EMC Navigates the IT Transformation Waters

IT is always changing, but the breadth and depth of the transformation now seems to be occurring at an especially rapid pace. IT professionals have to be rooted in the here and now, but they also have to understand how those changes will continue to affect them, and what actions they need to take.

Events such as last week's EMC World 2013, where the theme was "Lead Your Transformation" along the lines of cloud, big data and trust (security), are designed to help IT professionals better understand both what they need to do in the present and the near future.

Let's examine EMC itself briefly and then touch upon two of the most important technology directions featured at EMC World.

EMC as a Company and a Federation

With $21.7 billion in revenue in 2012 and about 60,000 employees, EMC is obviously an important IT vendor. Although its storage heritage is and will continue to be a pillar of EMC proper, much of the value added to the company directly by storage comes primarily from software. Of the 12,000 or so engineers employed by EMC, only about 500 are hardware engineers. The rest are software engineers.

EMC views research and development as an engine of growth. Joe Tucci, chairman and CEO of EMC, stated that about 12% of company revenues are spent annually on organic (that is, internal) R&D and about 10% (depending upon opportunities) on technology acquisitions. Tucci pointed out that the phrase that EMC likes to use for these acquisitions is "string of pearls," and that acquisitions are determined according to what unique or valuable technologies the acquired company brings to the table, rather than acquiring market share.

Now, EMC is not perfect in its acquisitions, but I suspect that many other IT vendors are envious of its batting average. The R&D mix balances the need to move current products forward and the need to bring bright ideas that enable EMC to fill current gaps in its product line or to move into emerging product adjacencies where EMC wants to establish a presence. For most of the last decade or so, that strategy has proved successful.

One of the outcomes of the acquisition process has been EMC's organizational structure. Now, some product acquisitions have been absorbed within the company structure and the original company name abandoned (Legato, for example), even though the technology persists virtually as it was, but that is not always true. For example, RSA has retained its name, as the RSA brand has a high recognition factor among security-focused customers.

There's another way to look at the EMC "family," and that is through the use of the word "federation." During a Q&A session with analysts, Tucci pointed out that EMC is the federation of three companies--VMware, EMC (notably EMC Information Infrastructure) and the recently announced Pivotal (a partnership consisting of EMC and VMware, with a minority share held by General Electric).

Each is a separate legal business entity with the common thread that EMC holds the major equity stake in both VMware and Pivotal. The most important thing to remember in this federation model is that each entity retains control of its own internal affairs. VMware is the poster child for this concept. EMC recognized that VMware's success depended upon its system vendor partners, many of whom are direct competitors of EMC. If EMC had tried to take unfair advantage (from its competitors' perspective) of its ownership of VMware, then they would have felt compelled to find alternatives. VMware has demonstrated its own independence so it has not had any significant defections (and one can bet that EMC has not always felt that decisions that VMware made were in its own best interests).

EMC's Pivotal Decision

The creation of Pivotal as a separate legal entity is an attempt by EMC to reprise its success with VMware. What does Pivotal plan to do? A lot of discussions at EMC World focused on the need for a new platform that enables data fabrics instead of the traditional application fabrics. The new Pivotal One platform as a service (PaaS), which acts like an overarching OS across multiple compute platforms, is the approach EMC intends to pursue.

However, describing exactly what Pivotal is and will be is something that defies a 30-second elevator pitch. With some trepidation, as the phrase has been overused (but no one from EMC or Pivotal brought it up), Pivotal is attacking "the next big thing." The industrial Internet (that is, the Internet of things), big data, fast data, advanced analytics and extending the lessons of the megacloud (think Amazon AWS products or the "consumer grade" Internet) are all simply parts of the whirlwind of areas that Pivotal will tackle.

And that is simply too big for a single company to tackle, even one as large as EMC. The creation of the new entity invites others (including, as with VMware, EMC's direct competitors) to participate. And that requires an independent entity that acts in its best interests (with VMware as the role model).

Pivotal starts out as a company with 1250 employees and $900 million in investment, so it is a significant company from its birth (what a big baby!). And that start-up funding includes $110 million from GE. Quite frankly, EMC and its federated partner VMware did not need the money GE is contributing, but that makes GE a true "blood brother" committed to the cause.

More importantly, the presence of GE in Pivotal is vital. GE is not a stodgy industrial giant, but rather a large enterprise that is in the process of reinventing itself to meet the challenges of the industrial Internet. See its "brilliant machines" work with healthcare as just one example (hopefully, Jack Welch is smiling). Although much progress has been made, GE seems to understand the magnitude of the challenges and the problems that remain. It needs Pivotal to help address those issues, and Pivotal benefits from the real problems that it can tackle.

Moreover, GE is a magnet. Other companies that can benefit from the industrial Internet as well as IT vendors that want a piece of GE's business will be attracted. The model for everyone is that a small slice of a very large pie is better than a large slice of an apple turnover. Now, is all that apple pie and motherhood? Well, maybe, as Pivotal faces a lot of daunting challenges, and a string of incremental successes may not yield the critical mass that is necessary for transformation to a true "next big thing." However, IT professionals need to pay attention to Pivotal and take into account the progress that is made over time.

One final point on Pivotal: The assets of Pivotal initially represent a good use of CEO Tucci's "string of pearls" acquisition strategy. EMC seeded Pivotal with assets, including Greenplum (such as its Hadoop distribution) and Pivotal Labs (agile software development consulting), while VMware contributed vFabric (including the Spring application architecture framework) and Gemfire (in-memory analytics), as well as Cloud Foundry (open platform as a service for clouds) and Cetas (real-time streaming analytics).

Next page: Striking Software-Defined Storage with ViPRNow, where Pivotal is composed of companies that EMC and VMware acquired, EMC's new ViPR (which stands for virtualization platform reinvented) represents in-house organic development of some two years. With ViPR, EMC is coiled to strike at the software-defined storage space.

Software-defined storage is about using host-based software to provide an overarching management control point for multiple, heterogeneous storage arrays for provisioning, virtualization and other services. The announcement is a pre-release, as ViPR's initial general launch will not be until later in 2013 and extended features are not expected until late 2014.

The first release will support EMC's VMAX, VNX, VPLEX, RecoverPoint, Isilon and Atmos storage, as well as NetApp arrays. Thus, the first release will support just one third party vendor and no commodity hardware. So, why is EMC announcing this early?

Note that while this product will be used by enterprise customers, the target audience for ViPR is service providers that can now include ViPR in their long-term planning process, even if specific capabilities are not yet available. Service providers may also provide input on the ViPR development process and priorities, as well.

Consider, also, that whoever sells software-defined storage software has the virtual high ground (king of the mountain) for managing storage. That does not preclude choice of particular arrays, but puts the SDS vendor--in this case, EMC--in a stronger point of influence.

A key architectural distinction in ViPR is the separation of the control plane from the data plane. The control plane is where the commands for managing the storage infrastructure travel and is separate from the data plane where bits and bytes of information travel. That is an important distinction. Storage controllers support valuable software-enabled management services, such as back-up snapshots and data mirroring. ViPR does not interfere with those services, or critical features like SRDF (Synchronous Remote Data Facility), which have the robustness and reliability that enterprises require.

Of course, EMC wants to preserve its own controller-based storage array architectures. Those would be threatened if a competing SDS with its own storage management software could be used with commodity storage hardware without the need for sophisticated controllers. While such solutions could be cost effective and good enough for many use cases, in practice, many large enterprises are risk adverse (making IT question whether potential cost savings are worth the loss of their jobs if or when something goes wrong).

So in the case of ViPR, continue to watch the space.

Mesabi Musings

That IT organizations are continuing to transform themselves is not debatable, but questions regarding what, where, how and when those changes arise are commonplace. As we saw at EMC World 2013, the company is attacking the problem and enabling change through both an attack on the "next big thing" (Pivotal), which is a revolutionary; and dealing more effectively in the storage management world (ViPR), which is evolutionary.

Pivotal represents the new data-driven world view, and while some of its technologies are available today (such as the Greenplum Hadoop products), Pivotal largely represents an initiative whose impact will be in the future. In a sense, EMC's ViPR represents a new view of the current world of storage that is being extended into and enhanced by new development, such as SDS, and will hopefully be an important component of the software-defined data center.

IT professionals have to continue to extend their knowledge and understanding of what ViPR will offer, but will have to learn to be prepared to include the new world view of Pivotal as appropriate. EMC is indeed strongly navigating the transformation waters of IT, and everyone needs to pay attention.

EMC is a client of David Hill and the Mesabi Group.