Creating Value: Hardware + Application Virtualization
Recently VMWare announced its intent to acquire the privately-held Java tools vendor SpringSource. Perhaps it’s one more indication that applications and workloads will increasingly rely on a virtualized (and eventually a cloud-based) infrastructure. For service providers, this represents yet another area of opportunity for them to create value for their customers. Should we anticipate that the industry-leader in server virtualization will more tightly integrate application deployment and management with the physical and virtual server infrastructure? According to Tony Baer, senior analyst at Ovum, “The deal marries hardware virtualization with a form of application virtualization, where logic is abstracted from the underlying plumbing. Combine the two, and you have a stack carrying the elasticity of dynamic, economical cloud deployment, regardless of whether it is deployed inside or outside the customer’s data center.”
So where is the opportunity for SPs? I believe it lies in continuing to do what SPs do well, and expanding those capabilities “up the stack”. What do I mean by that? Well, SPs strive to provide superior infrastructure services to their customers. Whether it’s the transport infrastructure of a traditional carrier or the data center infrastructure of a “managed services” provider – the SP’s inherent value is in its ability to deploy and manage infrastructure at scale. Right now, SPs are focused on the cloud computing category called Infrastructure-as-a-Service (IaaS). This is no surprise. Most SPs typically lack the software development capabilities of a Google to deliver Platform-as-a-Service (PaaS) or the vertical application expertise of a Salseforce.com to deliver Software-as-a-Service (SaaS). Granted, some SPs have partnered very successfully to deliver SaaS offerings, but that is the exception rather than the rule. SP roots are in infrastructure, not applications. Where it gets interesting is in the middle ground of cloud computing – the area often termed “Platform-as-a-Service.” Until now, delivering PaaS offerings seemed to involve complexity closer to the complexity of SaaS. Further, not only did the service require the platform, but that platform needed a following of application developers in order to be successful. When a provider delivers a virtual application environment capability on-demand, billed on a usage basis – and without exposing the complexity of the underlying infrastructure – well, clearly that’s the essence of a solid PaaS value proposition. If SPs build their cloud computing infrastructure in such a way that they can easily accommodate the eventual migration of VMWare’s capabilities “up the stack”, then they will be well-positioned to capitalize on this coming wave of developers migrating their applications to a cloud platform. Do you agree? I welcome your thoughts.