So now we know what Paul Maritz’s real job is going to be. VMware has announced on its Executive Team Blog that they will form a new organization, The Pivotal Initiative, which will collect EMC’s and VMware’s application focused people and assets into one group. This will include the people and the technology assets from EMC’s Greenplum and Pivotal Labs organizations, and the people and technology assets from the VMware vFabric (Spring and Gemfire), Cloud Foundry and Cetas organizations.
The Significance of the Name “The Pivotal Initiative”
It is extremely significant that this new entity is named after one of the components of the entity that has not gotten a lot of attention except from the customers for whom they have done outstanding work. Pivotal Labs is a contract software development company that uses “extreme programming” and Agile Development techniques along with some unique processes (like programmers always working in pairs) to make an enormous amount of progress on a product or a project in a very short period of time.
This notion of solving a problem in software quickly gets at the core of one of the most important issues facing the entire IT industry—both the vendors who provide solutions, and the companies who consume solutions. That issue is that the demand for business functionality implemented in software is infinite. The demands to change it as soon as implemented are inexorable. And the economic requirements to address these two issues with more readily available and less expensive developers represent life or death issues for many companies. By way of example, if Target cannot compete with Amazon.com in the online retail space, then Target will simply cease to exist—and based upon Target’s emphasis on its online operations Target is clearly well aware of this fact.
If we assume that the smart people at EMC and VMware picked the name of the company for the reason listed above, we can then analyze the remaining components in the light of a strategy designed to reinvent how applications are built, deployed, and managed in production.
VMware and an Applications API Strategy
It is important to understand that the key executives who have been and are now in charge of VMware have deep roots in the development of the “WinTel” ecosystem. This ecosystem is based upon the concept of horizontal layering of value to customers. Intel owns the chip layer (VMware’s new CEO Pat Gelsinger understands this quite well). An OS vendor (Microsoft, Google, Red Hat, VMware, etc, owns the layer of software that abstracts the hardware to the operating system and the applications (Paul Maritz understands this quite well). These same vendors own the layer of scheduling resources across competing workloads. The key innovation that VMware provided is that vSphere (actually ESXi) took over the role of abstracting the workloads from the hardware and scheduling the hardware resources in the hypervisor instead of having the OS perform that task.
But there was one key task that the vSphere platform was never able to capture, which was the function of providing the interface to application developers and to application development tools for next generation applications. Windows succeeded because it was able to provide abstraction from hardware, resource scheduling , and an API for developers in one product. VMware vsSphere succeeded at the first two of these task creating a tremendous threat to Microsoft and Red Hat in the process, but failed to present a compelling alternative application development story to developers.
The Role of Cloud Foundry and vFabric
With vFabric, VMware has a “better Tomcat than Tomcat.” This was a first and very important stake in a developer API war with the entrenched J2EE crowd (Oracle with WebLogic (BEA) and IBM with WebSphere). Cloud Foundry is an entirely different initiative. Cloud Foundry is about taking a clean sheet paper to the questions of how to build a next generation application platform, evolve that platform over time (with some sort of a managed open source model), and make money in the process (something that VMware has not either figured out or announced as of yet).
The Gemfire, Cetas and Greenplum Assets
It is clear that some significant portion of new highly scalable applications are being built around in-memory database architectures like Gemfire and big data architectures like Greenplum. This is where things get confusing. It is clear that some applications need in memory data management. It is clear that some need massively parallel databases and their associated analytics as provided by Greenplum. Cetas adds another wrinkle to the picture with a class of analytics beyond what vendors like Netuitive (who invented self-learning performance analytics, Integrien who was acquired by VMware and subsequently integrated into vCenter Operations, and ), Prelert provide. What is not clear is what percentage of the applications that have not been built yet require these advanced capabilities.
Back to the Economics of Application Development and Maintenance
At VMworld 2012 in San Franciso, Paul Maritz said, “The story of the Software Defined Data Center has been written. We are not going to be the only vendor of a SDDC, but all of us who are going to provide one, understand how this is going to end.” Paul Maritz then said, “The story of how application development, deployment and operations is going to be reinvented is just as compelling, just as important, but has not been written. The participants in the creation of this story do not know how it will end.”
The thesis of this post is that VMware and EMC have decided to combine technology assets, people and money in order to pursue a reinvention of how applications are built, deployed and operated in production. Paul Martiz presided over a revolution in the system software business while at Microsoft driven by third party adoption of a platform. The thesis now is that the platform can be abstracted from the cloud, from the virtualization layer, and from the operating system. At the end of the day, everything is all about application functionality delivered to the business in the fastest and most cost effective manner.
The Pivotal Initiative is is a bold bet to reinvent how applications are built and upon what platforms they run. While VMware has revolutionized data center operations, and public cloud vendors like Amazon, Microsoft, Google and the OpenStack cabal have revolutionized Infrastructure as a Service upon demand, the future of how applications are built, deployed and managed in production remains unresolved. Maritz and Pivotal are putting a bold stake in the ground as to one way that this will turn out. The value that the IT industry brings to the business hangs in the outcome of this battle.