The Data Forecast: There is No Such Thing as the “Private” Cloud!

COL02x - feature image - EDFree lunch. Unicorns. Automatic faucets that actually work. None of these exist.

“Private” Cloud also belongs on this list.

To understand this bold claim, let’s break down the argument. First, we will refresh our understanding of Public Cloud, and look at why this is such a valuable addition to the computing landscape. We’ll then dive deeper into “Private” Cloud, and why it is great marketing, but also total baloney.

Understanding Public Cloud (the cloud that actually exists)

The Public Cloud, at its most basic, is computing resources provided through a separate organization by decoupling physical hardware from the consumption of these resources. The physical resources are shared across many customer organizations, ostensibly leading to the numerous benefits outlined below.

Amazon Web Services, Microsoft, and Google provide this capability most extensively, in a generic capacity — but there are many other niche and market challengers carving out a position in the Public Cloud space.

In last month’s Data Forecast, we looked at some specific reasons that organizations should be doing data analytics in the Public Cloud. Go read that article for more details, but the value proposition of Public Cloud is largely attributable to:

  • Power
    • Public Cloud provides services that can be consumed a la carte, letting customers lease more powerful systems than they would be capable of outright purchasing on their own.
  • Cost
    • Public Cloud providers serve many customers, achieving economies of scale far beyond what any individual customer could on their own. The lower cost to the providers lead to lower costs for the customer, and the customer no longer needs as many personnel performing infrastructure-related roles.
  • Scalability
    • Public Cloud providers buy computing hardware at massive scale, with sufficient headroom for clients to be able to limitlessly scale their consumption of computing resources.
  • Security
    • Public Cloud has dedicated teams of the most capable security personnel in the world to secure their physical and virtual assets. Most customers will never be able to secure owned hardware in a comparable way.
  • Flexibility
    • Public Cloud allows for automated, instant provisioning of computing resources to align with consumption needs of their customers. This leads to new kinds of event-driven and serverless workloads that would not be feasible with on-premises environments.
  • Speed to Delivery
    • Public Cloud brings together all of the above values in an ecosystem made for speed. Instead of spending weeks or months researching the best path, customers can try things out, fail fast, and learn quickly with rapid iteration.

Public Cloud has effectively created a utility model of computing consumption. Just like electricity, water, and gas — computing power is now accessible through our existing Internet connections, and for most organizations that’s exactly where they should get it!

The Broken Promise of “Private” Cloud (the imaginary one)

“Private” Cloud is intended to mimic the Public Cloud, but resides within customers’ individual data centers. It provides some of the same interface mechanisms of the Public Cloud, but obviously without the benefits driven by sharing across other organizations.

“Private” Cloud is sold as the best of both worlds. It claims to give the customer all of the control they desire, with most of the benefits of Public Cloud.

These claims are simply untrue.

Let’s break it down:

  • Power – NO!
    • “Private” Cloud implies that the computing hardware is owned, or is fully leased, and resident in a customer data center. So the customer has access to the purchased power, no more and no less — the way it already is in a fully on-premises solution.
  • Cost – NO!
    • “Private” Cloud requires the same high-cost structures of on-premises solutions, because it actually is an on-premises solution. Customers may even need more personnel because of the added overhead to support the cloud-like interfaces.
  • Scalability – NO!
    • “Private” Cloud customers have access to what hardware they have. If they want more, they need to procure more, find data center space, wire the racks, install the hardware, provision access, ensure security, etc.
  • Security – NO!
    • “Private” Cloud customers are in charge of their own fate. In a world where thousands of organizations fall victim to ransomware on a long-ago deprecated OS, doing security work is not a core competency for most organizations.
  • Flexibility – NO!
    • “Private” Cloud provides some flexibility – like behavior on the micro-scale – so long as the capabilities have already been built previously, the computing capacity exists, and the interface to those capabilities have been sufficiently developed. But this is “fake” flexibility: building and paying for excess capacity in order to have the illusion of flexibility is not actual flexibility. It is inefficiency posing as flexibility.
  • Speed to Delivery – NO!
    • “Private” Cloud has none of the above benefits of Public Cloud, and therefore drives no actual speed to delivery. Failing fast, rapid iteration, trying out the new tools that didn’t even exist yesterday — none of these are truly possible in the “Private” Cloud.

The only rational conclusion to this analysis is that the “Private” Cloud is not a cloud at all.

It is a repackaging of the old ways with new terminology, promoted by legacy organizations that are trying to keep their cash cows relevant long past when the steaks should have been served.

Do not fall prey to this clever marketing. Do not try “Private” Cloud as a proof-of-concept on whether you should move to Public Cloud. You will receive virtually nothing of the benefits of the Public Cloud, because your “Private” Cloud has none of the benefits – because there is no such thing as the “Private” Cloud.

Good luck — and until next time, go make an impact!

Share

submit to reddit

About Anthony Algmin

Anthony J. Algmin was the first Chief Data Officer for the Chicago Transit Agency and is now Chief Data Officer for Uturn Data Solutions, a Chicago-based consultancy that helps companies use data and cloud technologies to get better at what they do best. For more information on OpenGrid or Uturn Data Solutions, contact Anthony at aalgmin@uturndata.com.

Top