For some time now I've been advocating to my core network provider customers (and anyone else who would willingly listen) a concept that I think is both central to the future of cloud computing and one of the great opportunities this market disruption presents. The idea is simple: who will be the cloud service aggregator to enterprises, large and small?
Think about it. Even the smallest of businesses require multiple software systems to both meet minimal data management and creative expectations and (I would hope) to establish come competitive differentiation where it counts. Large enterprises scale this need to the hundreds or even thousands of services, meeting the needs of everything from simple departmental database applications to core ERP and CRM systems on which the enterprise itself is managed.
Picking and choosing vendors to meet these needs makes sense; no one vendor is likely to deliver and end-to-end, differentiated IT offering for each and every business it serves. Mix-and-match is the name of the game in the cloud, and it should be expected that every enterprise will have a variety of functionality delivered to it from a variety of cloud service vendors and platforms. Many of which might be on internal cloud systems.
So how does one manage such a hodge-podge of business systems, serving a large number of stakeholders with unpredictable, often dynamic functional, integration and operational requirements?
What I've been advocating is the need for a cloud aggregator and broker service. I don't claim that is unique, or even new. The term "cloud broker" is one piece of this, as is integration, aggregation and customer support services.
Allan Leinwand, founder of open source switch platform, Vyatta, wrote a very well thought out post recently describing a cloud network aggregation business that he nicknamed a CloudNAP (Cloud Network Access Point). His description is of a private networking service that would specialize in delivering value to cloud consuming enterprises in a very specific way:
This new provider -- let's call it CloudNAP (Cloud Network Access Point) -- would solely be in the business of providing a toll road between the enterprise and the public cloud providers. The business of selling connectivity to the Internet, or transit, is a common ISP offering. The CloudNAP transit service would be different, however, in that it would be focused on delivering connectivity solely between enterprises and cloud services providers and not between enterprises or between clouds. In order to make network connectivity to the toll road cost-effective for an enterprise, CloudNAP would offer POPs (point-of-presence) in multiple geographies. Each CloudNAP POP would have dedicated leased lines to the networks of the major cloud services providers such as Amazon Web Services, Microsoft Azure, Google AppEngine, the Rackspace Cloud, etc.
The CloudNAP network could guarantee performance between the enterprise and the cloud by working with the service providers to enable the use of quality-of-service techniques that are not available over the public Internet such a Multiprotocol Label Switching (MPLS) classes for WAN connections or IEEE 802.1p priorities for LAN connections. Perhaps CloudNAP could even restrict the use of connections to cloud service protocols and services like REST (representational state transfer) or HTTPS (Hypertext Transfer Protocol Secure) -- thus preserving the network for its intended use by the enterprise.
I think Leinwand is on to something here, though I think he is a little bit pessimistic about what could be done with existing core Internet services. I am certainly not sure that an independent business would--or even could--win at this game. My own thinking has led me to believe there are two likely winners in the cloud network brokering game: either the biggest cloud infrastructure service providers will be the enterprise's entry point into the cloud, or the telecommunications giants will extend their network service offerings to address cloud computing.
An independent firm would just be too beholden to both the cloud service providers and the ISPs to make the end-to-end story work seamlessly. If such businesses are created and establish footholds, I would predict rapid consolidation through acquisition into one of the two market options I described.
Now, my personal preference would be for the telecoms to take this market, as it would set up independence from the services being provided, assuming there is no attempt to push users into a service wholly owned by the telecom. The telecom option would also allow for aggregation of billing, monitoring, and auditing streams across most if not all of the marketplace. Having Google or Amazon or Microsoft also provide network services to their clouds and those of their partners would in effect lock the customer into a slice of the overall market, which would appear much less desirable.
Unfortunately, for a variety of reasons, the telecoms just aren't seeing this massive opportunity. Instead, they are all working very hard to build "enterprise-ready clouds"--those compute, storage, and occasionally software services that just about every other cloud service provider is targeting with dollar signs in their eyes. There is no doubt in my mind that the telecoms will capture a reasonable share of that overcrowded marketplace.
Add to that an insight into AT&T's revenue model that I got from Joe Weinman, the leader of that company's business portfolio development for emerging technologies (and my fellow panelist at CloudWorld last week). Joe noted once that AT&T does not recognize revenue that is passed on to another business. In other words, if they resell somebody's cloud service, the only revenue that AT&T would recognize would be the cut they took off the top. Thus, according to Joe, they are driven towards building all of these services themselves, as it allows them to recognize the full range of revenue for the services they sell.
Update: Jeff Kaplan of ThinkStrategies, and the moderator of the CloudWorld panel, emailed me to point out an excellent post he wrote covering the telecoms' dilemma in more detail.
However, just imagine for a moment that AT&T and their peers put aside the desire to "out enterprise" Savvis or Terremark or even Amazon or Google, and instead focused on doing what they do best: being the provider of services that act as a gateway to a much wider market of services. Imagine those services are sold much like mobile phone plans, with monthly base charges for a certain level of service, and overages charged by CPU hour or whatever other metric makes the most sense. The service, in turn, provides a single face to almost the entire cloud marketplace.
I personally think that would be a very welcome addition to the existing ISP services offered by the big guys, and something that would make cloud computing palatable to a larger number of enterprises that they would have been otherwise.
Oh, and just imagine if the telecoms were to further extend the Internet infrastructure itself to provide much of the QoS and other services that Leinwand noted. For example, imagine technology that extends the enterprise network into the Internet infrastructure itself, allowing the VPN and other services to behave exactly as if they were under the control of the enterprise network administrators. No need to buy a separate, dedicated private network for your cloud systems, but instead subscribing to a list of services directly through the Internet.
That's something that only the ISPs and telecoms can do, and it is--in my opinion--a requirement for the Inter-Cloud of the future. I only hope one of the core networking greats will take the chance and be the first doorway to the cloud.