Peter Christy

What Does that Cloud Look Like to You?

It turned out to be a very Cloud-themed, meeting-oriented 3 weeks for me. First of all, Cisco, EMC and VMware (with a little help from Intel) announced a set of cooperative efforts aimed at accelerating the uptake of large-scale, virtualized computing. Then EMC had an Industry Analyst Meeting that was very much focused on the "Private Cloud." Finally, at Microsoft's PDC in Los Angeles, Microsoft's Cloud services (Azure) was at the center of much of the discussion. I'll talk about each of these briefly in this issue. They each see the Cloud through their own unique lens. Each is quite interesting (IMHO).

EMC and the Private Cloud

I missed last year's EMC Industry Analyst meeting, and this one was completely different from 2007. Two years ago it felt like EMC was a manufacturer of complex storage boxes that dabbled in software and applications as needed. In 2009, from Joe Tucci down, EMC seems very much like a systems company, and one that understands (at least strategically) that software is very much a key part of its future. To be sure, EMC is still very much a storage company and makes money by selling boxes. But a modern storage system is mostly complex software, and (it turns out) sometime in the past EMC moved to using X86 instead of alternatives like MIPS or PowerPC. In a virtualized world, it may well make more sense to package storage functionality as virtual machines than as boxes in the infrastructure. In fact, Chad Sakac, EMC's VP of VMware Technology Alliances, in his demo showed off some virtual Clarion systems running on the virtual storage fabric which is much easier (still not simple) if it is already running on an X86 platform. In his presentation, Brian Gallagher, the head of the Symmetrix business, let drop that he had 1,600 engineers, all software engineers (and all writing X86 code to boot). EMC seems to clearly understand that they need to make their customers succeed with storage for virtualized environments, even if that journey includes selling more software than hardware and aggressively bringing the cost of storage down. In the last year EMC introduced FAST (fully automated storage tiering) that lets flash and SATA drives be mingled with the traditional expensive drives in a Symmetrix system. The results can be dramatic: cheaper, faster and lower power. Next EMC has their sights set on what they call storage "federation" — how many block storage systems can be linked within data centers and across data centers including capabilities like storage motion to match application motion, and FAST and storage federation are essentially just software. Managing data within and across public and private clouds is certainly a key part of the vision of utility and cloud computing. In contrast to Cisco with significant businesses in both enterprise and service providers, EMC's bread is very much buttered by Enterprise sales, and their current focus is very much on the journey to the "private" cloud — the use of virtualization to evolve enterprise data centers to utility, self-service computing. I personally agree that this is the most exciting story (not the public cloud). EMC seems to have set the stage for a remarkable metamorphosis from the big box past to something quite different. Of course setting the stage and giving the performance are quite different, so only time will tell.

Autonomic Computing (Redoux)

In the last issue I wrote about how VMware's acquisition of SpringSource gives them a way to interact directly with application developers, and how valuable a framework may turn out to be in Cloud and Utility computing as a way of solving important infrastructure optimization issues (like edge acceleration) without requiring the application developer to understand all the nasty bits. VMware has something even more important in mind I think: what their Analyst Day talks hinted at was that they thought that applications built on the framework could be instrumented and automatically managed as a result. If you remember, IBM tried "Autonomic" computing some time ago, and couldn't really make it work. It seems pretty clear that an executable binary is much too cryptic to make sense of at run time for broader management purposes. Things improve with virtualization because it's much easier to see what the application is doing (if not to understand it). With a framework, things improve still more because you see more of the program structure at the framework abstraction layer, and as VMware notes, you can instrument the framework to give you more useful information about what the program is doing. Are these changes enough to make something like autonomic computing really work? Remember, the goal is a smart data center than automatically dynamically adjusts to accommodate evolving load, hardware and software failure, issues of scaling, ... Our friends at Microsoft would argue I think that this isn't going to work for VMware. Microsoft thinks that by the time you get an executable program all the interesting information about the program (the design intent, how it performs and scales, how you can tell it's in trouble) has been removed or encrypted beyond all hope. Microsoft's approach (Dynamic IT) is based on the idea of capturing and preserving key design and operational data in a declarative form, and using all this in the data center to achieve the goal. Virtualization changes the game enough to make VMware's direction worth looking into, but if you asked me now, I wouldn't bet the rent money they can make it work anymore than IBM could.

Service in the Cloud

On the heels of VMworld, SalesForce held a delightfully upscale introduction luncheon to announce the Service Cloud 2. Sales force automation — SalesForce's primary offering — is a key pillar of CRM, but CRM is a much broader topic and customer service is another important pillar. To date, the customer service part has been typically done with in-house systems that organize problem/solution knowledge bases that are then used by call center agents. Enabling customer self-service is a less-expensive solution, and done well a fundamentally better solution, especially when augmented by community support where users help each other. But of course customer self service and community support are outward facing applications with scaling and distribution aspects that make them perfect Cloud candidates. About a year ago, SF acquired InStranet, a knowledgebase technology company (a key underpinning of all forms of customer service) and they have been integrating and extending that capability since. The SF Service offering is still only a small part of SF revenues but is growing much more rapidly that the core SFA business. This most recent announcement incorporated innovative Facebook and Twitter integration. For example, a vendor can set up a Facebook page and using the SF system, service questions can be posted entirely within Facebook, but fully integrated with the other service delivery channels. It's going to be interesting to watch how this catches on. It's fair to say that SF is way ahead of the traditional application vendors in offering a SaaS/Cloud service. For companies wanting to be on the innovative side of support (self-service and community-service early adopters) but not wanting to operate it themselves, the SF solution is pretty compelling.

Virtualization and Security (Redoux)

Up until now, VMware treated security by saying as little as possible about it, knowing that arguing that virtualization didn't open up new security surface area just gave credibility to the thought that maybe it did (let sleeping dogs lie) which could only impede sales. Now that VMware is looking beyond the individual virtual machine, and thinking about the Cloud, they have started to talk (credibly) about how virtualization actually helps security. The issues are ultimately complicated (thank goodness there are bright analysts to think about them J ) but here's the basic story line. (1) If you want to talk seriously about the Cloud or even utility computing you need to talk about entire applications — virtual applications or Vapps — that include all the bits needed to get the job done, not about single virtual machines. (2) In order to deploy a Vapp to the Cloud (or even to a private utility or DR data center) you need to specify how it is to be protected. At first blush, a good answer is "just like you did before" except now the security can be largely provided by virtual appliances rather than the physical appliances used previously. (3)When specified on a Vapp by Vapp basis, and implemented by virtual appliances, the security can be as specific and customized as desired. For example, you can think about offering data leakage protection (DLP) tailored to the function and specifics of the application. You need to bind these security specifications in with the other application details so you know what else needs to be provisioned when the application is provisioned and you know where you can and cannot run the application (e.g., what other applications you can co-mingle with). Where it makes sense, part or all of the security can be specified parametrically (e.g., "PCI protection") so that as the regulations change, the specifics of the protection change automatically. VMW's observation is that binding security to the application and assuring it is provisioned as specified is a lot better than what we have now, where in most cases this is partially a manual process, and often requires coordination between multiple teams, all of which is complex, expensive, and prone to errors. The automation also makes it possible (and reasonable) to be much more specific about each application which is also good. Thus, virtualization makes security easier and better.

The Cloud (Dessert Topping or Floor Cleaner)?

 John and I have done a lot of interviews with Cloud "experts" over the last few months (this is the fun part of our business, hunting down and talking to really bright people). On the one hand, a lot of the analysis bandied about is malarkey (to be polite). Enterprises won't jump into public clouds to save buckets of money any more than we've all given up our cars to drive Hertz Rent-a-Cars instead (shouldn't they be a lot cheaper for the same reasons, huh?). On the other hand, we've been talking to many of the most successful ventures using Amazon (primarily, and other Cloud providers) and the stories are very interesting. There seem to be two important takeaways for me: (1) The cloud is software gurosity unleashed. A team of brilliant programmers and system architects can make incredible things quickly when unfettered from the nasty bits of buying and running the hardware infrastructure; (2) For speculative ventures, it's hard to beat the economics of pay-as-you-go. VC's like the Cloud because they don't have to by capital equipment with great impact during the early stages of the venture in terms of getting real product development bang for the buck. But it's also clear from these stories that the same benefits carry forward into real service operation especially if the success of the service is unpredictable, especially on the upside. One more time, we seem to be collectively overestimating the short term impact but probably underestimating the long term impact.

Google Wave

The announcement at Google I/O that made the biggest waves was Wave — a provocatively interesting collaboration tool. Wave consists of a Cloud component — the distributed infrastructure that enables a worldwide set of collaborators to share a structured document (the "Wave") and a client component that actually provides the collaboration tool. It's really important to understand that Wave is not a demonstration why the Cloud obviates the need for a PC. The Wave client is implemented as a "browser" application (part of the Windows/HTML5  debate discussed briefly in the last issue). Wave wouldn't be interesting without the part that runs on your laptop. The most interesting criticism of Wave has come from Ray Ozzie. His comments at the Churchill Club were very thoughtful — if anyone appreciates a clever collaboration application it's Ray since he's been building them since college. His point is for most purposes Wave is overkill and much more complex than needed, comparing Wave to Live Mesh and its precursors Groove and Lotus Notes. We run our business on Groove, and Ray is right — for 99% of what we do Groove is sufficient.  In Live Mesh, Ray has pared down the complexity of Groove enormously because it assumes that the heavy lifting is done by Cloud services (permitting the client footprint to be smaller). To implement Microsoft's desktop/phone/Cloud integrated vision Live Mesh is much more suitable than Wave, no doubt. And there is every reason to believe that the Cloud part of Wave could be implemented in Azure and the client component in Windows far easier than in HTML5. This does not distract from the fact that Wave is one of the best examples of a "killer" Cloud app — something that will deliver real value (Ozzie's question is the breadth of the use cases compared to those that Groove or Live Mesh enable) and that absolutely could not have been developed without the integrating Cloud collaboration services.

Cloud Wisdom

The Structure 09 panel of successful Cloud practitioners (all the right places and people) had some real insight. Last year the key takeaway was scalability — if you didn't know how you were going to scale it would certainly kill you if you started to succeed. The wisdom this year was at a higher level, and the first part might be summarized by (a) think at the system level, (b) think probabilistically (bad things happen rarely but frequency with large numbers) and (c) it's really useful to own your own stack and be able to make everything work well together (this all speaks well to the Cisco UCS value proposition but that's a longer story). The second part was that humans are still the key to effective solutions and the ultimate cause of most problems (that was the Google perspective after 10 years and maniacal focus on tools and automation!). The broader wisdom at the meeting was that it really is mostly about virtualization (as a key enabler of pre-existing concepts) and utility computing, and it all makes sense if we stop elevating "Cloud" to something new and magical (as we here at IRG have concluded as well).

The Cloud for the Rest of the World

One of the most interesting talks at Structure 09 was by our old IBM friend Willy Chiu. We first met Willy when he did performance work related to WebSphere and then talked to him as he built their high-performance test labs to help customers with large Web applications. For the last few years he has been driving the construction and use of large Cloud labs around the world.  What Willy talked about at S09 was some of the really interesting use cases he's discovered "elsewhere" (not the  high-tech, geeky focus that dominates much of the Cloud dialog). These are really interesting. Suppose you are a country like Vietnam that wants to invest in industry development. The Cloud (and we would add wireless communication) is complete game changing. A government or economic development agency can put in a modest (by Google or Amazon comparison for sure) Cloud data center and quickly create on-demand infrastructure to use (this is analogous to the use of Amazon by startups today except startups are a very small part of the US IT economy but could be the dominating part of a development economy).  This not only provides a lot of painless infrastructure, removing from entrepreneurs the task of buying, installing, connecting and running their own computers, but also lets the government provide standard solutions or stacks further reducing the IT knowledge needed to use IT. Maybe we've all been looking in the wrong place for where the Cloud will take off. The implications are broad. If this vision is right then the issue isn't the trickiness needed to enable the construction of the next Twitter, Facebook or Wave — the focus is how to make the basics really easy to use for what we would think of as completely boring (but certainly green field) applications.

Clouds and CDN's

As long time CDN followers we're fascinated by the question of how to get the best performance out of Cloud applications. This certainly includes incorporating traditional CDN capabilities (such as edge caching) but the more interesting question to us is what a higher level platform (Azure for example) can do to achieve the performance without requiring the programmer to understand the issues or CDN's. We know the Azure team is thinking about this (MS has built out an internal CDN for many separate reasons) but so far they haven't given many details. At Structure 09, CDN's were often discussed, but in a somewhat confusing way. Paul Sagan, President and CEO of Akamai, gave one of the more coherent business presentations focusing on the high level benefits of Cloud computing (e.g., energy efficiency) and then the role CDN's must play making Cloud applications high performance.  Many of the big Cloud guys talked about CDN's too, but what they said was that CDN's had become commodity technology, and although they all believed in the value of edge caching (and internal caching for that matter) most just rolled their own and/or happily played the leading vendors off against each other. We think that the question becomes simpler and clearer if one views "private Clouds" as what people are doing with virtualization in private, consolidated data centers. For these many customers rolling your own CDN isn't an option, and we think that these utility data centers will be more important than public clouds (economically) for a long time to come.