Skip to content

Problem statement

Embarking on building a private cloud is daunting. Cloud computing has been around for at least 15 years, but even now building your own cloud is still something organizations approach with great caution. There are several reasons for this.

Properly integrated solutions don't exist

While many cloud solutions will claim to be integrated, there are no truly integrated solutions out there. Some have historically confused hyperconverged infrastructure with integrated infrastructure. Others claim to be exhaustive but still require operators to do the heavy lifting, or to complete the final integration piece. And any segregated or unsupportable segment of the solution can (and will) endanger the entire solution.

Size doesn't matter, it's how you use it

In most cases, throwing more resources at the problem doesn't help. More resources often means greater eventual delivery burden, and this is rarely a non-linear curve, unless you're operating at the scale of a hyperscaler.

Proprietary software can be dangerous

Repatriating data from the public cloud might sound like it will inherently offer more agency and control, but there aren't many names in town when it comes to proprietary cloud software. Picking one of these often means handing over a large part of the control that your organization had been seeking from private cloud in the first place. Commercial contracts tie the solution down for years, and become exorbitantly expensive very quickly. In addition, most proprietary solutions require many different products to build a full private cloud, each priced and licensed differently.

Open source software can be dangerous

With open source, agency is rarely an issue - there are so many tools and solutions, and many claim to enable building a full stack. But while open source incredibly powerful, it also requires consolidating thousands of projects into a single solution. Many have done this - with mixed success. Such projects tend to be multi-year beasts, require enlisting the support of a significant number of crack software and infrastructure engineers, and in most cases... it all led to failure in the end, anyway.

Many products are thinly veiled professional services engagements

There are dozens of products that claim to solve the problems we've outlined above, but in reality, they're just professional services (PS) contracts used to tie other proprietary or open source software together. These will almost always be time and materials contracts with no end in sight. In the product world, some PS is sometimes necessary to scope out custom requirements and assist with deployment aspects, but products shouldn't require a majority of the cost to be PS.

The problem we see is that if:

  • you're looking to deploy a private cloud solution next week that you own,
  • that you can embark on without long PS engagements,

  • that doesn't lock you into a proprietary solution...

  • ...you're going to have trouble.