This blog appeared first at the Amalgam Insights blog.
The past few years have been exciting ones for containers. All types of tools are available and a defined deployment pipeline has begun to emerge. Kubernetes and Docker have come to dominate the core technology. That, in turn, has brought the type of stability that allows for wide scale deployments. The container ecosystem has exploded with lots of new software components that help maintain, manage, and operate container networks. Capabilities such as logging, load balancing, networking, and security that were previously the domain of system-wide software and appliances are now being brought into the individual application as components in the container cluster.
Open Source has played a big part in this process. The Cloud Native Computing Foundation, or CNCF, has projects for all things container. More are added every day. That is in addition to the many other open source projects that support container architectures. The ecosystem just keeps growing.
Where do we go from here, at least through 2019? Pretty much on the same path. 2019 will be a year for rounding out and growing container technology to make it more palatable to large enterprise applications. With the basic technology done, the work to make container networks secure, resilient, and manageable will be the primary focus for containers.
That doesn’t mean there won’t be new and exciting technology added to the container ecosystem. Serverless computing, which has been built on containers, will now itself be turned into a container technology. The KNative project to create serverless computing in a Kubernetes cluster is an example of interesting development that needs to be tracked over the coming year. For many developers, having to deploy container clusters is, in of itself, too much work. They would prefer a new level of abstraction that allows them to ignore all the workings of the cluster and just write code. KNative might just do that.
Another area to watch will be hardening containers. While the capacity utilization of containers is better than virtual machines, they also less safe than VMs. Inhibiting the ability of code running in a container to access the host operating system is an interesting way to make containers more secure.
Finally, the emergence of service meshes for containers is an important development for containers and microservices. Services meshes are set of network services controlled by a central controller accessible from the container cluster. This offers the possibility for much more flexible clusters that can access centralized services that compliment the internal components of the system. Service meshes help provide a balance between centralized and localized services.
2019 is not going to be “exciting” for containers in the sense of blockbuster new technology. Instead, this is the year when the container ecosystem grows up, filling holes in container architectures. The refinement of the container ecosystem is critical to long term health in the space. It won’t be exciting but it also won’t be boring.