Docker and other container services are appealing for a good reason – they are lightweight and flexible. For many organizations, they enable the next step of platform maturity by reducing the needs of a runtime to the bare essentials (at least, that’s the intent).
When you dig into the benefits afforded by containers, it’s easy to see why so many companies have started projects to:
The software development pattern of small things, loosely coupled, can go even further with an architecture built around containerization.
However, I’ve discovered that there is no shortage of misunderstandings about Docker (no surprise given the rapid growth and pace of change) and other container services in terms of:
Containers certainly offer plenty of benefits, and it makes good sense to explore whether and how they could work for your organization. But it is also a good idea to take off the rose-colored glasses first and approach this technology realistically.
Many organizations today are running tons of AWS instances to spin up new apps, services, databases, etc., and otherwise grow their businesses. While it’s super simple to scale, they’ve realized this comes with various types of overhead:
This can lead to sprawl, inconsistencies in core images, and process and budgetary challenges. Finance wants to know how Ops teams are modeling their growth and spend, security teams are trying to keep their arms around the growth to ensure that the organization is meeting the goals of its security strategy, and engineering wants to know they can get the flexibility they need to deploy new components quickly. At the same time, the business needs to grow.
So a key question becomes: How can we optimize our processes, optimize our AWS environment (to save money), and still do what we need to do?
Docker — at least on the surface — appears to offer an answer to this conundrum.
One common theme I’ve heard from organizations with a lot of AWS instances is the thinking that they can reduce the number of raw instances by increasing the size of individual instances and running containers.
For example, if you have 600 AWS instances that are 1 CPU and 4–5 GB of RAM each, maybe you’re thinking you could use Docker containers to reduce that to 100 instances, with 32 CPUs, and 64GB of RAM. Then you can significantly reduce your AWS costs, since you’ll have fewer instances. Great, right? Well… It’s not that simple.
In the short term, the shift described above may work for some use cases. But in the long term, as with many technology choices, you’re trading one set of complexities for another. Why?
A New Tech Stack
Well, as soon as you start to run containers at scale, you need to invest in a management and orchestration platform to manage your containers and their resources. This requires a whole tech stack of its own.
And because container usage patterns are still a relatively new subject, there aren’t a lot of best practices to rely on, so figuring out the strategy will mean continuous iteration towards production and organizational buy-in that this may impact delivery schedules, depending on the implementations.
Obstacles with containers include the following:
As far as that third bullet, today, I am seeing a lot of “Docker rationalization.” By this I mean that a lot of organizations are moving to containers because they can, and figuring out the use cases as they go along. This isn’t an inherently bad thing, but when it comes to determining the impact to your platform’s availability, security, and cost-efficiency, it’s better to lay out a clear set of use cases with goals ahead of time.
While it may make sense on the surface to move your workloads to containers, the devil is always in the details. You need visibility into what exactly is going on in your container, when, and from where. Security can be really challenging when it comes to containers, because there just aren’t tried-and-true best practices that you can rely on yet.
As you scale, you’ll want to understand how to exert appropriate controls over the images you’re using, how they’re built, and the scope of access provided to processes. You should know the answers to questions like these in advance:
Layout out clear answers to these questions at the beginning will help tremendously in keeping your implementation clean and clear.
Companies are opening up about the challenges they are facing with the complexity of containers, and we hear all the time from people who are running into difficult questions such as:
It is important to keep in mind that while Docker can indeed help you run faster, having a powerful engine doesn’t get you far if you don’t have the rest of the car built to support it.
It’s clear that containers are a huge part of the future of cloud infrastructure, however, like any other technology, it should be approached with a healthy mix of optimism and skepticism. If you properly optimize your processes, you’ll be able to fully take-advantage of everything Docker (and other containers) have to offer.
By Chris Gervais