Chris Hoff (now a colleague at Cisco, but long a phenomenal blogger in his own right) described in a recent post a fascinating analogy for the inevitable adoption of cloud computing--the adoption of the Apple iPhone:
While I have often grouped Cloud Computing with the consumerization of IT (and the iPhone as it's most visible example) together in concert in my disruptive innovation presentations, I never really thought of them as metaphors for one another.
(Credit: CBS Interactive)
When you think of it, it's really a perfect visual.
The iPhone is a fantastic platform that transforms using technology that has been around for quite a while into a more useful experience. The iPhone converges many technologies and capabilities under a single umbrella and changes the way in which people interact with their data and other people.
Hoff goes on to note several specific parallels: our willingness to be locked in to specific providers to gain the iPhone's benefits; our admiration of those who work to innovate beyond proprietary boundaries through jailbreaks and "unapproved" application marketplaces; the desperate scramble by a variety of vendors to attach their star to the iPhone brand; and the constant rate at which features appear and evolve.
In the end, Hoff notes:
The thing I love about my iPhone is that it's not a piece of technology I think about but rather, it's the way interact with it to get what I want done. It has its quirks, but it works...for millions of people.
The point here is that Cloud is very much like the iPhone. As Sir James (Urquhart) says "Cloud isn't a technology, it's an operational model." Just like the iPhone.
Hoff is referring to my earlier post pointing out that the cloud isn't about new classes of technologies, but about technologies written to support the cloud model of self-service, on-demand, at scale operations. Corny nicknames not withstanding, Hoff also picked up on two fascinating things in his riff:
The iPhone isn't a technology--at least, not a new category of technology--but rather a platform that delivers a new operations model for mobile device consumers. iTunes and the iPhone App Store are services that simplify iPhone diversification to the point that it meets the needs of an incredible variety of users; from the biggest geeks to the technology illiterate. (OK, so I wouldn't be surprised if the former dominated the latter.)
Cloud computing will evolve much in the same way that the iPhone itself has evolved; early versions of the technology itself now lacks key features, but the operations model is what is compelling, and it keeps early adopters coming back for more.
You have to wonder if Hoff shouldn't have included the Google Android in his analogy, as Steve Oberlin suggested on Twitter. There are choices in the cloud space--it's not dominated by any one vendor, though Amazon may be the Apple of the cloud today.
I'm generally sold. I wonder what you think. Will cloud computing see massive adoption as more and more people (and companies) are seen benefiting from related services, and more and more compelling applications and services are available from the cloud? Or are both just trendy subjects that will eventually give way to more traditional technologies and getting things done?
Software development "in the cloud" has been one of the really interesting developments to come out of the cloud computing market so far. While many early players, such as Zimky and Coghead died on the vine, there is a pretty robust Platform as a Service (or "PaaS") market out there today, with Google App Engine taking the most visible lead, and a pretty solid stable of Ruby on Rails-based hosting providers telling a compelling story of their own.
Such success is driving some new players to seek the spotlight, however. I wanted to highlight two that I found most interesting. They are very different from one another, but those differences highlight the breadth of opportunity that remains in the PaaS market.
... Read more
IBM launched late Monday a new portfolio of products and services for the enterprise cloud computing market, which the company claims builds on lessons learned from earlier cloud initiatives.
Targeted at providing standardized platforms for specific computing workloads, the products and services, launched under the Smart Business and IBM CloudBurst monikers, aim to change the way IT organizations build and deliver IT services.
"Cloud is an important new consumption and delivery model for IT and business services. Large enterprises want our help to capitalize on what this model offers in a way that is safe, reliable, and efficient for business," Erich Clementi, general manager of enterprise initiatives at IBM, said in a statement. "Today's Smart Business announcement demonstrates that we take this responsibility seriously with cloud investment and solutions targeting the early opportunity."
I spoke with Dennis Quan, director of autonomic computing at IBM, and asked him to break the announcement down with a little more detail. Quan started by noting that IBM has been pursuing cloud and autonomic computing initiatives for some time now and that these initiatives have taught the company lessons that were invaluable in developing Smart Business.
"For instance," said Quan, "new workloads like large-scale business analytics and information processing are driving new specialized approaches to computing."
Quan noted that standardizing these approaches, through automation and service management, is critical to meeting new IT demands.
Initially, IBM is focusing on two core workload types in the Smart Business offerings: development/test and virtual desktop infrastructure (VDI). According to Quan, the company will expand the list of workload offerings over the coming months and years.
Quan highlighted three delivery models for Smart Business cloud products and services:
Public cloud. IBM will provide its Smart Business portfolio as software services delivered directly from its IBM Cloud data centers located around the world.
Private cloud services. For Smart Business workloads, customers can engage IBM Global Services to assist them in building cloud computing infrastructures within their own data centers.
CloudBurst private cloud infrastructure. Initially only available for development/test workloads, IBM will offer a 42U data center rack of pre-installed and configured hardware and software as a "drop-in" private cloud option.
Data Center Knowledge had some more details about the CloudBurst rack:
The basic CloudBurst package includes:
- 1 42U rack
- 1 BladeCenter Chassis
- 1 3650M2 Management Server, 8 cores, 24GB RAM
- 1 HS22 CloudBurst Management Blade, 8 cores, 48GB RAM
- 3 managed HS22 blades, 8 cores, 48GB RAM
- DS3400 FC attached storage
- IBM CloudBurst service management pack
- IBM Tivoli Provisioning Manager v7.1
- IBM Tivoli Monitoring v6.2.1
- IBM Systems Director 6.1.1 with Active Energy Manager; IBM ToolsCenter 1.0; IBM DS Storage Manager for DS4000 v10.36; LSI SMI-S provider for DS3400
- VMware VirtualCenter 2.5 U4; VMware ESXi 3.5 U4 hypervisor
James Staten, principal analyst at Forrester Research, told me that IBM has an advantage here.
"IBM came to these solutions from a point of credibility based on conversations they started over a year ago," said Staten. "Many other systems vendors haven't done that, and it could hurt them in the long term."
Staten also noted, however, that IBM has chosen to handle the public cloud elements of Smart Business entirely on its own, choosing to compete with service providers rather than partner with them. This could put IBM into direct competition with some large customers, such as telecommunications companies and large hosting providers, according to Staten.
The Smart Business services are available now, while the IBM CloudBurst system will ship Friday.
HP, Intel, and Yahoo are announcing Monday at the first Open Cirrus Summit that they have signed on three more research organizations to their joint cloud test bed. The new institutions include the Russian Academy of Sciences, South Korea's Electronics and Telecommunications Research Institute, and MIMOS, a strategic research and development organization under the Ministry of Science, Technology and Innovation in Malaysia.
Open Cirrus, described by the companies as "a global, multiple data center, open-source test bed for the advancement of cloud-computing research," was launched in July of 2008, and represented one of the first large-scale systems deployments targeted at teaching and researching large-scale cloud architectures. IBM and Google have teamed up on a similar project.
According to the announcement, the three organizations will take on a variety of roles targeted at developing "tools and best practices," and "will help further benchmark and compare alternative approaches to service management at datacenter scale". Specific roles were described as follows:
Russian Academy of Sciences, the first Eastern European institution to join Open Cirrus, encompassing three organizations:
- Institute for System Programming (ISP) will conduct fundamental scientific research and applications in the field of system programming.
- Joint SuperComputer Center (JSCC) will engage in the processing of large arrays of biological data, nanotechnology, 3D modeling, and other applications, and port them to cloud infrastructure.
- Russian Research Center Kurchatov Institute will explore how cloud computing is different from other technologies, and apply its techniques for large-scale data processing.
Electronics and Telecommunications Research Institute (South Korea) plans to conduct research and development on the management architecture and content retrieval of massive data sets.
MIMOS (Malaysia) will develop a national cloud-computing platform to deploy services throughout Malaysia, focusing on enabling services through software, security frameworks, and mobile interactivity, as well as testing new cloud tools and methodologies.
"We're excited to see the growing momentum behind this extraordinary partnership--the Open Cirrus test bed," said Andrew Chien, vice president and director of Intel Research. "The new sites bring growing critical mass and more contributors to our vision of an open-source cloud stack as a strong, large-scale platform for research and development."
Update: Updated link to Hoff's terminology map and related image to his latest version.
I was privileged to be a part of the Enterprise Cloud Summit that took place at the beginning of Interop in Las Vegas a few weeks ago. The program was excellent, with an all-star list of cloud experts and a surprisingly large number of attendees who were new to cloud computing and trying to get a sense of what it was all about.
What was different from prior cloud-related conferences, however, at least for me, were the types of questions this inquisitive audience was asking. Almost nobody asked around defining cloud computing, but many took advantage of the show to ask panelists and speakers to describe how they could put the cloud to use in their own businesses.
The cloud conversation is moving from "what is it?" to "how would I use it for my business or institution?"
(Credit: Chris Hoff / Rational Survivability)I find this very exciting--and, quite frankly, very refreshing. The amount of energy spent on presenting and defending terminology and taxonomy has become a huge time-sink for those trying to advance the cloud discussion. It's not that I mind walking people through the differences between cloud computing and virtualization, but I'd rather focus my efforts on business cases and customer success stories (or even failures).
It's not that the industry has arrived at a common cloud definition--though the NIST definition has some legs, and I'm a huge fan of Chris Hoff's terminology map (pictured here). Rather, the market seems to have come to the conclusion that cloud computing has a lot in common with obscenity--you may not be able to define it, but you'll know it when you see it.
Perhaps the most beneficial aspect of this shift is the fact that we should start seeing some real business cases, use cases, and best-practice discussions appear in the cloud-computing discussion.
Best Buy running on Google App Engine; stories about impressive gains by the venerable New York Times and Animoto when they used Amazon Web Services; and Eli Lilly's tale of redefining research projects: all these serve as examples of cloud's value in the right contexts. We know from these examples that "batch jobs" are great cloud fodder (such as grid computing and image processing), as are applications with unpredictable scale.
We need to see more such examples publicized, however. Where are the financials with their complex models and data mining? Biotech with its constant data processing demand? Manufacturing with its "just in time" supply chain management?
Perhaps the examples will continue to be more of the same, but that's OK to me. Then we know where cloud's strengths and weaknesses are, and we can move the conversation forward from there.
The widely reported sale of key technology and talent assets from utility computing infrastructure vendor Cassatt to CA on Tuesday serves as a lesson to us all about the challenges that lay before enterprises transitioning to virtualization, automation and cloud computing.
I had reported that the end was near a few weeks ago, and former Cassatt Director of Product Marketing Ken Oestreich (now at Egenera) posted a beautiful overview of what CA is getting from Cassatt, and the opportunities that now lay the new owners.
I wanted to take a moment myself to reflect on what I learned during my two years with Bill Coleman's second company (the first being BEA, the legendary middleware acquired by Oracle). I have huge respect for Bill's technical vision, as well as that of former and current Cassatt technologists, such as Dave McAllister, Brian Berliner, and Rob Gingell.
Below are five key lessons I think Cassatt learned the hard way about selling cloud infrastructure to the enterprise, and the core insights I think both vendors and customers should walk away with.
Cloud computing is an operations model, not a technology. I spoke about this concept in length in my last post, but I should note that the core concept behind this came from my time at Cassatt. Coleman actually understood this better than most, even five or six years ago.
Cassatt was formed because Coleman believed that, if the network is the computer, it needed an operating system (which he saw as middleware, i.e. BEA WebLogic) and an operations system (i.e. Cassatt). Operations automation was the core of what Cassatt was pursuing, with a focus on optimizing infrastructure usage.
Yes, Virginia, there is a private cloud. As cloud computing models exploded into the collective IT consciousness about a year ago--yes, it was conceived of earlier, but the massive explosion of interest came about June of 2008--Cassatt argued that the utility computing model was at the heart of both internal and commercial cloud infrastructures. Add self-service provisioning, along with metering capabilities, and much of the agility and cost savings benefit of the cloud model could be applied to wholly owned corporate infrastructure.
One of the really interesting lab experiments attempted at Cassatt was utilizing ActiveResponse as a control system for delivering images to both internal and external cloud environments (notably Amazon EC2). Had this been commercialized--the experiment was nowhere near production ready, the last I heard--it would have made Cassatt the first private cloud computing environment as I currently define the term; a wholly-owned enterprise cloud control environment that can manage both internal and external resources.
You cannot forklift a revolution. As much as we'd like to believe it's all about return on investment (ROI), it just isn't. I think Coleman himself said it best in a recent article:
"What frustrates me is my own naivete," he says. "I thought I could give companies something radical that had a proven return on investment, and they would be willing to change all their companies' computer policies and procedures to get that. Right now it's hard to get people to get beyond proof of concept tests or a data center energy analysis."
You can't force a company to change every element of its policies, procedures and investment on a dime, no matter what the short or long term savings will be. There has to be a transition plan; has to be a way to allow the IT department to adapt to a new paradigm, hopefully under committed leadership that sticks around for a while.
This is what those that deny the value of an internal-external cloud transition miss entirely. No matter how much savings you offer, no matter how many new paradigms you introduce, the transition for enterprise has to be as gradual as the corporate culture and IT portfolio require. Baby steps will beat giant leaps in enterprise clouds hands down.
(The story is different for start ups and many small and medium sized businesses. Both culture and investment are easier to deal with at that scale, and the public cloud is much easier to embrace.)
It's easier to swap out a server than an attitude. Where does the resistence to change in the data center come from? It is certainly not the CIO or CTO. In the two years I was a sales engineer at Cassatt, I can't remember one company where the IT leadership wasn't sold on the ROI story.
However, when we turned our attention to getting buy-in from the "rank and file", the fun began. I wrote about this some time ago, but essentially the "big three" organizational siloes of the data center: network, server and storage administration, saw a threat to both their procedures and their jobs, and dug their heels in deep.
Lest you think this is a "virtualization problem", I'm hearing some of the same things from developers who have bypassed their IT organizations to utilize Amazon or some other public cloud service: the mantra, "we'll never move our production applications or data to the cloud", rings loud and clear throughout the IT world these days.
Bye-bye siloes. Unfortunately for die-hard data center "traditionalists", the end is certainly in sight for those "big three" siloes. For example, my employer, Cisco, announced today new certifications for "Data Center Management", targeted at the converging skill sets required to run virtualized workloads on converged computing environments. I hear the title "Virtualization Administrator" is increasingly popular amongst enterprise IT practitioners.
Perhaps this is the real window of opportunity that Cassatt missed; the reason it was well before its time. As IT organizations take baby steps to address the changing cultural and operational conditions brought on by virtualization and cloud computing, the environment will become more conducive to cloud models. That is perhaps the biggest reason that late-comers to the enterprise cloud party should all give Cassatt a tip of the hat this week.
Oh, and I can't wait to see what Bill Coleman gets up to next.
One of the most common questions I get from those exploring cloud computing for the first time is "what is the difference between cloud computing and virtualization?" It is an excellent question, as most IT departments are currently exploring the ways in which virtualization enables automation and provisioning agility. Given the fact that cloud is often touted for providing similar benefits, it can be confusing to understand why the two terms aren't equivalent.
My response to that question requires a bit of explanation, so let's step through the differences between the two concepts.
Virtualization is a technology.
(Credit: CNET/James Urquhart)When you run software in a virtual machine, the bits that represent the program's instructions run through a layer of software that "pretends" to be a dedicated server infrastructure, the hypervisor. The hypervisor is the heart and soul of server virtualization, and is the enabler of the consolidation and agility values that virtualization brings to the data center.
It is because of the hypervisor that virtualization is the true disruptive technology that enables cloud computing on a massive scale. Hypervisors allow servers to be multi-tenant without rewriting applications to be multi-tenant. Hypervisors allow operating systems and applications to install to a consistent hardware profile, even though they end up running on a variety of actual physical system implementations. Hypervisors also allow servers to be manipulated by software APIs, which greatly simplifies the act of automating IT operations.
Cloud computing is an operations model, not a technology.
When you run an application in a public or private cloud, there is no "cloud layer" that your software must pass through in order to leverage the physical infrastructure available to it. In the vast majority of cases, there is probably some virtualization involved, but the existence of hypervisors clearly does not make your data center resources into a cloud. Nor is the fact that Amazon EC2 uses Xen hypervisors the reason that they are a cloud.
What makes a cloud a cloud is the fact that the physical resources involved are operated to deliver abstracted IT resources on-demand, at scale, and (almost always) in a multi-tenant environment. It is how you use the technologies involved. For the most part, cloud computing uses the same management tools, operating systems, middleware, databases, server platforms, network cabling, storage arrays, and so on, that we have come to know and love over the last several decades.
Specific technologies, of course, gain significant importance in a cloud computing environment, such as policy-driven automation, metering systems, and self-service provisioning portals. However, all of these technologies--with the possible exception of the self-service portal--existed before cloud computing became a much hyped paradigm.
There is no doubt cloud borrows much from long established technologies. It is also true that cloud has borrowed from many long standing operations models, such as mainframe service bureaus. However, the combination of on-demand, at scale, in a multi-tenant infrastructure is relatively unique for the post client-server era, and is the reason why cloud computing is disruptive, rather than just another operations fad.
In a press release Wednesday morning, Salesforce.com and Google announced the availability of Force.com for Google App Engine, a platform as a service integration that allows App Engine developers to more easily integrate their applications with data from Salesforce.com.
Force.com for Google App Engine provides developers a set of tools and services to build new kinds of Web and business applications entirely in the cloud. Instead of having to manage and maintain their own on-premise infrastructure, developers can use real-time cloud computing infrastructure from Google and Salesforce.com to develop and deploy new applications on the Web. Force.com for App Engine will accelerate the success developers have had with App Engine in building Web applications, and makes it easy for them to integrate their App Engine apps to enterprise data stored in Force.com.
According to the release, Force.com for Google App Engine provides developers with a set of Java libraries that perform actions against Force.com APIs and services. Features are reported to include:
- A SOAP-based Web service to access Force.com resources directly from App Engine.
- Java libraries for utilizing the Force.com service specifically tuned for App Engine.
- Complete documentation, getting started guides, and a sample application
With this announcement, Salesforce.com has gained a Java platform to complement its own Force.com platform, which relies on the proprietary APEX programming language. Given the high penetration of Java in the enterprise applications market, this could open significant new doors for Salesforce.com among developers.
It should also be noted that this is not the first joint deliverable between Salesforce.com and Google, with connectivity between Force.com and Google Apps being announced last month. I'm keeping an eye on the two companies, as they are increasingly building a tight partnership, and have the potential of blurring the line between their respective partner and customer communities, creating potentially the largest cloud developer platform ecosystem by far.
In conjunction with the well attended Interop and Enterprise Cloud Summit conferences in Las Vegas this week, cloud infrastructure and service vendor 3TERA announced the 3TERA AppStore, an online portal containing a variety of "cloud ready" components for use on their AppLogic platform. This is the latest commercialization of cloud image stores, and another example of how cloud computing enables marketplaces that were difficult or impossible to do before.
(Credit: 3TERA)One of the earliest example of this trend comes from none other than Amazon, which provided a commercial payment system (called DevPay) for their Amazon Machine Image store some time ago. What has come of that experiment is in fact an amazing adoption rate for even some of the biggest software system companies in the business, including Oracle and IBM.
Others are working to catch up in this space as well. I know that at least one of the major IaaS providers and one of the cloud management vendors are working to make a play in this space. Getting people to buy pre-configured, pre-architected enterprise software to run in various cloud platforms is going to be an integral part of the cloud experience, and probably a very profitable one.
The 3TERA offering focuses, of course, on their platform, but AppLogic is one of a few platforms that take a true "virtual data center" approach to the problem. Sun's new cloud is another. My gut tells me that these guys have a bit of an advantage when it comes to "packaging" enterprise apps for the clouds, as they can easily include network and storage with the server architecture in one SKU.
Initial AppStore partners include CohesiveFT, Layer 7 Technologies, SOASTA, Tap In Systems, and Zeus Technology.
While most readers probably think "portability or mobility" when they think of cloud interoperability, the vendor community sees a shorter-term opportunity in standardizing the operation of clouds and cloud infrastructures. It's not that vendors don't care about image portability; it is an especially critical opportunity for so-called "cloud operating system" vendors.
However, the cloud operations opportunity--building a full-featured operations API and user interface for a cloud--is a daunting task, requiring tools for provisioning, management and monitoring, among others.
(Note that I am calling the term "operations" tools, not "management" tools. It may seem like semantics, but when I published part 1 of this series, one of the systems management tools vendors quickly came back and said "hey, that's us!" Unfortunately, I'm not talking about monitoring and responding to problems with specific cloud payloads here, I'm talking about the interfaces you use to provision and control the clouds themselves. Of course, there is often overlap between the two concepts, just to make things...er...cloudier.)
In some ways, cloud operations interoperability is your basic API standardization story; allow several different independent management tools interact with several different independent cloud platforms and providers. Create a loosely coupled market dependency between the tools you use and the systems you manipulate with those tools. This is always a desirable goal in any open software marketplace.
However, there is a little more at stake here than simply enabling tools to be used across platforms. Consider this scenario: an enterprise has a distributed application system that runs components on two different major cloud provider infrastructures, relies on services running on two or three more, and stores data in their local data center. What would be the most desirable way to address the management of this system?
I would argue that you want as few "panes of glass" (i.e. disparate management applications) as possible, preferably one. To do that, you need standards that the single, all-encompassing management platform can rely on to access and manipulate all of the cloud platforms involved in the system.
In fact, you probably want to combine SOA and cloud management into a single pane of glass, so you probably want the standards to cover just about everything in the computing stack, from servers, storage and networking to specific application components and data stores.
Today, the focus is on providing a unified API for Infrastructure as a Service operations. In addition to standardizing how systems are provisioned, when they are active and what policies apply for situations like component failure or load spikes, it is also critical that this API unifies the way in which images are imported and exported from each provider's platform. A cloud operations API needs to cover as much of the system life cycle as possible, including provisioning and deployment.
There is so much effort being made in this space right now, by so many groups, that it is at times a little overwhelming. Luckily, I found a resource that has helped me organize not only the predominant operations API effort, but also the image/data and application/service interoperability classes. Believe it or not, it is a wiki dedicated to cloud efforts in the federal government market. (I'm telling you, the feds seem to be way ahead of the pack when it comes to organizing cloud activities these days.)
The wiki lists the following key efforts for "Governance and Management":
Management of Virtualized Resources - DMTF (http://www.dmtf.org/initiatives/vman_initiative/)
Open Cloud Standards Incubator - DMTF (http://www.dmtf.org/about/cloud-incubator)
SLAs -at SOI (http://sla-at-soi.eu/)
Compliance - Industry organizations (e.g. HIPAA, PCI) (http://www.hipaa.org/)
I was surprised to see three organizations missing from this list: the Cloud Computing Interoperability Forum (CCIF), the Open Cloud Consortium (OCC), and the Open Grid Forum's Open Cloud Computuing Initiative (OGF OCCI). It turns out the CCIF and OGF are acting more as organizing entities here, and the OCC is working standards around data and application interoperability, which I will talk about in part 4 of this series.
There is a "Cloud Standards Summit" planned for July 15th in Arlington, Va., that may clear further organize "who's doing what" in this space. I think the fact that such a meeting is taking place at all at this point is phenomenal. It is clearly an artifact of cloud computing's distinction as the first major IT disruption to take place in a world that assumes openness and transparency, thanks in large part to open source software.
As much as anything, cloud computing has borrowed from open source in terms of its governing principles, which could well be open source's lasting contribution to the cloud. It took a few decades of Microsoft dominance to really get the open-source movement in full swing, but it only took a few months for things like the Open Cloud Consortium to spring to life. Open source has taught us to expect openness by default. The cloud is no different.
