Jeff Jaffe’s Blog

In the last couple of posts I described some very high level thinking that I presented at Infoworld’s Executive Forum on Virtualization in September. The main points were:

• Virtualization is much broader than an I/T thought – it is the fundamental driver of business
• Implicitly, this is the main point in Friedman’s book – “The World is Flat”; better titled “The World is Virtual.”
• That his book is not about business, but about technology. Technology is the fundamental driver behind the virtual world concept.

I believe that Open Source methodologies are the accelerant to this key driver. Actually Open Source is already referenced in Friedman’s book. Specifically, one of Friedman’s ten “flatteners” is “Open Sourcing”. But Open Source will be a larger driver in the future.

In the last posting, I referred to some 30 new technologies that are driving this flattening according to Friedman. One way to look at these technologies is to divide them into three categories. One category is “Rapid Growth of Technology”. In the 1960s-1980s that was where technology action was. Expensive investments by large companies in fabs created exponential improvement in processor speed, memory densities, storage, communications bandwidth, and many associated technologies. A second category is “Point Innovations”. These are the breakthrough ideas that can come from anywhere: university labs, individual inventors, government funding. The third category are inventions that improve communications or “Sharing”.

In the figure below, I have mapped the 30 technologies into these three categories. The interesting observation is how the vast preponderance of them relate to sharing. While the research action in the 1960s-1980s properly belonged in the large fab, the more recent research environment – the one driving the virtualization of the world – takes place where sharing is most natural.

Some of these innovations have taken place in the Open Source community, as indicated by Friedman in his one chapter on Open Source. In fact, most of the recent ones have taken place in Open Source. The primary reason that some sharing innovations did not take place in Open Source (e.g. email) is that they were too early. The important bottom line is that we should expect future sharing innovations to take place in Open Source, now that this methodology has matured.

Why is the Open Source methodology the most natural one for introducing new technologies related to sharing? There are five reasons for this:

1. Sharing technologies are best developed by diverse teams that see the world differently, but collaborate. By bringing diverse views to the table, the solution gets a level of completion that it would not get if it were done by one person or one company.
2. A unique aspect of collaboration technologies are that for them to work they must be introduced into multiple systems. This is at all levels. Every different user workstation or hand-held must support the innovation. Often, every network device must support the innovation. In the data center, each compute architecture, switch, or storage system must implement the innovation. This happens much faster – and with greater consistency, if the code is shared. Open Source.
3. Even within a single system, often the innovation takes place in multiple places. Different subsystems, different layers of a software stack might all need to implement the same capability. Again, best if the code is shared. Open Source.
4. Open Source is great to introduce sharing because it is great to introduce all new technologies (cf. my August 21 post for the reasons).
5. Similar to point 2 above, Open Source generally drives standard acceptance (also note the aforementioned blog entry).

Last posting, I introduced the topic of my Virtualization Executive Forum keynote: The World is Virtual.This is a direct take-off from Thomas Friedman’s book, The World is Flat. Friedman discusses the ten forces that have flattened the world: (1) Berlin Wall and Windows, (2) Connectivity, (3) Workflow Software, (4) Open Sourcing, (5) Outsourcing, (6) Off-shoring, (7) Supply Chaining, (8) In-sourcing, (9) In-forming, and (10) The Steroids – Digital, Mobile, Personal, Virtual.

A good book and a good read. These ten forces allow Friedman to provide many vignettes about how business models are changing. The sourcing of a company’s efforts – inside the company versus outside the company – is changing. Globalization is a major factor. Three billion people – many well-trained – are coming in to the world economy. Horizontalization is a factor: you can divide tasks into finer granularities and handle each fine-grained task in a different place – optimized for cost, proximity, or quality.

What’s the common thread? Friedman says that the common thread is that the world is flat. By this, I believe he means that the world is getting smaller (although my math background tells me that you can have very large flat surfaces and very small spherical surfaces – but that’s a detail). To me, a more apt metaphor is that the world is virtual. The computing and communications technologies cited in the various chapters in Friedman’s book – and the granular grinding of business steps into independent movable processes – means that the physical world no longer has bearing on business and policy makers. The world has become virtual.

This virtualization – at the macroscopic business level – has a direct metaphor to the computer virtualization that we are talking about at the I/T level. In both cases there are multiple places (physical locations, processors, CPUs, operating systems, etc.) where workload can get done. In the virtual world a decision maker (via a management process, a workflow engine, a hypervisor, workload management software) is assigning tasks to places of execution. Criteria include cost (labor costs, CPU costs, management costs) or capabilities (skilled labor force, unique software available). Business virtualization and I/T virtualization are directly analogous.

It is interesting that Friedman apparently (subconsciously) realized this. His tenth “flattener” is what he calls the Steroids. And his fourth steroid is “Virtual”. Somehow he recognized that he was talking about virtualization. But he was enamored with his Flat Earth slogan. So he couldn’t bring himself to properly title his book.

Much as the book purports to talk about business, government policy, and the impact on people, it is really a book about technology. Each chapter, in describing some aspect of the increasingly virtualized world, introduces a large collection of technologies which drive the world towards the increased virtualization. Here is a subset: Windows desktop, world wide web, HTTP, browsers, HTML, fiber optics, Internet, Email, XML, Paypal, AJAX, Web 2.0, Business Web, Web App servers, Wikipedia, Firefox, Linux, Blogging, Y2K, Lean manufacturing, ERP, MRP, RFID, GPS, distributed computing, wireless, optimization algorithms, search, TiVo, Digital, Mobile, personal, virtual. Sure feels like a book about technology.

Let’s develop this further. The world is going virtual. The underpinnings of this change are the rapid development of some 30 new technologies. Why is this happening now? What is driving behavior in the technology space to create these new technologies? Important to know – because it will drive the future changes.

My conclusion is that Open Source is the fundamental methodology that is increasingly responsible for the technology change driving virtualization. How do I come to that conclusion? That will be in the next post!

Last week Infoworld sponsored an Executive Forum on virtualization. A few observations about the event, and I want to share the highlights of my keynote address.

The event

There was a discernible feeling that Xen based virtualization had arrived as one of the key virtualization technologies (cf. my posting of August 2). While there was debate about how it compared to older technologies, the momentum was clear. I would say that the timing of this first Executive Forum for virtualization was driven by the maturity of Xen, evidenced by its inclusion in SUSE Linux Enterprise 10.

There was great sponsorship for the program: EMC, Intel, Polyserve, Red Hat, VMWare, Data Synapse, Egenera, Emulex, IBM, Kingston Technology, Microsoft, and of course Novell.

Four keynotes, and roughly 50 presenters, made sure that every aspect of this issue was addressed. The bottom line: virtualization technology is transforming the data center. Consolidation and cost reduction are taking place. Virtualization is disruptive.

The making of a keynote address

Having been invited to prepare a keynote address, I spent some time thinking about what the audience would find interesting. Based on the agenda, everything was already covered! Virtualization, SOA, detailed engineering design, customer views of the impact on the data center, hardware enablement, testing, planning, and deployment. What could I possibly add?

When I attend a conference, I expect that the sessions will take me through the details of the subject matter. But I expect something different from keynotes. Broadening. Perspective. A point of view. What was the broadest possible picture I could paint for Virtualization?

Four major themes

Running multiple virtual machines on a single computer is a technical thought, and not a particularly original one. After all, virtual machine architectures were invented more than 30 years ago. But the broader impact of virtualization is huge.

Thomas Friedman, noted author and global affairs analyst, wrote a best selling book, The World is Flat, about the major trend in business, government policy, and the way that work gets done in the 21^st century. My first theme was that his book title – while playful (a double entendre hearkening back to pre-Galileo days) – was not the most accurate title one could think of. A better title would have been: The World is Virtual.

The second theme of my talk was that the virtualization of the world has less to do with business and more to do with technology. Friedman’s book is actually a book about technology. The third theme – open source methods are critical to the virtualization (or flattening) of the world.

My fourth theme was to relate the broader discussion of the virtualization of the world to what we are accomplishing with I/T virtualization.

I will develop these themes more in my next posting.