If I had a dollar for every time I heard or read the word ‘Cloud’ over the past year, I’d be a millionaire. And, this week would have been great for my bottom line as I attended Pacific Crest’s Emerging Technology Summit in San Francisco. In the halls, at the receptions, and at the sessions the ‘Cloud’ was pervasive.
However, given that the ‘cloud’ is on everyone’s lips, I thought I’d try to relay some insights on Cloud Infrastructure on display at the conference, talk a little about why what Dyn does (for better or worse) fits into the ‘Cloud Infrastructure’ rubric and, lastly call out two of the meatballs that fell from the ‘Cloud’ at the conference.
Why Is The Cloud Infrastructure Movement So Popular?
The consensus opinion at the conference about why the ‘Cloud’ has been such a draw revolves around three advantages that Cloud Infrastructure solutions present. The first advantage is that Cloud Infrastructure lowers cost/technical barriers to market entry and service/product deployment. More specifically, cloud infrastructure allows entrepreneurs and corporate IT staff to get technical resources on-line at a fraction of the cost (both CapEx and OpEx) of buying a piece of hardware and getting it running in your basement or installed in your corporate datacenter.
The second advantage is that Cloud Infrastructure provides technical flexibility that allows for rapid scaling. More specifically, the multi-tenant virtualized architecture of Cloud Infrastructure allows a user to consume x capacity today, 1000x capacity tomorrow and 100x capacity the day after that.
The third advantage is the business corollary to the second: Cloud Infrastructure enables amazing agility on controlling costs. More specifically, you can pay for x capacity today, ramp up spending tomorrow and buy 1000x capacity to run room for huge growth and adjust spend downward the day after that to make sure it is in line with margin/burn rate requirements.
Dyn’s Place in the Cloud
I am a bit of a Yankee, so I am a bit too curmudgeonly to join a movement that is people think is the latest “hot” trend, but given what we do here at Dyn (DNS, email delivery and deliverability) and how we do it, being part of this trend is pretty much unavoidable. Two of the main traits of any Cloud Infrastructure service are its use-based billing methodology and the fact that it dislocates the customer’s previously “on-premise” hardware.
Dyn’s services have had these traits since long before that meant the service was part of the Cloud. The cost of both Dynect and SendLabs services are use-driven and we happily allow our customers to use their out-of-date DNS appliances and mailservers as coffee tables. However, in addition to being Cloud Infrastructure services themselves, Dynect and SendLabs services are key to enabling other Cloud Infrastructure services. A DNS solution that enables dynamic routing and is truly scalable, like Dynect, is essential to delivering cloud storage and cloud applications services. An email solution, like SendLabs, that allows for reliable delivery of the transactional emails (account activation, password resets, event notifications, alert notifications etc.) is an absolute prerequisite in any Cloud service.
Meatball #1 – Don’t Follow the Money: Over the course of several presentations and panels, I heard more than one person talk about how the technical flexibility is what is really driving cloud adoption. I am calling “meatball” on this statement. It is true that the scalability enabled by Cloud Infrastructure is riding shotgun on the drive toward use of Cloud Infrastructure, but the threshold cost savings (i.e., you don’t have to buy a box) and the ongoing expense management agility it gives (i.e., you can manage costs by using more or less capacity) are in the driver’s seat.
Whether it is the penniless AppDev entrepreneur, the cost/savings-directive driven corporate IT manager or the CapEx averse venture capitalist, the business people, moreso than the technologists, are driving people toward the Cloud (the Cloud, after all, isn’t driven by a new chip, but by a new business model). If it were cheaper to buy and run your own infrastructure appliances, the “Cloud” would be a wisp in the wind.
Meatball #2 – The Cloud is everywhere and everything. There are “public” Clouds. There are “private” Clouds. There are “hybrid” Clouds. Software companies are part of the Cloud. Hardware companies are part of the Cloud. IaaS and SaaS companies are part of the Cloud. I am calling “meatball” on this too. If everything is part of the Cloud, then nothing is. I’m not sure that I should be the one to put the Cloud in a box, but unless the industry coalesces around some meaningful definition, the “Cloud” will continue to remind me of the “Stuff” in that Snapple commercial.