The Case for Architecting Plug and Play

By May 17, 2017 No Comments
Architecting plug and play

The concept of vendor lock-in isn’t new—it’s always been a valid concern for IT professionals. What we’re finding is that architecting to avoid vendor lock-in is sometimes a new mindset for IT leaders. Many leaders still lean toward wanting only best-of-breed technology (okay, who doesn’t?), even if that means 17 different vendors and huge management overhead. Others are partial to the idea of having everything fall under one vendor for the convenience of a “single ecosystem”, ease of management and (hopefully) the heightened service levels it might deliver. In either case, they look at the technology as a monolithic, self-contained stack.

To be clear, we aren’t saying that IT leaders want vendor lock-in. After all, who wants to be tied to one vendor’s feature set, or rely on their roadmap to improve your capabilities? Or, what happens if you need to switch to another tool or vendor down the road because your original provider raised their prices or can’t deliver the solution you need? One byproduct of that thinking happens when migrating IT services to the cloud. The idea of vendor lock-in is enough to put many companies at a standstill. So, they look to the vendor to provide a cloud offering to drop into their relative environments instead of trying to build a new service or solution to fit the needs of the business.

We’re here to tell you that you can have your cake and eat it too. Yes, you can have the best technology. And no, you don’t have to be locked into any one vendor. Here’s what we mean.

The answer: architect for plug and play

The problem with vendor lock-in lies in the potentially massive expense you’ll have to shell out if and when you decide to migrate from one vendor to another. The solution: think ahead, at the design and architecture stages, and make smart decisions that allow for more flexibility as your needs change. Think of your capabilities as modules you can call on, and add a layer of abstraction before that capability.

Here are a few examples of how you can change your approach to avoid vendor lock-in:

  • Go with industry standards. Don’t paint yourself in a digital corner by operating only to your vendors’ proprietary standards—stick with the industry standards and you’ll have more freedom to port your application or platform wherever you choose.
  • Don’t fixate on cost. Making critical IT decisions based on the lowest cost solution is short-sighted. What may appear like an easy and affordable solution on the surface may turn itself into a complex web of interactions that can be difficult to unwind and expensive to remove. Many vendors will lure you in with too-good-too-be-true upgrade costs or service contracts that require you to use specific features. Read the fine print. Think long term.
  • Get comfortable with the idea of replicating your data. Your data is only going to continue to grow, and you’ll need lots of storage to accommodate it. Good news, storage is more competitive and affordable than ever! Invest in it and consider replicating your raw data elsewhere so you can do what you want with it, when you want. Don’t get stuck with only a normalized copy, locked into a proprietary format.
  • Build in a layer of abstraction. This is the biggie. Insert a layer of abstraction between your technology and its functionality. Architect a layer that allows you to move between technologies seamlessly. Don’t send everything straight to the vendors’ technology. Here’s an example: add a load balancer layer between your logging systems and your log management. Not only will this give you much more freedom when it comes to what technology you choose to use for log analytics (you’re the boss!) but it allows for A/B testing and gives you greater negotiation power with vendors. They know you can move away from them at any time; they need to present their best prices.

Plug and play is a core Chronos design principle

Our goal with every client engagement: ensure no clients are trapped by any vendor. We design absolutely everything with an eye for plug and play. In fact, we believe it’s the only way to be nimble and stay competitive.

Here’s a snapshot of our full core design principles:

  • Own the intellectual property (IP): Don’t use a vendor that takes your company’s IP with them. We see this happen a lot with data analytics projects.
  • Cloud first: Unless our customer is in the infrastructure business, we help them stay out of the infrastructure business.
  • If it’s not automated, it’s broken: Always consider microservices, queuing and automation at the early design and architecture phases of any project.
  • Plug and play: And of course, avoid vendor lock-in by using layers of abstraction (some of you refer to it as a “service bus”) into every aspect of your environment, application and infrastructure.

Architecting for plug and play immediately reduces your business risk and cuts through your dependencies. Dealing with a layer of abstraction, for example, gives you far more freedom to execute versus relying on any one vendor. Plus, automating your processes reduces your costs and efforts when it comes to managing your environment. It’s a no brainer. And finally, taking a microservices approach to your infrastructure architecture gives you the adaptability to achieve your goals one piece at a time, versus an old-school method like “Big Iron in the Core.” It’s a design concept that we happily debate with vendor engineers any time! We’ll touch on infrastructure microservices in a future blog.