The evil that lurks in the intelligent network

The evil that lurks in the intelligent network

Summary: The spectre looming over the convergence of Web 2.0 and enterprise SOA is a future in which network operators control the intelligence in the network.

TOPICS: Networking

The nightmare scenario that flows from the convergence of Web 2.0 with enterprise SOA (the "Global SOA"?) is that network providers regain their monopoly control of our access. Network equipment vendors are already eager to sell them the kit to do this. Here's a PR pitch from Solace Systems, a vendor that markets XML appliances to network operators (my emphasis added):

"As the need for system interoperability increases, XML- and SOA-enabled infrastructure becomes the standard for next-generation networks. The idea that the Internet can become a massive SOA infrastructure on-demand is becoming a reality."

Writing last month on The Register about Cisco's AON product line, which offers similar capabilities, Tom Welsh outlined the dark side of this proposition (again, with my emphasis added):

"... it’s by no means clear that the rest of us should uncritically welcome 'putting intelligence into the network' ... there is a downside to the AON dream. At first glance, three serious issues raise their heads: security, unfair competition, and the potential demise of the internet as a content-neutral medium ... Before we know it, we could be back in the AOL/Compuserve universe – paying extra for every piece of information and every 'special' service."

It's thoughts like these (along with Doc Searls' November essay about Saving the Net: How to Keep the Carriers from Flushing the Net Down the Tubes) that make me nervous when people start talking about Web 2.0 and the on-demand space in terms of a utility model like power or water. As Dan Farber pointed out early last year, those utility markets have transparent pricing, whereas "an industry standard definition of CPU per hour usage doesn't exist. There is no equivalent to kilowatt hours." ComputerWorld's Mark Hall recently spelt out the implications of that rather more forcefully (my emphasis added):

"For utility computing to fulfill its promise, there has to be a standard pricing model that all users can apply to their operations. Until then, on-demand computing will be just another complex, proprietary pricing strategy vendors use to keep you from fairly and accurately comparing one service to another."

This is the spectre of evil that looms over the convergence of Web 2.0 and enterprise SOA — a future in which network operators control the intelligence in the network, giving them the power to set proprietary tariffs not only for their own utility services but also for any rivals you might wish to subscribe to over their network.

I promised yesterday, when I asked Who will rule Web 3.0?, to explore some examples of "either fragmentation or concentration of economic opportunity and power." Tomorrow I will move on to examine a happier, more fragmented potential landscape. But it's one we won't be free to enjoy unless we maintain the clear and absolute separation of network access and networked functionality that the Web enjoys today.

As soon as my access provider can start dictating which services I can and cannot use and how much I pay for them, every benefit I expect to gain from Web 2.0 and service-oriented architectures is negated. Bear that in mind the next time some network provider offers you a tempting bundle of network access and applications. There's a catch. Oh boy, is there a catch.

Topic: Networking

Phil Wainewright

About Phil Wainewright

Since 1998, Phil Wainewright has been a thought leader in cloud computing as a blogger, analyst and consultant.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Spot on

    Mr. Wainewright correctly points out a significant danger of the SOA, SaaS, and Web 2.0/3.0 situation, which is that when monetized, pricing is Byzantine at best. Mr. Wainewright and I have carried on a nice series of blogs and comments in the recent past regarding the dangers of third party vendors (TPVs), in a nice Hegelian way. We both ended up in a middle ground (, which is that none of this will work without massive amounts of transparancy on the TPV's part to enforce transparancy.

    Sadly, the pricing models behind these types of services simply don't work for either the TPV or the customer. Let's try examining a few:

    * Per-user usually hurts the vendor. Without an adequete user limitation scheme (look at the lengths Microsoft goes to just to keep only one user using a copy of Windows!), cusomters have a bad habit of forgetting (accidentally or on purpose) just how many users they need to buy licenses for.

    * Per-byte transferred runs into a severe rediculousness factor. It is nearly impossible for the customer to gauge just how much they will be using when signing a contract until they actually start using the service. In addition, there's the "bathroom light" factor: when the bill isn't paid out of the user's pocket, they have a tendency to use a lot more than they need. This is bad for the customer.

    * Per-CPU hour is, as Mr. Wainewright points out, nearly impossible to measure and properly value, especially before signing a contract.

    Mr. Wainewright also points to a disturbing possibility, one in which network carriers sit astride your connection and limit your usage or charge you extra for certain types of traffic. As it is, carriers are considering allowing websites to pay them to give traffic to/from those sites increased priority. I am sure that consumer oriented providers (DSL by the Baby Bells, cable companies) that offer VoIP are thinking about whether or not they can get away with limiting or preventing users from using third-party VoIP providers like Vonage. If one reads their contract carefully for home use, there are all sorts of unpleasant gotchas in there. Mine, for example, *mandates* the usage of anti-virus software. Theoretically, if they detect that my computer is sending out virus generated spam, they could cut me off. That sounds well and good, until you realise that they could also detect other types of traffic for ToS violations, or even to push me towards their own service in a particular way.

    Remember, that with the execption of high-priced connections, you really don't have an SLA on service. For example, even a T1 line typically doesn't give you an SLA on uptime unless you opt for the (quite expensive) managed service add-on. With no uptime guarantee, what is your recourse it your connection seems to get bumped or the speed drops everytime you start sending a lot of XML data or VoIP traffic?

    The only protection possible against these types of problems is to bring as much of these services in house. I predict that SaaS and Web 3.0 will be monetized by vendors selling specialised applications to customers to place within their own networks. That is the best way for everyone to be happy, and to keep the network carriers (remember: they are the spawn of AT&T/Ma Bell, have had a 100 year monopoly, and act like it!) out of the loop.

    Justin James
  • Pick your single point of failure

    A server on your platform or the internet connection to your SOA apps.

    In terms of network providers taking control over you - there is an alternative. The WiMAX Steamroller will allow for the MESH, which (eventually) can be turned into universal free access (free from ISPs).
    Roger Ramjet
    • WiMAX solves what exactly, and how?

      I've been hearing a lot of posts about how WiMAX is going to knock network carriers out of the loop. How exactly is this going to happen? WiMAX is MAN level technology, at best. It is NOT WAN connectivity. WiMAX is NOT going to connect LA to NYC, or even LA to Oakland. WiMAX is merely a very effective last mile.

      How does it free people from the entwork carriers? Do you think that those WiMAX access points don't get connected to something (backbones, for example) with a copper wire at some point? What about the people who cannot afford or who are not allowed by either company policy or by law to have their data being broadcast? What about the people who are so security concious, they need to use STP and/or fiber instead of UTP? What about the people who live in "metropolitan areas" that still have dirt roads (I live in Columbia, SC, the state capital, and we still have quite a number of dirt roads within city limits)? I doubt that their local governments are going to be setting up WiMAX anytime soon. Or the people who live in areas where they beleive that Internet access is NOT the responsibility of the government to provide? No, WiMAX is not going to kill the network carrier and their copper/fiber lines anytime soon.

      Don't get me wrong, WiMAX is a very cool technology, and I love the potential for what it can do and what it does. But eliminating the network carriers? I don't think so, not by a long shot.

      Justin James