The application container market is set to grow from $762 million in 2016 to $2.7 billion by 2020. This according to 451 Research’s latest Cloud-Enabling Technologies Market Monitor report. Despite only being a small portion of the overall cloud technologies market, application containers will see the hottest growth, estimated at 40 percent through 2020.
Why? Well, it’s hype mixed with need, with a bit of success on top. Containers have a valid place in the cloud computing technology stack, beyond the hype. In other words, they solve a core problem facing those moving applications to the cloud or building net-new in the cloud: portability, scalability, openness, and consistency.
However, they are not the solution for everything. The largest problems I see with containers and container orchestration (Kubernetes), are the misapplications of this technology. Let’s look at three issues:
First, application architecture is key. You can certainly shove code into a container and get it running, but containers work best when the application architecture is created or changed around the notion of containers.
Containers are, in essence, distributed and processing-oriented. Typically, in order to use containers in an optimal way, you need to be able to change or even break apart applications. Moreover, if your applications are tightly coupled to the data, unless you’re willing to decouple data from the app, you’ll not find much success with containers.
Second, containers cost more than traditional application development. The application changes needed to leverage containerization is part of the “container tax." This is the additional money you’ll need to spend to modify an application for containers or to build net-new applications that are container oriented. Although it is tough to set a solid number, I’ve found that the average is about 35 percent more than traditional app dev costs.
Of course, 35 percent more could well be offset by the savings you get from portability and scalability, as well as newfound agility. Your mileage will vary. I suggest you take the time to do a business case prior to jumping into containers—or any other technology for that matter.
Finally, your success with containers depends on having the right container talent. This should go without saying: Organizations moving to containers need to understand that best practices are still forming, and it’s easy to do containers incorrectly if you don’t know what you’re doing.
To avoid this issue, splurge on hiring and splurge on training. You should do a skills gap analysis if you can, and create a training and hiring plan to close any gaps.
Hopefully, this helps some of you avoid false starts with containers.