[Update 13 March 2009: I decided to update the title on this series of posts from “Talking Points” to “Architecture Strategy” as it better describes what I’m talking about.]
[Update 2 April 2009 added reference to previous post show why GSI projects are just so hard to pull off. The conclusion is fairly easy to make that our technical capability to engineer the hard & software infrastructure is outstripping our social ability to implement it.]
Global Single Instances as an applications strategy, particularly for ERP, are something that Oracle has advocated for more than 10 years now.
The concept has become broadly popular in the IT industry as customers consolidate their systems in search of greater efficiency, flexibility and of course lower costs.
These days the software and infrastructure can deliver this capability with (relative) ease.
The issue however is that such projects are extraordinarily difficult to pull off successfully. I talked about this in my post on Enterprise Architecture as Strategy
Over the next couple of months I'll be publishing a series of posts on Global Single Instance (GSI) Architecture Strategy to cover a number of recurring themes that come up in customer conversations. A glorified FAQ if you like, based on 10 years of doing these projects.
As a first grab on some of the typical questions, I’ll list a few out here. Readers are (of course) invited to contribute additional suggestions for subjects!
- GSI’s: How many and where? Functional splits?
- Multilingual. Do we really need to install all of these languages?
- Localizations. The quagmire of dealing with exotic local (statutory and/or fiscal) reporting
- Test environments for GSI projects. How many is a good thing?
