the Future of Work
Utility v. Opportunity
by Danny Kolke
published on: 01/28/08 4:55:03 PM
A conversation with a partner sparked more thoughts about the shifts going on in the application marketplace. They asked, "So, Etelos is now in the utility computing business?" My immediate reaction was to stare at them in amazement.
The answer, of course, is not at all. Etelos is in the "Opportunity Computing" business.
What is utility computing?
Utility computing is gaining momentum as an accepted term in the industry. In fact, people are now interchanging the words "the cloud" along with it and many other things. But what really is Utility Computing? Here is what Wikipedia says:
Utility computing (also known as
on-demand computing) is the packaging of
computing resources, such as computation and storage, as a metered service similar to a physical
public utility (such as
electricity,
water,
natural gas, or
telephone network). This system has the advantage of a low or no initial cost to acquire hardware; instead,
computational resources are essentially rented. Customers with very large computations or a sudden peak in demand can also avoid the delays that would result from physically acquiring and assembling a large number of computers. Conventional
Internet hosting services have the capability to quickly arrange for the rental of individual servers, for example to provision a bank of
web servers to accommodate a sudden surge in traffic to a web site...
... "Utility computing" usually envisions some form of virtualization so that the amount of storage or computing power available is considerably larger than that of a single time-sharing computer. Multiple servers are used on the "back end" to make this possible. These might be a dedicated computer cluster specifically built for the purpose of being rented out, or even an under-utilized supercomputer. The technique of running a single calculation on multiple computers is known as distributed computing.
... The definition of "utility computing" is sometimes extended to specialized tasks, such as
web services.
In my opinion, utility computing is the combination of:
-
Infrastructure on demand (storage, cpu, etc.); with
-
Cheap hosting: pay for what you need and when you need it; that is
-
Reliable and always available; and sometimes with
-
Other services you may need, (if they got 'em) that you can buy cheap.
What's great about utility computing?
The utility computing model gets you out of managing infrastructure elements like database clustering, bandwidth metering, failover, etc... It also allows you to scale as you go, paying incrementally for your usage. This is a great model, but is only part of what you need to be successful as a business.
Sidebar: All true but there are dangers. You aren't immune to bad database design and endless loops in a utility computing model. If you are no good building your apps, it's still going to hurt. If you are under the impression that you can design a really bad app and scale it; that isn't necessarily true. Especially if you are building inside a virtualized container model. I remember some of the early tests I did with virtualization where we kept running out of memory even though the loops weren't endless.
What's bad about utility computing?
In reality, I can't say there are a lot of bad things about utility computing at all. I think what Amazon is doing is great as an example. I think it becomes dangerous if you are locked down/locked in. To use Amazon again as an example, the earliest models at Amazon didn't really give you a persistent database in their hosting. So you had to engineer your solution around that.
As you go about looking at utility computing options, what's potentially dangerous is the limitations that you choose in the various models. What are your constraints? Can you code in the language you like? Can you customize the tables in the database? Do you have flexibility and portability? If you can live with all those things, then utility computing is great.
But perhaps the real question is what's missing in the utility computing model you are investigating? That is where I think the new paradigm of "opportunity computing" begins.
What is Opportunity Computing?
Opportunity computing enables me to recognize an opportunity for my business and respond to it faster than ever before. Opportunity computing encompasses utility computing. But it also gives you the tools, partnerships, API support and go-to-market capability that you need to be successful as a business.
Opportunity computing is designed for distributing applications in a software as a service model. Does utility computing have the tools built-in that enable that? Do they have user management, account management, licensing enforcement, billing, marketing, monetization support for free distribution?
Figure 1: The stack in Opportunity Computing consists of:
Platform->Accounts->Users->Services->Apps
You see, opportunity computing is really a superset of utility computing.
With Opportunity Computing you get the on-demand infrastructure you need to allow you to scale as you go, (utility computing) but you also get tools to rapidly build out apps and integrations based on that infrastructure. This means you can very quickly respond to market opportunities without having to re-invent things like accounts, users, groups, etc for every application. You automatically get the ability to synchronize data across apps and, most importantly, the ability to go to market with your app immediately. All this without having to write to a specific set of proprietary APIs.
Additionally, you now have access to opportunities through partnerships that you never had before. With the new application sync capabilities of Etelos Application ServerTM, Version 6, potentially any app in the Etelos MarketplaceTM can share information with any other app, enabling portability of data and integration at the data layer without explicit APIs. For example, I can move to the Etelos platform, set up the sync thing, and now all of the other applications users are potentially my users. This enables an opportunity for cross marketing, pollination, etc. among a superset of users and accounts, empowers the users to chose best of breed solutions in a plug and play SaaS way.
Let's look at the tools available in Utility vs. Opportunity computing. When you start looking at Etelos and what's different about our stack, the difference is we've added an extra layer for opportunity.
Figure 2: The tools for Opportunity Computing consist of:
Account Provisioning->User Licensing-> Usage
Licensing ->
Billing -> API Sychronization -> APP
Synchronization ->
Standard Web Services -> App Functionality
-> App Customization ->
First Level Support/Tools -> Marketing
Tools ->
Partnership Programs -> Ad Platform
At first glance, this seems like a lot of tools and services. It is. These are all necessary solutions that each developer of Web-based solutions must address in their development and business operations. At each major level, you are addressing what these mean and what to do with it.
In your model, how do you provision accounts? What's a user? Do you charge for usage and how is that enforced?
What do you synchronize with and what do you support? What services to you integrate with? Do you support customizing your application, and if so, how do you deal with code branches?
Not to mention, once you are through all of these things; you still have to market your application. And "Marketing," with Utility Computing, you are still on your own. You have to figure stuff out on your own. In Opportunity Computing, you are part of an ecosystem that is vibrant and figuring stuff out collaboratively.
Partnerships.The net-net is that Opportunity Computing creates an environment for expansion through partnerships. It enables you to partner big and drive adoption.
Again, much more to say on this subject and must cut it short today.
More soon.
Cart
Account
