Steve,
Your email is very interesting. Here, there are two issues that are
determining the future of mainframe and client/server Open systems
environment (ie UNIX).
The first issue is the philosophical issue that it can be done
cheaper on client/server. Management refuses to look at the total
cost of operation (TCO) of the client/server setup. I am simply told
to get with it, this is the way of the future. Yet they see the big
cost of the mainframe. They never compare apples for apples. It was
recently announced that a HP Superdome was installed here. In
a nut shell, a mainframe version of client/server, but we wont call
it that.
For what we have spent in client/server hardware over the past 10yrs
to for People Soft, we could have bought several mainframes from IBM.
It recently came to light that we have not had good backups of all of
the client/server systems. One manager started to panic and a
mainframe solution came to light . Work began on implementing
it. Then it was quickly halted as it was....a mainframe
solution. We would shudder to think of having shoddy backups of
the mainframe. Yet it is ok for other systems?
The second issue here is the funding model. The mainframe has always
been a cost recovery model. We charge our mainframe users for CPU
and disk space based on what it is costing us. Yet the systems
supporting People Soft, are centrally funded. More or less a black
hole as they pretty much get what they want. As more
departments use less of the mainframe, we have to cut more and more
of the mainframe budget. There appears to be no constraints on the
open systems side.
I wont even touch the topic of money spent on consultants here for
People Soft. We were even mentioned in Information Week 10yrs ago
about our investment in consultants for People Soft and it continues.
Fortunately the mainframe will be here long for me to get my 30yrs in.
Duane
(As with Colin, these comments are my own.)
