The biggest requirement for expanding VMware is improving the storage subsystem and reducing the storage tax. EMC and VMware opened the kimono on the future direction of VMware at VMworld 2011. There are some good ideas, but Wikibon believes that the business approach to implementing these ideas could lead to a stifling of innovation.
Wikibon understands that the basic idea is to create a storage container including both the data, the metadata, and the policy data. Logical channels connect the storage to the external storage ports (think IBM System z multiple logical channel subsystem, SMS and VTOCs, plus object storage). Wikibon understands that VMware plans to work with the normal storage partners (Dell, EMC, Hewlett Packard, IBM, and NetApp) to provide APIs to help these traditional storage vendors add value, for example by optimizing the placement of storage on the disks. There is a three-to four-year roadmap to this world of storage without LUNs.
So why not celebrate a world without LUNs? Wikibon believes this business model is in part influenced by EMC and other large storage suppliers to preserve current product models for as long as possible. These large suppliers form a type of storage 'cartel' where a very few players control the direction of storage solutions for VMware. Our concern is this model presents risks to innovation that are counter to the approach normally taken by independent software vendors. Specifically, ISVs are normally storage agnostic and are consistently trying to lower the cost of storage solutions-- to the point of commoditizinig storage where possible (e.g. Microsoft's stance on SANs for Exchange). The inclusion of EMC and other members of the storage cartel in the formative stages of critical API specs (and specific exclusion of smaller players) is justified by VMware as a way to reduce the engineering overhead of supporting many smaller companies. While we recognize VMware does not have unlimited resources, we feel it is responsible of us to ask the question: "Is VMware and its owner, along with what's become a storage oligopoly, stifling innovation?"
Activities at VMworld 2011 underscore that storage innovation is coming at a faster pace than at any time in the last decade. Flash-only vendors such as SolidFire, Nimbus, Pure Storage, Texas Memory, Whiptail, and many more are providing better and more consistent performance at equivalent price points to the traditional disk-based storage vendors. Vendors such as Fusion-io are providing new architectures of memory as an extension of RAM storage, and with it new application opportunities. Coraid are innovating with new protocols such as AoE (ATA over Ethernet). Vendors such as Cleversafe are introducing erasure coding solutions that are an order-of-magnitude cheaper for storage archiving solutions. A myriad of software vendors are providing capabilities to manage storage better, for example by taking an application view of performance, and storage as a component. The contributions from the established vendors are important, but they are incremental improvements to the status quo. The question we are asking is could VMware be doing a better job of including these new storage innovations early on in the cycle and thereby accelerate the pace of business value delivery to its customers?
It's a question that no doubt will be met with attacks and denials from VMware, EMC and the balance of the storage cartel. Indeed, we have no information that suggests EMC is 'stacking the deck' in its favor and the company's largest competitors have always stated that VMware treats them as an equal to EMC. Moreover, we've always believed that EMC wouldn't purposely sub-optimize VMware's market value in an attempt to gain advantage over competitors. However as an independent observer witnessing an explosion of storage innovation, we feel it is necessary to raise the question and catalyze discussion on the issue. Specifically, if VMware were not owned by a storage company would it be more aggressive about accelerating disruptive storage innovations? Is VMware influenced by its owner in a way that favors the traditional storage model and is this the best outcome for VMware customers? If VMware is going to provide the best storage subsystem of the future, it needs to work with the innovators in the storage industry, and skate to where the puck is going to be. If it doesn't, we believe it will, over time, open the door to alternative platforms and weaken the VMware ecosystem.
Action Item: Shareholders and VMware customers should tell VMware that they expect new storage systems to utilize the best storage technology and ideas available and not bake the past into future designs. This will also drive the traditional storage vendors to increase their rate of innovation and ultimately result in greater business value for customers.
Footnotes: