The new announcement by Intel regarding a Cloud computing alliance is very interesting and poses many questions. It represents more than 70 top global businesses unified by Intel. The alliance is supposed to drive interoperability, flexibility and industry standards for the cloud and next generation data centers. More about the alliance can be found on their website.
Will they discard the NIST Taxonomy and definitions for cloud computing? Why did Intel decide to start a new alliance when open source projects like DeltaCloud exist? The good news is that the cloud buying community is an integral and essential part of this alliance. How will this alliance collaborate with members who have signed the Open Cloud Manifesto?
The term service orchestration was used in several contexts and no wonder Wikipedia called it a buzz word. I felt compelled to dive deeper. With help from some of my colleagues I was able to piece together something helpful.
Is service orchestration the same as a work flow automation? Workflows and processes are used in different domains. At one level there are level inter-company business processes and at another level there are processes to get a network operational. To keep matters simple, we will define an orchestrator as the entity that manages complex cross domain (system, enterprise, firewall) processes and handle exceptions. Since an orchestrator is valuable in the fulfillment, assurance as well as billing processes, in it’s most advanced service aware incarnation, should be capable of adjustments based on feedback from monitoring tools. At the most basic level an orchestrator is a human. The main difference between a workflow automation and orchestration is that workflows are constrained to complete a process within a single domain.
Wikipedia defines a service as a set of singular and perishable benefits
• delivered from the accountable service provider, mostly in close co action with his service suppliers,
• generated by functions of technical systems and/or by distinct activities of individuals, respectively,
• commissioned according to the needs of his service consumers by the service customer from the accountable service provider,
• rendered individually to an authorized service consumer at his/her dedicated trigger,
• and, finally, consumed and utilized by the triggering service consumer for executing his/her upcoming business or private activity.
Cloud Service orchestration therefore is the
• Composing of Architecture, Tools and Processes by humans to deliver a defined service
• Stitching of software and hardware components together to deliver a defined Service
• Connecting and automating of workflows when applicable to deliver a defined service
It is critical in the delivery of cloud services because
• Cloud is all about scale – automated workflows are essential
• Cloud Service delivery includes fulfillment assurance and billing
• Cloud Services delivery entails workflows in various technical and business domains
The $15 million funding of Nimbula by Accel Partners this week received a great deal of attention. After a quick look at the website I decided to see who their competitors would be. I was startled to see the number of offers in the market. There were a number of lists and all seemed to be partial and many included vaporware. The range was fairly big starting with open source projects like openstack yet to have their first release, all the way to decade old products which have evolved to fill in the needs of cloud computing.
Clearly service management in the cloud is a very important and broad topic. The sheer number of tools implies that a consolidation is bound to happen where the leaders will surge ahead. However it would be unfair to pool all the tools into one single group for cloud management. It is no wonder that coalitions (Eucalyptus,newScale, rPath) are forming.
Again I found the best way to group tools was to separate basic infrastructure ( hardware including hypervisors) control software as IaaS management tools (providers) and the others as PaaS ( Change/ Config) control software (providers). I excluded any tools which only monitor changes or provide existing configuration.
My initial list based based on gleaning the following is shown below:
a) Cloudbook.net Directory
b) Serachcloudcomputing article on open source tools
c) Syscon-media list
IaaS Orchestration tools
- Eucalyptus
- Deltacloud
- vCloud Director
- Nimbula
- OpenStack
- RightScale
- DynamicOps
- OpenNebula
- Enomaly
- Univa VD
- Opscode Chef
- HP Insight
- Tivoli Provisioning Manager
- Abiquo
PaaS Orchestration tools
- BMC Cloud Lifecycle Manager
- CA Applogic
- HP Opsware
- Cloupia
- Newscale
- Cloud.com
- Surgient ( Now Quest )
- Enstratus
- Cloudera
- rPath
- ManageIQ
- Incontinuum
- embotics
- Kaavo
- Elastra
- VMLogix
- CohesiveFT
I do expect this list to grow rapidly and also have some way to classify the products/companies and make it easy to understand.
A NY Times article reported that Google and Verizon are close to a deal on giving priority to youtube content on the Verizon network.
The blogosphere is abuzz with discussion as this hits the heart of the net-neutrality discussion. HuffingtonPost declares it the “end of the internet” as we know it.
AT&T’s move earlier this summer to change the pricing from an all you can eat for access to a tiered model, made the change on the consumer side. The Google deal is on the content provider side. Both emphasize the ability of the communication service provider to create value for Internet users and cloud service providers.
These developments have serious implications for CSPs. In my view this was bound to happen sooner than later. So what other changes should we expect on the internet? Are the Googles and Amazons the new partners for Communication Service Providers (CSPs).
A recent article on the Economist claims that most of the smart phone applications will not generate sufficient revenue. It makes a valid argument that many of the applications are teasers and access mechanisms to services and content on the Internet (cloud). Of the 225000 apps in the Apple App Store I am sure many do fall in this category, but that still does not diminish the importance of useful applications. I personally know of an application priced at $25 replacing stand-alone electronic devices sold for $100.
The Economist article also makes a naïve comparison of apps to music. Just because songs and applications can be stored digitally does not make them the same. Songs are expressions of creativity and the audience can get bored of songs over time. Software applications (not games) are not the same, and once users recognize their utility they don’t get bored of it. How many users of Excel do you know of, that have stopped using it because they got bored with it?
The Apple App Store is a marketplace, which gives application developers a way to sell their products, and Apple keeps 30% of the sale price. It is the ecosystem and network effect of the App Store, which is very powerful. Depending on the type of application and it’s ability to leverage the unique capabilities of the hand held device, an application is here to stay. Software development and application development is a viable business. Unfortunately it is content providers like the Economist that are unable to distinguish between musical content and software content.
“What is the Cloud?” This has been a topic of ferocious discussion for some years. The Wall Street Journal even chastised high-tech company executives for confusing users about this paradigm. Conceptually it is not new – we can draw parallels from the Utilities industry. In the past, before electric utilities were reliable, it was not uncommon for factories and enterprises to own electricity generators. Today, electric utility companies have proved their reliability, and it is economic for customers to purchase from them. The utility provider produces electricity for thousands of users, exploiting the economies of scale. The provider theoretically supplies an unlimited quantity of electricity on-demand to users. The utility companies take care of peak demands by bringing generators on board as needed and charge the customers based on their usage. Another utility that comes to mind is the Telco industry. They provide communication (voice, cell phone) as a service and charge customers based on usage. Wireless Internet services even experiment with the pricing structure. AT&T recently announced that it will discontinue its monthly fixed price plan for unlimited bandwidth use after it found that a small percentage of subscribers were responsible for disproportional use of network bandwidth. Today, we are witnessing the cross pollination of ideas between different industry verticals and are in the midst of a convergence of the Telco and Information Technology industry.
Cloud computing is the application of this utility concept to computation power. Computation is accomplished with four components – Network, CPU, Memory (RAM) and Storage. Cloud computing makes these components available on-demand and on a pay-as-you-go basis. Buyers can now forego the capital expenses of computing equipment and just pay for it as an operational expense. It is like choosing between purchasing office space and renting it. Purchase of computing equipment and software entails maintenance expenses, which often exceeds the initial capital expenses. When computation is routine and not a differentiator for an enterprise, it must consider the option of outsourcing it or purchasing it from a utility provider. This is a big driver for cloud computing. So is Cloud Computing the next Utility ?
Ranjit Nayak
After a short hiatus I am back to pursuing my passion of seeing cloud computing make a meaningful impact for businesses. High tech company executives have been chastised for confusing users about this paradigm. Rightly so because many existing products were renamed as cloud computing products. In defense of the high tech companies, the fact is that cloud computing is not a revolution but an evolution;E commerce has been around for more than a decade. Some of the ideas such as virtual desktops have been around for several years.
So what is different now? One big difference is that we now have more reliable networks, be it data or voice. More Users have broadband connections and have come to depend on the network. Cloud computing is an evolution where users are willing to use compute power in the network, using various access devices including smart phones. It is still a change from the past because users need to trust the cloud computing service provider to render services without interruptions, and secure all personal data. In the past, service interruptions were due to disk crashes or virus attacks on the personal computer.
A friend recently asked me how she could get up to speed on cloud computing, and which books would help. My initial instinct was to say there are no books, but then I went to check on Amazon. Sure enough there were no less than 293 books on “cloud computing”, several available for pre order. The books reminded me of blind men describing an elephant. This is not to demean their efforts in anyway, because I would explain it in a way my thinking has been shaped over the years. Books should help us synthesize the collective thought process and define the common vocabulary where there is a consensus. Since this is such a big and emerging topic, it takes time for the dust to settle. NIST, a US government agency had to step in and provide a definition of cloud computing we could agree on. Analyst firms like Gartner and Forrester have variations of this.
In the information age and with a topic as fluid as Cloud Computing it is imperative to keep in touch with daily news and blogging. I have found the following to be great sources.
a) Gigaom http://gigaom.com
b) Phil Wainewright http://blogs.zdnet.com/SAAS
c) Sandhill.com http://www.sandhill.com/
d) Cloudbook.net http://www.cloudbook.net/
e) And to top it all (no pun intended) Guy Kawasaki’s new venture “Alltop” – a news aggregator. http://cloud-computing.alltop.com/
f) Google groups – cloud-computing@googlegroups.com
- Ranjit Nayak
The convergence of Telco and Information Technology. For some reason the convergence is not an recognized as an obvious trend. I first saw this notion in a research study by IDC. The research indicated an intersection of the two industries and hence opportunities. As a person on main street (nice terms for Not a Pundit) I see that my telco has added new services such as Satellite TV. One identifier I have seen for such telcos is CSP ( Content Service providers). Last week I saw another article about Telstra the Australian Telco which reinforced the notion of convergence (http://www.billingworld.com/articles/telstra-integration-on-demand-smb-market.html)
I was at the Management World conference in Orlando in 2008. Although one of the keynote speeches did go into Amazon’s cloud services and intrduced Web 2.0 services to attendees, the average attendee did not see how online services were related to the services telcos offered. A very knowledgebale senior technical staffer even asked what “SaaS” was, and admitted that he had just heard about “Salesforce.com”.
This is a topic being discussed far and wide. And the politically correct answer is it depends! One look at this article (http://www.mercurynews.com/business/ci_11130755?nclick_check=1) on the San Jose Mercury and the ambivalence is evident.
The logic favors SaaS companies in an economic downturn. Buyers do not incur capital expenses, instead they incur a recurring expense which is accounted for as an operational expense. A recent survey by ScanSafe confirms this sentiment. We have counter arguments from none other than Larry Ellison. I for one do expect that SaaS companies will do well, but then seeing is beleiving!! Surveys and opinions can only go so far.
However there is some truth to the argument that in general, SaaS companies will have lower revenues due to the economic downturn as all investments are put on hold. This is very disturbing because even good investments will now be stalled. The CFO’s mantra in uncertain times is cash conservation.
A blog post by Phil Wainewright even suggests that an acceleration of the SaaS market could be disastrous for the software industry.
Until the 90’s, most software applications were written for mainframe use. In the 90s, personal computing and client server applications began their dominance. The 90s also saw the rise of the internet and a networked globe. Today, the pendulum is swinging back to server centric computing, much like the mainframe era. Software complexity contributes to this trend. Maintaining software by applying patches is cumbersome and time consuming. Personal computers, unlike data centers, are not very secure and reliable. There are several cases of laptops with classified data being stolen or lost. Users losing important data due to hard drive crashes and lacking any backups is not uncommon. Individuals and small companies cannot afford the time or manpower needed to maintain IT resources.