All,

We are revisiting this scenario pretty seriously.  I originally posted
this back in April of
this year.  Has anyone done this yet?  What was the performance like?
David Boyes
responded back then with some very impressive numbers.  I'm wondering if
he would
mind passing along any additional comments?

Our cost case is based primarily on cost avoidance, without this change,
we will be forced
to upgrade the traditional mainframe side within 12 -18 months for
capacity.  We have done
a 3 year financial plan, and even after investing in the technology,
training etc, it appears to
be a $200,000 savings in that time frame, assuming we can forestall the
upgrade and the
associated hardware/software upgrade fees.  We want to grow
the DB2 environment considerably by converting the applications running on
20+ instances
of INFORMIX running on AIX to DB2 on Linux on a z/800.  If we can do this,
the savings
more than doubles based on the retirement of the INFORMIX license charges,
even if
we need to add an additional IFL.

Of course there is a lot of FUD running around with the DBA's right now
because this is a
new idea for them.  Now that we are getting more serious on this idea,
they will be
getting involved, but I want to give them some confidence that this is
being done today and
is a viable situation.

Any feedback, or pointers to Whitepapers, etc would be helpful.

Dave



David Boyes <[EMAIL PROTECTED]>
Sent by: Linux on 390 Port <[EMAIL PROTECTED]>
04/04/2003 08:51 AM
Please respond to
Linux on 390 Port <[EMAIL PROTECTED]>


To
[EMAIL PROTECTED]
cc

Subject
Re: Thoughts on DB2...






I'm just finishing up a cost case for a similar environment, and the
numbers look very good if you can live with some of the technical
restrictions. As others have said, there are limits on the size of the
databases supported by UDB, however with the ability to run multiple
copies of UDB in the VM system, this is not necessarily a show-stopper
if you distribute individual tables to separate databases. The setup for
DRDA over TCP is pretty straightforward, and the use of hipersockets is
actually a very nice setup for this type of operation.

The cost savings is pretty dramatic -- in the case I'm working on,
reducing the size of a z/OS partion by 1 standard engine and enabling
similar workloads on 2 new IFLs reaps a cool $350K savings in the first
year (by the time you include the reduction in z/OS licensing and the
corresponding reduction in ISV prices for a smaller z/OS machine
footprint, you're talking about real money).

Is it technically perfect? Not yet. Still, $350K/yr will buy a lot of
remediation for the current problems.

-- db

David Boyes
Sine Nomine Associates


> -----Original Message-----
> From: Linux on 390 Port [mailto:[EMAIL PROTECTED] Behalf Of
> Dave Jousma
> Sent: Thursday, April 03, 2003 2:34 PM
> To: [EMAIL PROTECTED]
> Subject: Thoughts on DB2...
>
>
> All,
>
> On our traditional mainframe system, we are running a
> z/800-001 that is
> maxed out on CPU.  We are trying to figure out ways to relieve this
> without buying another GP engine, and pay the big software costs.
>
> What I'm wondering is would it be possible, and if so, how smart would
> it be to try and move our DB2 workload to a Linux/390 server, and then
> have all the "traditional" mainframe apps access it remotely?  Is this
> a path worth looking at?
>
> The reason I ask, is that and IFL bundled with z/VM is pretty
> cheap, and
> and I could utilize hipersockets.
>
> Waddya all think?
>
> Dave
>
> __________________________________________________________
> Dave Jousma
> Lead Systems Administrator - Information Technology
> Spartan Stores, Inc.
> PO Box 8700
> Grand Rapids, MI 49518
> (616) 878-2883
> Email: [EMAIL PROTECTED]
>

Reply via email to