At the @Cloud Connect event this week in Silicon Vally, CA, writer and cloud economics guru Joe Weinman, who is also VP of Strategy and Business Development at AT&T, discussed the inevitability of the cloud and its benefits.
A few years back, Weinman coined the term “Cloudonomics,”a treatise on why the cloud makes sense economically. I’d like to repeat these 10 commandments, rather, laws, that Weinman came up with in this space because I think many of us in the IT industry need a refresher from time to time. Perhaps we get too caught up in the insecurities of cloud computing and we forget about the benefits.
After all, if there were no benefits to cloud computing, we wouldn’t be insecure about them, would we?
What enables the “sustainable strategic competitive advantage of clouds” as Weinman puts it?
- On-demand services found in a common pool of “dynamically allocated resources”
- The huge scale of large cloud providers
- Reduction of costs via dispersion – versus the consolidation and concentration of enterprise data centers.
But it’s the 10 Laws of Cloudonomics that really fine tune the economic benefits of the cloud. They are:
#1: Utility services cost less even though they cost more. While cloud providers may charge a higher cost per unit time than if it were owned, financed or leased, they cost zero when they’re not used. So customers save by replacing fixed infrastructure with clouds when workloads are spiky, “specifically when the peak-to-average ratio is greater than the utility premium.”
#2: On-demand trumps forecasting. Cloud providers can rapidly distribute and provide capacity when it’s needed, and that means your business earns money, e.g., from higher web sales. On the other hand, forecasting is “often wrong.”
#3: The peak of the sum is never greater than the sum of the peaks. Enterprises deploy capacity to handle their own peak demands, for example, a retailer for Black Friday. Done this way, the total capacity that gets deployed equals the sum of these individual peaks. But clouds can reallocate resources across many enterprises with different peak periods. So that means a cloud provider winds up needing to deploy less capacity overall.
#4: Aggregate demand is smoother than individual. Aggregating demand from multiple customers tends to smooth out variation. So clouds get higher utilization, and that means better economics.
#5: Average unit costs are reduced by distributing fixed costs over more units of output. Big enterprises benefit from economies of scale but larger cloud providers can benefit from even greater economies of scale, such as volume purchasing, network bandwidth, operations, administration and maintenance tooling.
#6: Superiority in numbers is the most important factor in the result of a combat. This comes from the classic military strategist Carl von Clausewitz. He argued that, above all, larger armies were key to winning battles. When it comes to clouds, it’s the botnets attackers versus DDoS defenders. A botnet of 100,000 servers, each with a megabit per second of uplink bandwidth, can launch 100 gigabits per second of attack bandwidth. “An enterprise IT shop would be overwhelmed by such an attack, whereas a large cloud service provider — especially one that is also an integrated network service provider — has the scale to repel it.”
#7: Space-time is a continuum (Einstein/Minkowski) Computing tasks can often trade off space and time, for example a batch job may run on one server for a thousand hours, or a thousand servers for one hour. But a Google query is fast because processing is divided among a lot of CPUs. Yet, a cloud provider can offer unbounded on-demand scalability.
#8: Dispersion is the inverse square of latency. Reduced latency is so important these days. It means everything to effectively delivering such services as rich Internet applications, online gaming, remote virtualized desktops, and interactive collaboration such as video conferencing. However, to cut latency in half requires not twice as many nodes, but four times. It’s more economical for clouds to deploy more nodes than an enterprise.
#9: Don’t put all your eggs in one basket. No finite quantity of data centers will ever provide 100% reliability, but we can come “very close to an extremely high reliability architecture” with only a few data centers. Cloud providers that want to guarantee high availability services globally for latency-sensitive applications must establish a few data centers in every region. So look for cloud providers so equipped.
#10: An object at rest tends to stay at rest (Newton). A data center is a very, very large object – usually stuck in one place. However, a cloud service provider can establish virtual sites optimally.
Good stuff! Thanks Joe.