Digital Walkabout

CloudCamp 2009

I recently attended CloudCamp 2009, a worthy event, although some were of the opinion that the material in the the event—the CloudCamp is organized in an unconference format—was a bit basic. However, a lot of the questions I heard, and conversations I had, at the event suggested to me that many are still trying to get their heads around the basic concept of cloud computing. As a result, it behooves those of us who work in this space to remember that not everyone is quite so immersed.

Any event of this sort inevitably has a wide number of different simultaneous threads going on. However, here are a few that I thought were worth highlighting:

Is security really an issue? The security aspects of cloud computing are often presented as a matter of trust in service providers and getting comfortable with the loss of direct hands-on control. Those are part of it certainly. But a number of discussions made it clear to me that the situation is a lot more complicated than developing a “comfort level” with the technology. It also goes beyond point products such as data encryption.

Christofer Hoff (who was in one of the unconference sessions I participated in) wrote a post a few days ago that gives a nice window into some of the complexities here. This isn’t to say that security control and compliance concerns are stumbling blocks for moving all applications into a cloud, but they do have to be taken into account (in all their myriad complexity) for many core business applications.

What is cloud interoperability? Interoperability seems to be emerging as a bit of a contentious topic in cloud computing. This is partly because interoperability, like cloud computing itself, means different things to different people. Just about everyone agrees that base-level data portability (download your customer records from a CRM system in a readable format) is a must and that a nirvana of totally transparent computing delivery across providers is years away.

The confusion and argument comes in the vast middle ground.

A couple of things that people at the event seemed to generally agree with:

The higher you go up the stack from Infrastructure-as-a-Service (a la Amazon Web Services), the less interoperability you can or should expect. (And, indeed, the less interoperability IT folks are even necessarily looking for.) In other words, don’t expect to transparently switch from one business intelligence application to another even if you can move and transform your data relatively easily.

A lot of interoperability will be provided not by the service providers themselves (who after all don’t have a particular incentive to make switching away from their service easy) but by third-party “service brokers.” We see a good early example of this sort of thing is RightScale who is essentially mapping the APIs from different IaaS vendors like Amazon and Rackspace to common interfaces in their management software.

Where’s the business opportunity? Finally, in a session led by fellow analyst Judith Hurwitz, we discussed the opportunities for startups in cloud computing. One way to look at the opportunity is to break it down into two broad categories: companies and products such as development or management tooling that make clouds work better and those that are enabled by clouds. The latter includes Software-as-a-Service broadly but it could also refer to things like distributed testing that make use of geographically distributed public cloud providers.

As is the case with the software industry more broadly, there was general agreement that industry-specific products and services often offer better opportunities than big horizontal plays. For example, offerings that deal with data governance issues related to a specific vertical such as the pharmaceutical industry.

One thing there was consensus about is that setting up capital-intensive cloud infrastructure is not typically going to be a good startup play. (In general, I conclude that far more startups are going to be enabled by cloud computing—vis the explosion of mobile applications—than will be involved in creating the platforms that support it.)

Cloud computing covers a lot of territory and there are still enormous differences in how familiar different people and organizations are with the particulars. However, one of my takeaways from this event is that we’re seeing reasonable consensus on the basic taxonomy of cloud computing (even if we argue about terminology like “private cloud”). But, like peeling an onion, now that we’re winding down some definitional and basic value debates, we’re starting to get into more detailed and thornier questions about the particulars.

Rackspace Open-Sources Cloud APIs

"A week after adding API access to its Amazon-esque cloud, Rackspace has now open-sourced these APIs. On Wednesday, the company said it had opened up the specs for both its Cloud Servers APIs and Cloud Files APIs under the Creative Commons Attribution 3.0 license. Meant to encourage the development of open standards for all so-called infrastructure clouds, the move is in stark contrast to Amazon's much-discussed silence over its own API philosophies." 

These were opened up under the MIT license, and they're now available on GitHub here. More at http://www.theregister.co.uk/2009/07/23/rackspace_open_apis/

Microsoft's Azure Gets A Business Model And An Official Release Date

This is an eventful week for Microsoft for Cloud: first, Office 2010, which made a move towards the browser with cloud-based versions of Word, PowerPoint and Excel. Today, Microsoft is announcing the business model and launch timing for Azure, the company’s much hyped CloudOS. The platform also includes a Web-based relational database in Microsoft SQL Azure, and connectivity and interoperability with .NET Services.

Microsoft says that Azure will be offered for purchase through a consumption-based pricing model and will try to continue to offer promotional discounts to enterprise customers. Pricing for Azure’s OS is $0.12 cents an hour for computing and $0.15 cents per Gigabyte per month for storage. SQL Azure will offer a basic $9.99 per month plan and a $99.99 business edition, which has a database capacity of up to ten gigabytes.

Microsoft also announced that Azure will begin to be commercially available at its Professional Developers Conference, which takes place in mid-November. Azure was announced <http://www.crunchgear.com/2008/10/27/microsoft-windows-azure-cloud-os-platform-announced/> last fall, and developers have already been testing the cloud based OS via Microsoft’s Community Technology Preview.

Microsoft chief software architect Ray Ozzie has been saying that Azure will establish Microsoft’s place in the cloud, especially given the dual strength of Microsoft’s platforms and applications and because of its lengthy experience serving both the consumer and enterprise spaces.

Microsoft Azure for Private Cloud: H1 2010

http://www.microsoft.com/virtualization/cloud-computing/privatecloud.mspx

• The Microsoft Private Cloud enables:
•  Management of the data-centre fabric as a single pool of resources
•  Delivery of scalable applications and workloads
•  Focus on the management of the data-centre service and it’s dependencies
•  Federation of services across the full cloud continuum

Microsoft is enabling customers to build the foundation for a private cloud infrastructure using Windows Server and System Center family of products with the Dynamic Data Center Toolkit for Enterprises, availability scheduled in the first half of 2010.

The Dynamic Data Center Toolkit for Enterprises is a free, partner-extensible toolkit that will enable datacenters to dynamically pool, allocate, and manage resources to enable IT as a service.

•          An architectural roadmap, deployment guidance, and best practices
•          Familiar tools that are compatible with existing applications
•          Interoperability with public clouds

With this new effort, Microsoft is keen to maintain a lead in “packaged” or “one-stop” PaaS Cloud Computing, opening the doors to small software companies to build tools(management, monitoring, custom solutions) to accompany the on-premises offerings from Microsoft.

 

-Charlton.

NoSQL: leading Cloud's "NoBah" movement?

One of the remarkable aspects of the emergence of Cloud Computing has been the shift to addressing problems and providing solutions to them, from a focus on technology. The enterprise is still reeling from years of touting technology regardless of what problems needed to be solved. Whilst the industry loves to flog Microsoft, we need to point out other examples of this phenomenon, such as J2EE, which was sold as sledgehammer for every pin, tack, bracket, nail and spike - before innovators such as Rod Johnson started Spring, the MACR gang Flex/Apollo, and others Mule, Curl, etc. arrived to try to shift the balance back to sanity. The result has been a crimson tide of resources who were weaned on the coconut shies of J2EE, selling hammers for problems that couldn't any less be nails, and a current of architects/developers trying to solve real problems with Web/RIA and frameworks, whilst sorting out how to make any way against this tide. 

Cloud Computing has arrived, by virtue of 1) a lack of definition and standards, 2) riding on the coat-tails of Web and RIA, and 3) a backlash against the technology fun fair of the late 90s/early millenium, with a clear focus on solving real business problems. The drum beat of lower TCO/GreenIT has only grown louder with the development of the Cloud. Whilst technology purists may wince at the thought of something so "fluffy" having any real value, the evolution of Cloud from problem-solving, so-called "counter enteprise" (Web and REST, which position themselves as opposite to "enterprise software") movements is clear. These movements are very important, since, at a macro level, they are really pitched against the "every problem is a nail" set which "enterprise software" represents to them. 

These movements provided strong inputs into Cloud Computing, which has, if anything, positioned itself as a KISS approach to solving SMB (and by extension, enteprise) problems As Cloud has been somewhat challenging to pin down. it' has been targetted at scenarios (IT scalability, mobility, dynamic resourcing, adaptive) rather than technologies. Where as the "crimson tide" would attempt to sell you on something nonsensical as, "we're Cloud, just make your EJBs into Web Services," the Cloud and "counter enterprise" movements first ask, "what are you trying to solve?" and then sell their proposals. This focus on the problem/solution than hammer/nail has shifted SMBs and enterprises to a more comprehensive approach to service.

NoSQL is itself a remarkable proponent of this - by asking the question, "when do you *really* need a database?" There are enough scenarios where you do, and a growing number of scenarios where you don't. Flogging a database where you don't need one is madness, and by first asking the right questions as part of its fundamental value proposition, NoSQL may not replace the RDBMS, but is helping to replace an obsession with the latest vendors' offerings and hype with a better application of reason. In order to use specialised data sources, one needs to understand what they were designed to do. Likewise, RDBMSs are designed to model very highly and statically structured data which has been modeled with mathematical precision - data and designs that do not meet these criteria, such as data designed for direct human consumption, lose the advantages of the relational model, and results in poorer maintainability than with less stringent models.

NoSQL - and the problem-solution focus of it and Cloud as a whole - have come as a breath of fresh air, to me as I work to address problems at a strategic and architectural level, and to my colleagues as they work to implement solutions. Our ability to engage both these functions, without having technology shoved down our throats, leaves us free to develop the best and most appropriate solutions to real problems. I, for one, have encountered too many terrible architectures and designs which were driven by the need to adopt or pursue a particular technology. "Just make your EJB into a Web Service" as well as "you need a database" should send any reasonably intelligent architect and developer for the exits. Thus, I use the term "NoBah" - just as NoSQL seeks to challenge whether one needs a database, "NoBah" seeks to challenge whether every problem is a nail ("NoBah comes from the reaction I have to, "Bah, you can do that with Servlets/EJB/RDBMS/CORBA/DCE/EAI/COBOL"; you can likewise perform neurosurgery with a hammer, but it doesn't scale very well, or result in much success, does it?).

What CloudOS means for DesktopOS

The have been some interesting discussions around Google's ChromeOS and what it represents as a CloudOS. Yet, many still perceive ChromeOS as just a thinner for of a DesktopOS. I argue that it represents more than that - it represents an abstraction of the DesktopOS. Rather than pinning applications to a DesktopOS which itself is pinned to HW, the technologies that have evolved into the CloudOS abstract away this away from the applications. Applications can now be written to a higher level of system functionality and are no longer pinned to an OS; they are distributed and assembled componentwise, in a dynamic manner which can now target HW capabilities, but yet are as easily provisioned and deployed as traditional thin-browser based applications. 

The history of Chrome and ChromeOS themselves provides insight into this evolution. Google Chrome represents a new form and model of ‘browser’ (one that takes on the greater use of HW resources presently performed by RIA runtimes) – it is more an integrated runtime stack with both rich and simple rendering and processing capabilities, rather than a traditional thin browser with its traditional web sandbox. Chrome and the RIA runtimes each provide platforms for richer, more immersive/expressive and adaptive experiences than traditional thin browsers, in a way that abstracts (and thus “squeezes”) the desktop OS.  
ChromeOS, in a sense, is attempting to take this type of environment a few steps further by assuming/abstracting additional DesktopOS capabilities. 

The trend that is seen directly pressures the DesktopOS, challenging its relevance. Microsoft has recognised this beforehand and has sought to position Azure (and the client platform pieces of Azure, ‘scattered’ as they are at present across multiple RT bits) as its CloudOS. Each of these CloudOS do the same thing – they spawn Cloud/Web apps as local processes.
Microsoft and other purveyors of DesktopOS will take a hit financially from this. If they, as is Microsoft, act in a timely manner to capture the CloudOS market, they will continue to have a play, but the revenue stream from CloudOS has shaped only to be a fraction of DesktopOS, and doesn’t show any sign or likelihood of rising.

Google Chrome, the next CloudOS

When Google introduced the Chrome browser, a broad discussion about it and about how it could be transformed in the OS of the cloud emerged. The biggest secret that everyone know was realised when Google introduced their plans for Google Chrome OS in their post on the Google Blog . Although based on Linux, it is really a greatly enhanced runtime environment that still abstracts away Traditional OSs. The highlights below talk about Google's consistent vision – the web is the platform . "Google Chrome OS is an open source, lightweight operating system that will initially be targeted at netbooks. Later this year we will open-source its code, and netbooks running Google Chrome OS will be available for consumers in the second half of 2010. Because we're already talking to partners about the project, and we'll soon be working with the open source community, we wanted to share our vision now so everyone understands what we are trying to achieve. Speed, simplicity and security are the key aspects of Google Chrome OS. We're designing the OS to be fast and lightweight, to start up and get you onto the web in a few seconds. The user interface is minimal to stay out of your way, and most of the user experience takes place on the web. And as we did for the Google Chrome browser, we are going back to the basics and completely redesigning the underlying security architecture of the OS so that users don't have to deal with viruses, malware and security updates. It should just work. Google Chrome OS will run on both x86 as well as ARM chips and we are working with multiple OEMs to bring a number of netbooks to market next year. The software architecture is simple — Google Chrome running within a new windowing system on top of a Linux kernel. For application developers, the web is the platform. All web-based applications will automatically work and new applications can be written using your favorite web technologies. And of course, these apps will run not only on Google Chrome OS, but on any standards-based browser on Windows, Mac and Linux thereby giving developers the largest user base of any platform. Google Chrome OS is a new project, separate from Android." I have before argued that the balanced compute model presents the greatest relevance to the user's daily work. This is precisely there where I think that products such as the Google Chrome OS will have a strong impact - not because they are rich or thin clients to access the Cloud, but because they enable the Cloud to actually work for users in a way that has real meaning to them. It's obvious that in a commercial perspective, a complete Could ecosystem is something in Google's interests - precisely why they need to be careful to maintain an open platform. In any case, this is no doubt a major step forward for the Cloud Computing paradigm.

OMG Cloud Computing Standards Summit

The Object Management Group is holding the Cloud Computing Standards Summit July 13-15 in Arlington, VA, as part of their Standards in Government & NGO's Workshop.The Summit will focus on  standards for government clouds. Participants include:

Peter Mell and Tim Grance - National Institutes of Standards and Technology
Reuven Cohen - Cloud Computing Interoperability Forum)
Kevin Jackson - Dataline, LLC
Mark Carlson - Storage Network Industry Association
Winston Bumpus - Distributed Management Task Force
Robert Grossman - Open Cloud Consortium
Craig Lee - Open Grid Forum
Nils Puhlmann - Cloud Security Alliance
Douglas Johnson - Organization for the Advancement of Structured Information Standards (OASIS)
Jim Warner - TM Forum
James Odell - OMG SOA SIG Co-Chair, CSC,
Ralph Thrash - CSC

Recommendations from this summit will be made available to the July 15th, NDU IRM Cloud Computing Symposium attendees. 

The establishment of compatible industry and government standards are crucial to the broad adoption of cloud computing. Just as standards allowed the global spread and enhanced the utility of the Internet, standards will drive the utility of the cloud computing paradigm.

Registration for this event is available at http://www.omg.org/registration/GOV-WS/css/index.htm

Cisco Unified Cloud Services

I am certain many of you reading my scribblings on CC are closely watching the Cisco Unified Service Delivery. Attached a few screen captures from their recent launch webcast and transcript from Padmashree Warrior’s podcast. You can see the launch webcast here: https://cisco.webex.com/ciscosales/lsr.php?AT=pb&SP=MC&rID=39623962&rKey=6BAC6AA9C5663A3A

Pretty interesting coverage overall. A few things caught my eye in the unified cloud services webcast…

•         Highlighted VMWare partnership and key ingredient role.
•         Service provider role/influence (esp telcos) in the enterprise will grow as they make the play for various enterprise services – Just like how they expanded beyond voice to other data/media services.
•         Cisco believes the service provider clouds as the key for bridging cloud islands (inter cloud vision).
•         Their ultimate vision is providing a common platform building block for service providers that can provide business and consumer services to any device, anywhere. Desktop related services are something to watch for.

Will Cloud save the world?

Reading a generally insightful commentary in a post by Eric Knipp at Gartner: The Cloud Will Save The World, I started thinking more carefully about the value proposition of Cloud. In particular, I liked Knipp's statement that Cloud is not "an easy button - in the abstract. Someone, somewhere, has to do heavy lifting in any software development endeavor to build a high-quality, high-availability, highly-reusable Web architecture."

I couldn't agree more - security, audit, governance, certification, etcetera all represent considerable challenges in Cloud Computing. Yet,, "-- companies who learn how to screw down the cost drivers while simultaneously enhancing value drivers to satisfy customer needs have a competitive advantage in their industries." Indeed - satisfying customer requirements, and finding simplicity in complexity, are competitive advantages.

Nu, does this enable the Cloud to save the world? Knipp goes on to say "that as the world grows more complex, the only chance we have to head off the disintegration of modern society under the weight of complexity comes through technological leaps in the form of disruptive innovation. -- Could this new level of simplicity in complexity be the disruptive innovation that saves the world - or at least gives us a bit more time?"

I do believe the title does get a wee bit ahead of itself - Cloud Computing offers considerable opportunities for GreenIT and transforming both consumer and enterprise technology, but it is still a bit early to decide whether it will "save the world" or not. Yet, the sort of thinking that started the momentum of Cloud - and is verily needed to make Cloud succeed and fulfill its potental - is the same sort of thinking that is needed to conceive and execute on the broad transformations of which we are in dire need. Creative solutions almost always require a grim determination in the face of criticism and pedantry. Although it may not have come in time, Green tech is now recognised as a necessity. Likewise, Semantic Web, however the etymologists wish codify it, is now recognised as an enabler.. Cloud, and other movements, which drive this sort of creative, revolutionary-yet-evolutionary problem solving combined with the necessary persistence, will likewise have their day.