Let’s take a step back and consider the evolution of technology.
First of all, hardware has gradually migrated out of the enterprise and almost literally into the ether. From the mainframe to the personal computer, virtualization, and the cloud, hardware has shifted from something the enterprise maintained and managed at a high fixed cost, to something the enterprise can access as needed at a variable cost dependent on availability and usage.
At the same time, we’ve seen an interesting change of the role played by the system platform. In the past the hardware vendor tended to drive the enterprise’s choice of operating system and this OS in turn drove the choice of applications. Bottom line was, you invested in applications that ran on the OS that ran on the your hardware.
The advent of virtualization and the broader adoption of hypervisor technology, however, have effectively severed this tight linkage between applications and specific hardware/OS configurations. In fact, as soon as you could bundle just enough operating system (JeOS) with an application in a workload or appliance, there was no longer any reason to ship an OS with hardware in the first place.
To meet the demand generated by this shift, in addition to the traditional Java and .NET stacks, we now find a developer stack emerging for applications that sit in the cloud.
In the end, the most striking aspect of these transformations has been what might be called “the liberation of the application” in the form of total portability. Separated from specific hardware, and even a specific geographical location (“our data center”), indeed, carrying it’s OS with it, the application can go where it will and, ultimately, run anywhere.
Long story short, the way we have traditionally thought about enterprise computing – one integrated stack of physical hardware, operating system, middleware and application – no longer works as the paradigm for computing, which has become increasingly distributed, dynamic, and flexible.
This flexibility and dynamism presents us all with a new array of challenges. For developers, the many new possibilities mean they have to develop applications that can run in physical, virtual, or cloud environments, which might include VMware ESX, Microsoft Hyper-V, Amazon EC2, IBM Big Blue or some other cloud that hasn't even launched yet.
For the enterprise more broadly, the challenge is figuring out how to make the most of this new world’s great potential while managing its inherent perils, security, for example.
Overcoming this challenge will bring us to the next stage of technological evolution which will undoubtedly involve a new form of distributed intelligence in which the workload itself becomes intelligent.
Posted
Feb 17 2010, 11:37 AM
by
JoelRichman