Hi David,
I take the point that it's possible to make a case to *not* upgrade a
stable, working application.
Many of the Ingres enhancements (and there are some nice ones:
http://community.ingres.com/wiki/Ing...es_Per_Release)
have to be coded-in or 'taken advantage of' in some way.
If the App has been ticking along nicely for years, there is not much
incentive to make such changes.
I think a killer reason to upgrade is the license fees you mentioned,
Ingres 2.6 is on Extended Support (i.e. out-of-date support)
which *is* more expensive than Standard Supp, how many CPU's are you
paying for?
I agree that the move from 2.6 to 9.x is pain-free,
the only glitches we had were a couple of queries that fell through
the cracks of the optimizer enhancements; run-times increased by x10.
So I guess that the main operational reason for you to upgrade is that
you'll be _ready_;
Ready for an upgrade to new HW, ready for a new JDBC interface that
expects a DAS Server,
Ready to use some new feature that your not expecting to need right
now.
HTH
Steve
On Sep 29, 10:16*pm, David Brooks <dbro... (AT) visionsystemssoftware (DOT) com>
wrote:
Quote:
After looking at all feature addition and enhancements, and realising
that Standard Support versus Extended Support could mean a difference in
license fees, and also that 2.6 is 32-bit and 9.2.0 is 64-bit (on
Solaris) so that might be a additional edge, what would be the other
most compelling reason(s) to make one move from up, especially if your
installation is sitting there working fine, for your application,
without problems?
Is there a benchmark on performance between the various version of Ingres?
Thanks,
David. |