Public Cloud Computing: Signs of the Future

Overview

While many questions remain about this evolving space, there is growing evidence about what cloud infrastructure services (aka cloud-based infrastructure as a service or IaaS) and its cousins, shared and dedicated hosting, will become. Simply put, the idea of cloud infrastructure services as something separate and distinct from these other hosted services will fade, and customers will gain more flexibility and agility across all their hosting services.

More specifically, shared and dedicated hosting will become less distinguishable from cloud infrastructure services by inheriting what had been the defining characteristics of cloud computing. These characteristics include each of the five essential characteristics of cloud computing according to the National Institute of Standards and Technology (NIST): on-demand self-service, broad network access, rapid elasticity, resource pooling and measured service. More information about the NIST cloud computing definition can be found here.

However, it won’t simply be the above characteristics that become more uniform across the key hosting services. There will also be a convergence of technologies. For instance, the same platform technology (think OpenStack and others) will be extended and increasingly used to serve up shared and dedicated hosting, as well as cloud infrastructure services. Management tools – another key technology – will also advance in order to handle all types of systems (physical, virtual, cloud) together rather than separately. On the business model front, the ability to support pay per use for short time periods (once a cloud-only notion) will also be available for dedicated servers. The concept of a “dedicated” server will require finer granularity along the time dimension. It won’t just be, “How many months do you need that server?” It will be how many hours, or possibly minutes.

Why should this become the new reality?

First and foremost, this reality already exists in a few isolated pockets in the hosting and cloud service provider segment. In the last blog post, ”Cloud Computing and Innovation in the Hosting Industry,” SoftLayer was introduced as a leading innovator. The company’s focus on innovation, particularly in the area of automation, led in large part to its rapid growth, financial success and recent acquisition by its executive management team and private equity firm GI Partners. With SoftLayer, customers can already perform a broad range of automated tasks, from provisioning to ongoing management, on both dedicated and cloud servers from the same API and from the same self-service interface. Further, dedicated and cloud servers can be provisioned together on the same network for reduced latency within multi-tier applications (for example).

Since the acquisition was formally announced on August 4th, SoftLayer has also publicly acknowledged it is in merger discussions with The Planet, another leading hosting and cloud services provider. On a related note, GI Partners already held controlling interest in The Planet. The current plan is to complete the merger of the companies in the 4th quarter of 2010.

On Friday August 6th, I caught up with some of the key leaders from these companies – Lance Crosby, CEO / SoftLayer; Doug Erwin, CEO / The Planet; and George Karidis, Chief Strategy Officer / SoftLayer. George is also leading the integration effort for the combined company. I not only learned some interesting background, facts and direction related to the merger, I gained valuable insights to the future of hosting and cloud computing that reinforce the vision from the beginning of this post.

The Merger

Some additional relevant facts include:

• Lance Crosby, Chief Executive Officer for SoftLayer, will lead the new company.
• Headquarters for the new company will be Dallas; there are no current plans to curtail operations in Houston.
• The full management team for the combined company will be named during the integration planning process.
• The merger creates one of the largest dedicated hosting companies in the world.
• The combined company will have over 78,000 servers under management.
• Both companies already offered cloud infrastructure services.

Thoughts on Cloud Computing

This is not just any merger in the hosting industry. It is not just a story of consolidating market share. And it is certainly not a story of low margins that begged increasing scale for survival. In fact, SoftLayer had already achieved EBITDA margins of more than 50% and annual revenue per employee of $750K. Then Planet had impressive numbers of its own including 15 million websites and 48,000 servers under management.

Instead, my take away from this merger is that a powerful new kind of hosting company is being born, one that will lead the way in terms of providing cloud-like technologies, capabilities and benefits across a broader range of hosting services.

I asked the respective CEOs for their broad perspectives on cloud computing. You know… How do you view it? Why does it matter to your business? Etc. Lance responded, “We believe everything we do is cloud – not just our multi-tenant cloud services.” This perspective fully reinforces what SoftLayer has done in terms of automating the provisioning and management of dedicated servers in the same ways it has for its cloud servers. Same self-service portal. Same APIs. Across the hosting industry, cloud services were first to receive capabilities like rapid elasticity and automated self-service. But SoftLayer has also brought this to the more traditional hosting services.

Doug had another telling response. “(At The Planet) we knew we had to be in a position to deliver anything our customers wanted - shared, dedicated, managed, co-located and cloud. They use all these things and they want to use them all together.” Doug also brought in some historical perspective which likely resonates with many readers. “Cloud is a word that has taken off, but we’ve been talking about the same idea for a long time. It has a lot to do with ‘how to satisfy a customer’s need instantaneously.’” He talked about some of his history in the systems management space (e.g. as EVP and COO at BMC). He and others in the industry can attest to earlier work in the system management space toward the same objectives of cloud infrastructure services. While server virtualization and other technologies may have helped “cloud computing” gain initial popularity as a distinct service, some of its early roots and early objectives had physical infrastructure in mind.

Summing it up

In some ways, cloud computing is coming full circle. By rapidly offering the services customers need, whether shared, dedicated or cloud, hosting services are getting closer to what they should be. And soon, with a merged, dedicated hosting company that will be second in size only to Rackspace, many hosting and cloud service providers will have no choice but to follow this new direction. For traditional hosting providers, it won’t just be about how quickly they can add cloud services to their list of offerings. They will also need to improve the automation of their other services, making them more cloud-like. For cloud-only players, there will be increasing pressure to take a broader view of public cloud computing. One that includes mixed infrastructures. For customers, it is all good - more of what they want, when they want it.

What do you think? Do you see this future playing out for cloud computing? Do you see something different? Let me know!