Monthly Archive for August, 2010

Will virtualization become less important in cloud computing?

By Olafur Ingthorsson on August 30, 2010 in Cloud Computing and Virtualization. 1 Comment

It’ generally acknowledged that VMware are the leaders in virtualization technology. Products like the the vSpere has probably been installed in more data centers than any ofther virtualization platform. Not to minimize the importance of other platforms, like the Citrix XenServer and Linux KVM, the vSphere is the market leading virtualization platform that together with its strong installed base and head-start set VMware in a comfortable position for developing its cloud computing strategy. And then the question emerges – how is VMware cloud computing strategy actually playing out?

Well, first of all we know that by enabling multiple logical server instances on a single physical server, virtualization is usually seen as one of the fundamental enablers of cloud computing. But exactly how important is virtualization as a cloud computing enabler? By many, the importance of virtualization is considered to become less significant when cloud-based application platforms mature to the point where applications are built and deployed without any reference to current notions of servers and operating systems. Developers using Plaform-as-Services, like Google App Engine and Microsoft Azure, are never going to have to think about virtualized infrastructure. The same goes with Software-as-a-Service. So as companies migrate services to the cloud, they are less likely to purchase VMware’s vSphere virtualization platform, unless they are going to establish their own cloud computing environment.

VMware has unofficially acknowledged this transformation. Their CEO, Paul Maritz, speaking on the 2010 GigaOM Structure conference, said that “the cloud at the infrastructure level is the new hardware”. What Mr. Maritz means is that “as the cloud infrastructure becomes more complex, or hardened, there is a point where we will stop seeing the details behind the infrastructure making it agnostic” and, hopefully, interconnected clouds from multiple cloud providers (federated clouds). All this of course is heavily dependent upon standardization work. Several bodies are now working on cloud standards, including the International Telecommunication Union (ITU) which have formed a Focus Group for that purpose

So where does all this leave VMware when it comes to their cloud initiative, especially if virtualization is to be a given thing – and ultimately more restricted to cloud provider and large company data centers? Well, VMware is already offering cloud platforms for various needs, both for public and private clouds. Their cloud vCloud Express platform was built for IaaS providers in particular, to offer services to end-users and to create PaaS services.  Already, it seems that VMware, e.g. through its partnership with SalesForce around VMforce, is moving more into the realm of PaaS by its Java Cloud enterprise offering. In the longer term, it is likely that virtualization will become less important when it comes to PaaS offerings and public clouds, while it probably will last longer in private cloud environments.

Tags: , , , .

Are telcos not finding their cloud computing niche?

By Olafur Ingthorsson on August 20, 2010 in Cloud Computing. 1 Comment

Are telcos loosing in the race of becoming viable cloud computing service providers? At least it seems that many telcos have had a difficult time in positioning themselves in the cloud ecosphere or even decide what cloud services to offer. Obviously, this does not equally apply to all telcos. Verizon, for example, seem to have already come a long way in providing a robust IaaS services through their Communication-as-a-service (CaaS) offering, stressing in particular data security, bandwidth guarantees and SLAs, and BT in the UK are providing, or reselling, a bundle of third-party applications as SaaS services enhanced by value-added services like billing, security and more.

However, some analysts maintain that telcos have a difficult time in becoming a dominant cloud computing player or even identifying their particular niche. Analyst Jeff Kaplan, points out that telcos are still struggling to figure out managed services that already have been around for a decade, and SaaS is still far from their reach. Telcos are also not exactly known for their innovation pace and therefore are in danger of being bypassed by more agile service providers.

While this may have some truth in it, I tend to see a number of inherent strengths that telcos are in a particularly good position to capitalize upon. For one thing, telcos in general have a good reputation for reliability and security, based on their legacy of running mission critical systems (voice and networks).  People expect telcos to deliver services with high uptime and reliability. The five-nines, representing 99.999% uptime is a famous reference to the expectations and benchmark that telcos have traditionally placed on their services.

Telcos are in a good position to deliver better than “best-effort” service delivery on the Internet. They can for example provide guaranteed bandwidth in their networks, including MPLS and VPN connections and, through their traditional association with multiple peering- and exchange points, capable of establishing multi-carrier guaranteed bandwidth for business customers that require high-quality services.

Through their extensive communication services, telcos can provide value-added and bundled services with third-party application providers, e.g. SaaS providers. This can include networking and phone services. Another interesting aspect is the concept of  “mash-up” services with a vast range of applications available through the cloud. Telcos can provide and integrate various services and data, including identity management, location information and billing capabilities with other applications and thereby provide cloud services with added value to their customers.

Certainly there are many more opportunities available for telcos to capitalise upon in the cloud that will become more visible in the coming months and years. Let’s not forget that cloud computing is still in its infancy and a lot may still change before cloud computing becomes a mainstream and the primary IT provisioning and service delivery channel for businesses.

Tags: , .

Mobile Cloud Computing and Cloud Phones!

By Olafur Ingthorsson on August 10, 2010 in Cloud Computing and Mobile Cloud Computing. 3 Comments

I recently started leading a new multi-national research project about Mobile Cloud Computing, it’s development, impact and potential opportunities for communication providers (telecoms). It seems to me that Mobile Cloud Computing is going to become increasingly important in the the near future in terms of providing browser based access to more and more cloud-based applications and services. Instead of downloading and installing applications on the mobile handsets (smartphones, tablets, etc.) users will access them directly in the cloud and display through the mobile browser, i.e. analogous to Software-as-a-Service provisioning. Some predictions include:

In Gartner’s 2010 key IT predictions for organizations, mobiles phones are expected to overtake PCs as the most common Web access device worldwide by 2013.

ABI Research predicts that there will be nearly one billion end users accessing the “mobile cloud” by 2014. Smartphone applications will move from the handset itself to the cloud – creating a ecosystem for new kind of smartphones – sometime termed “Mobile Cloud Phones”.

There are many reasons for this to materialize. For example, developers are increasingly discontent by being forced to develop and maintain applications that are exclusive for certain mobile phone platforms, e.g. iPhone, Android. This is both costly and limits developers to adhering to the rules of the platform owner – and, sometimes a limited market segment. Many users are also frustrated to have no choice but purchasing a powerful, and more often expensive, smartphone to have the possibility of running mobile apps.

It is clear that with the emergence of the iPhone and the App store, there is already a huge supply of applications avilable. However, most of these applications require users to download them to the handset, install them and run on the handset. Concequenlty, smartphones are becoming increasingly advanced and powerful, embodying sophisticated computing architecture and operating systems. The best analogy could perhaps be the Wintel cooperation – whereas a new version of Windows OS required the user to upgrade to a more powerful Intel processor.

Instead of handset-centric installation, new Cloud Phones will be able to utilize 4G, LTE networks to their fullest and perhaps adapt specifically to certain cloud provider applications or infrastructure. This will be a very interesting segment to investigate and there are many questions that arise. For example, will some of the current smartphone platforms, e.g. the Symbian, MeeGo, Apple’s iOS or Android be dominant in new “Mobile Cloud Phones” – or will there be a completely new platform that has still to be developed!

Tags: , , .

What to put into the Cloud – and what not!

By Olafur Ingthorsson on August 3, 2010 in Cloud Computing. 5 Comments

There is an endless discussion and debate taking place on what services and applications should typically be moved into into the Cloud – supporting the benefits of Cloud Computing in general, and also what services are less appropriate in the Cloud.

The TM Forum has put together an interesting and simple list of those primary services, that should not come as asurprise for the veteran Cloud Computing user or Service Provider. Still they are important to keep in mind when determining the appropriateness of when to consider migration to the Cloud – and when not:

Promising Cloud services:

  • Compute, storage services
  • Development and test environments
  • Audio, video and web-style collaboration
  • Select industry applications (e.g. SalesForce)
  • Analytics
  • Contact center services
  • Data storage and archiving
  • Business continuity and disaster recovery
  • Desktop virtualization

However, not all is rosy – there are some challenging services and applications that should perhaps be avoided to be migrated to the Cloud, e.g.:

  • Sensitive data (e.g. moving offshore)
  • Highly customized applications
  • Complex transactions
  • Applications and data with regulatory restructions
  • Legacy batch applications

This is certainly not a complete list of all potential Cloud Computing services and applications, but gives the indication, not surprisingly, that the benefits of Cloud Computing are largely restricted to standardized provisioning of services and applications. In the long run, however, Cloud Computing can also become increasingly applicable to more customizable environments and with more comprehensive international legal frameworks, less sensitive towards data location.

Tags: .