Cloud computing: here we go again

Cloud computing: here we go again

Summary: The same pattern has emerged again and again in the history of IT. New ideas emerge. Vendors develop their own approaches. Camps form to support those approaches. Eventually standards emerge. The battle over cloud computing standards and approaches is only the latest repetition of the pattern


Here we go again. Cloud computing is the newest area for the age-old pattern we've seen throughout the history of information technology. The pattern goes like this:

  • A new idea emerges. Sometimes this is merely the rejuvenation of an older concept using new technology.
  • At first a single supplier offers a useful implementation of this concept and starts to see success selling their product.
  • Other suppliers jump in with their own approach to implement that concept. The approaches are often are incompatible with one another so a customer selecting one was locked into that vendor's offering.
  • Camps form around the most successful implementations. These may be called partnerships, alliances, trade groups, etc.
  • Customers find that each group has developed worthwhile features and want to build solutions that mix and match technology from different groups. When they face difficulties, they scream for standards.
  • The vendors, making a show of listening to their customers, engage in efforts to create standards without also giving up account control in their own customer base.
  • Early versions of the standard either are too complex to implement or leave thorny issues for later efforts and, thus, are useful in only certain circumstances.
  • The industry moves on and does its best. Eventually interoperability is hammered out again.

This process has repeated time and again. We've seen battles over items such as:

  • Processor architecture
  • Memory design
  • Network media, protocols and interconnects
  • Storage media, data storage structures, protocols and interconnects
  • Operating systems, APIs
  • Programming languages
  • Application frameworks
  • Database architectures, scripting languages, approaches to handling events
  • File structures for items such as documents, presentations, spreadsheets, graphic images, databases
  • Virtual machine software architectures and on-disk formats
  • Management architectures
  • Approaches to security
  • and now cloud computing architectures, tools and services

We're seeing factions build up behind the approaches offered by Amazon, CloudStack, Eucalyptus, OpenStack, VMware and several others. Customers that select one approach are likely to find that interoperability between and among these different approaches will take careful planning and execution.

It would be wise to start out with a clear idea of what needs to be accomplished before plunging headlong into an implementation effort. This way, it would be far easier to chose the approach that best fits this set of requirements. Thoughts of interoperability should be part of this careful design.

Topic: Cloud


Daniel Kusnetzky, a reformed software engineer and product manager, founded Kusnetzky Group LLC in 2006. He's literally written the book on virtualization and often comments on cloud computing, mobility and systems software. In his spare time, he's also the managing partner of Lux Sonus LLC, an investment firm.

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
  • "New" ideas

    "Sometimes this is merely the rejuvenation of an older concept using new technology."

    Although far more often it is the rejuvenation of an older concept using new marketing.
    • Old Hat...

      Yep! Cloud wars were first mentioned in the Bhagavadgita! ;)
  • Applies to Anything

    When is a new idea ever really a new idea? Not just in technology, but in every industry, we see steps in adoption for "new ideas" that are just really old ideas in a nice white box. I guess what I'm torn about is whether or not it's a good thing.
    • Basic self interest

      Virtually our entire economy is based on it. Initially the first mover(s) try to corner the market, but then reality sets in and the players are forced to compromise.

      It is painful and wasteful in many respects, but the alternatives are probably worse, just like democracy.
  • Almost very company uses the same DBMS architecture


    This is because the theory was properly thought out long before anyone produced an implementation.
  • : )

    It will be interesting to see who will win out in the cloud wars. IMO companies such as dropbox and googledrive will not survive due to security issues and user data compromisation. I recommend everyone to check out 4sync, they offer 15gb of free storage and have excellent security records.
  • Marketing before technology

    Seems that 'cloud' computing is as much of a Marketing, and speed-to-market as it is a technology. The premise that the "Net" never goes down is faulty. Although diverse SANs and virtual networking and SAAS might be part of a solution, when a building full of people can't connect or have slow connections, the 'cloud' will be rainy...

    As most (if not all) cloud companies require the requisite EULA before you continue to load and use the software, the fine print would tell us all that security is 'best effort' and if anything is lost, well, "we're not liable, and so very sorry."
    Until security, survivability, and adequate business-level SLAs are in place, computing via cloud is dicey at best.
    • Clouds occur

      When a huge body of hot air meets a large volume of liquid.

      As, for example, when a marketing department encounters a refridgerator full of beer.

      The marketing department "interacts" with the beer and becomes "creative".
      • fascinating

        marketing chemistry is truly fascinating.
        • Until I discovered that

          I didn't really understand why marketing speak is usually completely logically incoherent.
  • Marketing: "Reality without the implementation details"

    I think that the basic problem with this phenomenon, which is related to Gartner's Hype Cycle, is the fact that on average marketing folks actually have NO IDEA what they are talking about.

    I mean, yes sure, they have the "big picture", like watching the clouds from high above, but they have no idea about how these things interact with "reality" (e.g. winds, thunderstorms, low/high pressures, rainfalls, etc).

    Actually the core approach to all Marketing seems to be "Systematically label reality as 'implementation details', so that we can do our job with as little worries as possible".

    Now, tell me about interoperability...
    I was on the BoD of the 1394 Trade Association for 2 years.
    1394 = FireWire
    There were super-solid standards to ensure interoperability and STILL we had lots of interoperability trouble. We were setting up "plugfests" where companies would bring their devices to check compliance with the standards and interoperability with each other. These event were really much more 'entertaining' than one would imagine :-)

    The closer something is to the hardware the easier it is to create interoperable standards. Which is why things like USB, FireWire, PCI, Ethernet and mobile telephony exist in the first place.

    The further something is from the "metal", like VMs or Cloud, it is totally improbable that a standard will ever emerge, and most likely a standard CANNOT possibly emerge.
    It is like trying to make a standard for native binary EXEs that run both on Windows and Linux and MacOS. I know some people have already tried to have a few glasses of wine.
    I have to admit it looks kind of appealing even to me, but the implementation "details" are a colossal nightmare.
    Dimitrios Staikos
  • Discussing interoperability

    Thanks for the article Daniel. While I think you are right in saying that interoperability between cloud computing approaches is becoming a challenge, I think it’s more that we’re moving to the next phase of this cyclical IT pattern in which we are defining standards. However this time, perhaps the issue of interoperability is being addressed earlier in the standards phase than in previous cycles. But nonetheless, the good news is that it is currently being addressed, per The Distributed Management Task Force (DMTF) and version 1.0 of the Cloud Infrastructure Management Interface (CIMI).