|
|
||||||||||||||||||||
| Home > SOA News > Grid lock in | |
| SOA News: |
|
||
Market Analysis
Grid lock in For the sake of this discussion I will define "grid" as an environment where resources can be added and subtracted in a flexible and effective manner to support a varying workload. We have heard a great deal about these ideas from the industry leaders: IBM, Oracle, Sun, Microsoft and some initial thoughts from BEA. And there is the rub, each of these vendors has a somewhat different view of what grid is and an even more different view of how to make it happen. The biggest difference is how homogeneous or heterogeneous the workload or the resources can or should be. Oracle, for example, starts from the premise that all applications will run under Oracle10g and run on lots of five thousand dollar Lintel machines. This is a simplification of their view (for which I apologise) but has enabled them to provide grid solutions today. IBM, on the other hand, has a large customer base with a heterogeneous workload and technology that it needs to support; and a belief that standards and continuous change are the way of the world. So they are developing a much more flexible grid, specifying standards that will be required to support it, and delivering parts now. However a full grid implementation is not likely until 2006-7. This will be based on the Open Grid Service Architecture (OGSA) and will have a system management bus built on WebSphere. IBM will share the standards and interface definitions with the industry, but there will be a large incentive to get as much of the infrastructure as possible from IBM as it will all work together. Once a grid solution is implemented it becomes much easier to outsource it in parts, but only to a vendor that supports the same grid. Oracle 10g is built based on Oracle's experience of providing outsourcing for their applications; and therefore if an enterprise implements 10g Oracle outsourcing will be well placed to support them. Over the last few years there has been a move to open standards that will allow plug-and-play and mix-and-match; this has enabled enterprises to break free from the lock-in of the vendors. It will take a long time for this to spread effectively to the grid. The message that buying the infrastructure software and the grid management from one vendor because it is easier and will work together is also true of the infrastructure software and development tools. For instance buy BEA WebLogic and you should buy BEA Workshop, buy IBM WebSphere and you should buy WebSphere Studio. The conclusion is that over the next few years mixing and matching solutions from different vendors will become more difficult and therefore much less attractive. The problem is that this means a decision made in one part of IT will have a more profound effect on other areas. The operational (grid) environment chosen will impact the choice of development environment, which will effect the choice of infrastructure, which will impact the choice of applications, which will impact outsourcing decisions, and vice versa. Grid lock-in to one vendor may be unavoidable. Making the correct holistic architectural decision for a particular enterprise is going to be harder but also more important. This is a brief introduction of a complex area that I hope will generate further discussion and debate.
Copyright 2004. Originally published by IT-Director.com, reprinted with permission. IT-Director.com provides IT decision makers with free daily e-mails containing news analysis, member-only discussion forums, free research, technology spotlights and free on-line consultancy. To register for a free e-mail subscription, click here. For more information:
'); // -->
|
|
|||||||||||||||||||||||||
| About Us | Contact Us | For Advertisers | For Business Partners | Site Index | RSS |
|
|
|
|||||||