Like a new car that turns used as it rolls off the lot, all real world networks have legacy issues. Only the smallest and most trivial of enterprise and carrier networks are homogenous; the vast majority are combinations of past purchasing decisions, both successful and failed strategies and a mix between old, but proven, technologies and new, but unproven, promises. As technologists, we can fool ourselves quickly by starting with a clean sheet of paper. It is a pleasure that our customers rarely have.

Mashups - light weight applications written on top of web services or SOA architectures - accept this basic fact, and allow us to develop new applications from old components. Mashups take data and services from several sources and combine them in ways to present analysis or functionality in an interesting way. The sources of data and services may be from older equipment, or from newer purpose built machines. By using web services and SOA as the main integration mechanisms, the mashup can not only be created with high productivity web scripting methods, but the devices that create the services can be used for other applications in the future. Even when the legacy equipment does not have a web services interface, off-shoring a software development effort to write a custom library is straightforward and relatively inexpensive. For companies at the end of their product life cycles, put that feature on your checklist… many vendors now have a decent story for that. If your preferred vendor doesn’t… find one that does.

In my particular field of communications enabled business process (CEBP) development, I see these issues everyday. I would like to caution the larger enterprises out there to be careful with your vendor negotiations. Avaya may want you to upgrade all of their equipment to support your CEBP deployment, but be careful. You may be able to take your existing equipment and extend the functionality using custom software, or there may be other vendors such as LignUp, iPeria and BlueNote Networks instead of fork-lifitng your existing implementation. At a minimum, take the time to do the numbers before you sign the contract. Remember that the hidden value of web services and SOA based applications is that, since they are relatively easy to develop, they are also relatively easy to port and re-write. Before you go out and spend ten million dollars on infrastructure, pick yourself up a LignUp switch, try your application and measure your return. If you have success, then you’ll know how to move forward, either with your existing big iron vendor, from an upstart or not at all.

Posted by Thomas Howe @ 9:21 am | Filed Under Lead Stories |

Comments