We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.
Start a Discussion
Cloud Computing

Are private clouds bound to fail?

Vote 0 Votes
In the 10 cloud predictions for 2011, number two on Mel Duvall's list at Information Management is, You will build a private cloud, and it will fail.  Do you agree, and if so, what can be done about it?

13 Replies

| Add a Reply
  • The predictions were (quote)
    "* You will build a private cloud, and it will fail...
    * Hosted private clouds will outnumber internal clouds 3:1. "

    Should we read Public clouds would outnumber... Private clouds 3:1???
    Anyway, private clouds won't fail but they'll take time to build and they won't be as cost efficient as public ones.

  • It's a very generic statement. What is meant by failure? Costs are higher than public? Consumers don't show up? SLAs are not defined or not met? Private clouds are traditionally more supportive of elasticity and utilization models needs more so than self-service and multi-tenancy. In this manner, it will allow businesses to do more with the same data center infrastructure (e.g. power, cooling, UPS, etc.) Due to this value, identifying the private cloud initiative as a failure will be very difficult.

  • Agree that "private clouds will fail" is very vague. One benefit for organizations in building private clouds is they can dictate usage of it across their organization. They may not reap the financial benefits of using a public utility, but they can overcome the adoption hurdle that plagues public clouds.

    And some organizations will most likely never move to the public cloud - i.e., government and public safety - but they will build private cloud systems that will serve them well.

  • Private Clouds always makes me think twice. What is Private Cloud? Is it a form of Cloud Computing? If an essential characteristic is indeed elasticity, and that means that a lot of companies suffer from major misconceptions when it comes to Private Cloud. Achieving elasticity with a Private Cloud is very challenging and some special circumstances.

    If you say that Private Cloud is not a form of Cloud Computing and instead the software platforms including OS, virtualization, monitoring, failover and automatic provisioning then I personally feel that the concept Private Cloud would be better off with another name. Cloud Computing is way for a provider to offer services to a consumer based on certain technologies. Private Cloud is only about the technologies (for the majority of companies that I have seen).

    If you define Private Cloud as a form of Cloud Computing very few companies indeed will be able to succeed in getting the elasticity that Cloud Computing promises (see my blog post that is linked above). If you (re)define Private Cloud to be some kind of technology then I believe that a lot of companies will be able to succeed in that they will be able to streamline their procedures and reduce the downtime of the hosting platforms / products.

  • Well, I could certainly use that broad a brush to sweep the recent snow from the driveway if nothing else.

    I'm going to agree with JP on this one - it's hard to claim failure at a task for which the criteria for success are not defined. Those who build a private cloud to achieve X, Y or Z and do not attain those goals, fail. Those who achieve their goals, succeed.

    Then again, even if don't succeed in meeting their goals are those really failures? Perhaps they follow the Thomas Edison school of thought in that they didn't fail, they found a private cloud that didn't work in achieving X, Y, or Z.

  • These kinds of all-or-nothing statements make for great blog fodder... but there are simply too many unique challenges and environments out there. I tend to look at most things from a large enterprise (5000+ employee) perspective, and it would be a failure NOT to try a private cloud in a large company with any significant IT investment.

    Often, the servers internal to the company literally outnumber the employees - you have multiple team, development, pre-prod environments, and a private cloud would be so much more efficient and agile than needing to provision capacity for each new proposed task.

  • Indeed, failure is not to even try but having said that, it won't be smooth sailing for companies building one of their own, using own resources. Trial and error thing does not make for an efficient private cloud initiative. One, it doesn't maximize early adoption benefits and worse, it will be a source of frustration for its users who will have more reasons to go the public cloud.

  • There are many legitimate reasons why various public and corporate sector organizations are attempting to recreate their data centers into private cloud environments. While many will fail, just as they did in their attempts to create highly efficient data centers in the first place, others will significantly improve the responsiveness and effectiveness of their data centers by modeling them on the best practices of the leading public cloud providers.

  • I believe that only a limited number of companies can reach economic private cloud scale.
    "Hosted private clouds" should be the same as virtual private clouds, it more likely reaches economic scale.

  • I believe that Private cloud are bound to raise for Mid-Large organisations that can afford paying for business assurance of technology going in line with business (short time saving does not have strategic perspectives). At the same time, Public cloud become the realm of predominantly small companies and accidental use by Mid-Large organisations

    Private cloud differs from Public cloud only in the right, permission and ability of the client to see all and any details of the SW deployment on the cloud side, which is unavailable in Public cloud.

    If the SW deployed in the cloud is owned by the client, the latter must be in position to command any changes in the deployment in Private cloud; if the client uses 3rd party's SW, e.g. owned by the cloud provider, the client has to have full visibility into what is going on with this SW.

  • The term failure i assume is to suggest a reduced adoption rate. According to me private clouds are here to stay, atleast for a while if not forever. There are several governance, security and SLA standards yet to be defined crisply in this "Clouded" market.
    The lack of standards will surely fuel the existence and growth of Private clouds, where control is with the organization (maybe at a higher cost)
    When these aspects of cloud mature and structure into standards with time we may see the Public cloud dominating the market.

  • Hi all,

    Way behind on this - started replying at the time, it got a bit long, I got distracted and so I ended up making it a blog post instead, lol.

    Thought I would post the link to my response back here for completeness, though.


    To summarise I think that building a private cloud now misses the point wrt the massive shift occuring in the value proposition of IT and the opportunities for business integration this will offer. In this sense I think that efforts to build private clouds will be judged retrospectively as a failure of leadership and vision - irrespective of any perceived improvements they deliver to existing estates - given that they are steeped in an old model of IT ownership rather than embracing new models based on IT access.


Add a Reply

Recently Commented On

Monthly Archives