Blog Archives

Recently, Rob Mills wrote a post on Think Vitamin proclaiming death to traditional web. He was pointing out the widespread use of mobile devices and the next wave of iPad like devices and argued that this will make traditional web secondary to mobile sites and apps. 

The web is dead. OK, it isn’t but it might be dying a slow painful death when it comes to how users access online tools and the platforms they use to carry out certain tasks.

He, then, argues why he thinks mobile sites will take over the traditional web

There are huge advantages to iPhone, iPod Touch’s and the mobile web but it does mean that designers and developers now have new parameters in which to be creative, a world where attention to detail prevails, or at least it should. Will it be that the traditional web will be secondary to mobile/iPhone/iPad sites?

I think he is plain wrong. Let me explain why he is wrong and the web as we know is not going out anywhere.

1. First, and foremost, he has completely misunderstood the very nature of web itself. For some odd reasons, he thinks various websites and applications that are part of the web itself as the web. For me, it is a very simplistic view. Rather, I see web as a platform on top of which we build these many different websites and applications. A simple design problem doesn’t mean a death to the very platform itself. 
2. Unlike many other platforms, web is a continuously evolving platform. It started as a collection of documents and evolved into a platform on top of which applications can be built. Then, we added a social layer on top of the platform to do wonderful things. Next, it is evolving into a web of data, making the platform much more robust. The next evolution is going to help us develop more intelligent applications. The web as a platform will continuously evolve offering us newer ways to build applications on top of it. Such evolving platforms rarely die a fast death. Sorry my friend, the design issue you are quoting will not kill the web.
3. Plus, the author seems to be completely ignoring what will happen when HTML5 takes further foothold in the web. We cannot dismiss HTML5 as inconsequential. Who would have thought till few months back that Flash could face a potential danger of losing the dominance in the video marketplace. If anyone had talked about death to flash sometime last year, we would have strongly recommended a visit to his/her Psychiatrist. Apple’s refusal to allow flash on iPad has made talk about potential death to flash plausible. We should not underestimate what HTML5 can do to web in the coming years.

I think I have offered solid arguments against the post on Think Vitamin. I would love to hear from the readers of Cloud Ave about what they think about the issue. Feel free to jump in.

On Friday, Eucalyptus Systems, the open source software powering the private clouds in the enterprises, announced that the former CEO of MySQL AB, MÃ¥rten Mickos, has taken over as CEO of Eucalyptus Systems replacing Woody Rollins who will now be their CFO. This has taken the entire industry by surprise. I am also equally surprised because I just spoke with the Eucalyptus team including Mr. Rollins on wednesday and I got no clue about their impending move. In this post, I will briefly talk about Eucalyptus Systems and what this move means to them.

Eucalyptus started off as an NSF-funded academic research project at the University of California, Santa Barbara, developed by a team of researchers headed by Dr. Rich Wolski. They started Virtual Grid Application Development Software Project to develop a programming language and runtime system techniques for large-scale computational grid applications. At the same time, Amazon EC2 was gaining traction and they wanted to see how AWS can be tapped for their project. In order to do this, they developed Eucalyptus software to act as a local cloud platform. Essentially, they developed a platform that we call in today’s terminology as Private Clouds. They built their platform on top of unmodified open source components so that they can incorporate any future upgrades without any hassle and, also, run on top of any existing Linux distributions. From the beginning itself, they made a conscious decision to develop a modular platform so that it can be integrated into many different virtualization environments. Even though they supported only Xen (and, hence, the support for AWS) in the beginning, this decision to have a modular architecture is allowing them to support other virtualization platforms like KVM, VMWare, etc.. When I asked them about the support for Microsoft’s Hyper-V, they told me that they are looking into ways to integrate with Microsoft’s hypervisor.

From its academic roots, Eucalyptus Project morphed into a commercial entity with the name Eucalyptus Systems, supported by a $5.5 million Series A funding from Benchmark Capital. Once it became a commercial entity, they quickly expanded to add support for third party management tools like Rightscale, CohesiveFT, rPath, Ylastic, etc.. They also partnered with Canonical to support Ubuntu Enterprise Cloud. In Sept. 2009, they launched their first commercial offering, Eucalyptus Enterprise Edition, that offered support for VMware’s vSphere, ESX and ESXi. With this move, Eucalyptus announced their intention to take on the enterprise market. They expanded their offerings with partnerships with newScale and Terracota. The Ubuntu Enterprise Cloud 9.10 accelerated the adoption of Eucalyptus with a tighter integration which made the installation of Eucalyptus cloud a breeze.

The Eucalyptus strategy is simple. They will offer their core product for free and monetize with proprietary addons and support. Their addons with open source platforms will continue to be free but any addon/adapter with proprietary technologies is going to cost money to the users. It is a neat way to monetize an open source software and they are finding an enthusiastic response to this strategy. I did prod them to give out some numbers on the enterprise customers they have but they refused to disclose anything. However, they told me that they are very optimistic and they are in a strong position to close the sale with many enterprise customers.

Having seen the evolution of Eucalyptus Systems, let us take a look at the impact of the announcement of their new CEO. In my opinion, it is a smart move considering the role played by Mickos with MySQL. First and foremost, Mickos showed the world how an open source company can successfully monetize even while giving away the software for free. He was part of the MySQL team that helped grow the revenues from 14 Million in 2004 to an estimated 65 Million in 2008. In the business world, where the CEO’s words are filtered through their PR, he was brave enough to accept that MySQL gets one paying customer for every 1000 downloads. He was pretty clear about where he wants MySQL to go and the constraints it faces as an open source company. This will come very handy as he tries to steer Eucalyptus in a cloud world dominated by proprietary vendors.

MySQL started off as a darling to web developers and continuously tried to reposition itself as an enterprise player in a market dominated by Oracle, IBM and Microsoft. Mickos played a very important role in that repositioning and helped MySQL convince enterprise to take them seriously. In fact, strong database players like Oracle and IBM saw this smaller open source vendor as a long term threat to their business. Mickos helped MySQL navigate this competitive marketplace and this experience will come handy as he helps Eucalyptus Systems navigate a competitive marketplace dominated by some strong players.

Sun Microsystems surprised many pundits with its 1 Billion acquisition of MySQL. Even though it was a big surprise to many at that time, we cannot overlook the role played by Mickos that resulted in the high value deal. As he tried to thrust MySQL into the enterprise, Oracle tried to grab MySQL into its fold but Mickos pooh pooh-ed the attempt by Oracle and pushed MySQL ahead. Eventually, his team convinced Sun Microsystems to play a fortune to acquire them. In fact, many pundits are already wondering why Eucalyptus Systems is not acquired by one of those bigger players. I see the appointment of Mickos as CEO to be the first step towards such an endgame. This man knows how to do it. His experience in taking a small open source company across a competitive marketplace resulting in a $1 Billion paycheck will go a long way in taking Eucalyptus Systems towards a big acquisition. Plus, Mickos is not new to the cloud marketplace and he must be having a fairly good idea of the marketplace since joining Rightscale’s board as one of the directors. I am sure Benchmark Capital took all these into consideration before putting him on the driver’s seat at Eucalyptus Systems.

The next few years will be crucial for Mickos as well as Eucalyptus Systems. They need to show the world that they can monetize their product successfully by establishing a larger footprint. Then they need to use all the available skills, including that of Mickos, to find a buyer who can give them a fat paycheck. In my opinion, the process has already started and we have to wait for the endgame.

Cloud Connect is in full scale today. Yesterday, there was the Cloud Business Summit where top industry executives spoke about the impact of Cloud Computing on today’s business. I had a chance to talk to many leaders in the industry and I could feel the cloud in the air. People are pretty excited about how cloud computing is going to change the entire business landscape in the coming decade. The sessions ranged from discussions about which business models are winning to exploring different challenges and opportunities in the enterprise market to the investment scene and go to market strategies.

I had a chance to speak to Mr. Jason Green of Emergence Capital Partners yesterday. Emergence Capital were an early investor in Salesforce.com and their portfolio includes such big names like Successfactors, Echosign, Yammer, etc..  He sounded very positive about the state of the clouds. He told me that they exclusively support cloud vendors, especially on the applications side. He was telling me that he sees a big increase in enterprise adoption and expects it to stay that way in the coming years. 

Another interesting topic in our discussions was about SaaS players using other cloud based services (platform and infrastructure) for their needs. I had a twitter discussion long back on the same topic. We see SaaS and Web 2.0 startups using IaaS and PaaS for their infrastructure needs early on and move to their own datacenters as they grow big. I am really keen to hear from SaaS companies on why we see such a trend while IaaS and PaaS players are trying to get enterprises move into the cloud. It is also interesting because on one hand, they want customers to trust them and put their data on their services but, on the other hand, they don’t think they can rely on IaaS or PaaS players to deliver their apps. Ideally, I would like to see SaaS providers use the -aaS’ all the way down to the infrastructure part of the stack.

In short, I get a feeling that Cloud Computing has finally taken off. In spite of some hyping on the vendor side and scare tactics on the side of the traditional software vendors, it has finally reached mainstream. I spoke with a couple of ISVs and both of them told me that they were at the conference because they want to move their apps to clouds. Nope, they didn’t talk about cloud washing, they want to re-architect their applications so that they can offer them “as a service”. If anything this conference has highlighted, it is the fact that the cloud has arrived.

In my opinion, Google not only enabled somewhat bigger players to sell in their marketplace, they also provided a way for developers and other smaller service providers to play in the marketplace. For example, individuals can offer services like setting up Google Apps for $30. All they have to do is to guide the customer to signup for Google Apps, setup the DNS records and activate different services in Google Apps. Any marketplace that opens up many opportunities for small and big players alike is a true marketplace and it has the potential to develop into a vibrant ecosystem.

Somewhere in the hype about the marketplace, another interesting news got lost in the noise. Google has opened up the Google Apps script gallery to developers and users so that they can publish their scripts in the gallery for others to consume. 

Today, we are excited to make Google Apps Script available to everyone. Some of you may already be familiar with Google Apps Script within Google Apps, but in case you are new to it, Google Apps Script provides a powerful and flexible scripting environment that lets you automate actions across your spreadsheets, sites, calendars, and many other services.

This is pretty neat. This not only allows experienced developers to add value to Google services, it also opens up opportunity for relatively novice developers to showcase their talent by adding features on top of Google services with their own scripts. All these make Google very attractive for developers and has a potential to ensure a vibrant ecosystem around their services. What do you think?

Recently, there was a discussion on Twitter about the fate of small webhosts. I thought I will expand my thoughts here in this post. Traditional web hosting ecosystems is huge with big hosts offering enterprise level managed hosting to a college kid having a reseller account to sell hosting space to friends and family. As we move ahead with cloud based hosting, the natural question is about the fate of such small hosts. In this post, we will take a look at what is in store for them.

The cloud world will definitely be harsh on most of the smaller webhosts. There is no doubt about it. However, it doesn’t mean that we will see a world where there will be a consolidation of handful of monopoly cloud infrastructure players, which is a shortsighted idea. There are many constraints to such thinking along with other mundane reasons for a more federated ecosystem.

Two weeks back, I wrote a post about an European cloud infrastructure provider, ScaleUp Technologies, and I wrote about how one of the traditional webhost, Internet4you, has morphed part of their datacenter resources to offer cloud based services using 3Tera’s Applogic platform. Recently, I spoke with VMOps, a company developing software stack that helps service providers set up Infrastructure as a service easily. Their software stack comes with three components.

• A retooled multi-tenant hypervisor that supports dynamic resource provisioning and complete isolation of CPU, memory, storage and network resources for virtual servers.
• A management tool that helps service providers to package their offerings, setup and manage users and, also, an integrated billing solution.
• An easy to use self-service interface for end-users with necessary API to allow them to control the launching of applications in the cloud.

Companies like Reliacloud and Cloud Central have already tapped into the VMOps stack to offer IaaS offerings. I will dig deep into VMOps in the future but software like the ones offered by VMOps and 3Tera shows tremendous potential for smaller webhosts to jump straight into the cloud bandwagon. These software will help webhosts who have their own datacenters to reposition themselves as a cloud provider.

What about the shared webhosting providers who either rent a dedicated server or use a reseller accounts? A big chunk of them will eventually vanish to the pressures of market forces. However, there are still opportunities for these smaller hosts in the cloud world. They could add value on top of these cloud offerings and then resell it to low end users like some of the small businesses. For example, they could add some management layer and a customer support layer on top of raw EC2 instances and sell them to individual and small businesses who neither want to manage these EC2 instances or don’t know how to manage them. As I pointed out in my post about Blackmesh and the future of small webhosts, customers want support. They want to talk to a human beings and get personalized support. Then, there is the idea of regional clouds where some of the customers (somewhere in the long tail) want to do business with a cloud provider who has operations nearby. Now add the regulatory requirements and other factors based on diverse needs of the users, I see huge opportunities for smaller level players.

Yes, cloud era is going to drastically reduce the number of cloud infrastructure players. Yes, most of these smaller webhosts are going to shut down their business and go away. However, we will still see a vibrant federated cloud ecosystem and we will see software tools that will enable some of these small players to reposition themselves to play the cloud game. The future is in an open, federated cloud ecosystem.