a reliable box when it's > only running one or two static applications. Under those > conditions it *better* be reliable. otherwise it's a mere toy.
Reliability is only one part of the equation. It is almost trivial to demonstrate that you can build high availability systems from low reliability parts. It has been done that way in many disciplines for decades. It used to be in things like switching networks, phone systems etc. Now with the widespread availability of ultra-cheap chips it has made its way into commodity computing gear. Heck even your car probably has redundant systems built into it these days. > A mainframe is a different beast, offering many, many diverse > applications to thousands of users all on one box. Well occasionally that's true. The number of users depends on the application(s) and the diversity likewise. There is no doubt that the ability to manage competing workloads is still a mainframe unique quality of service - for now. CPU and I/O performance leadership are no longer mainframe features. We lost that title a good while ago and are unlikely to get it back. > I've yet to see unix or windows accomplish the same thing. See Ron Hawkins' last post on this. If you're running modern disk hardware you're running a fault-tolerant unix application that is solely responsible for the integrity of your data. You could be scared about that, or not. But it demonstrably works. > You have to compare apples to apples when talking about reliability. Rubbish. Reliability is a "nice to have" feature that costs a ton of engineering dollars. Go visit the IBM R&D labs to see just how expensive it is to deliver the sort of reliability you expect from P and Z series (largely the same people, and hardware and production line BTW) Availability is what users perceive and that can be delivered a lot less expensively. As long as the user doesn't perceive incorrect results or degraded service, (s)he doesn't give a rip how it is done. We as a community have to get over the idea that the mainframe is the only, or even the best way of delivering high availability. There are lots of other systems that do it quite mundanely every day. Mainframes have some wonderful features and still have some technical advantages, but those are rapidly diminishing. Chest thumping is out of place at this point. It is the ninth inning and we're behind. CC ---------------------------------------------------------------------- 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
