Kubernetes is now “boring” (that’s a good thing) and service mesh is the next big thing in containers. That’s what Red Hat had to say recently when we asked about the state of the container ecosystem.
We posed some questions to Joe Brockmeier, senior evangelist for Linux Containers at Red Hat, about the current state of containers, and what’s coming next. Here’s what he had to say.
Kubernetes is ‘Boring’
Brockmeier’s most interesting point was that Kubernets has now become “boring”—and that’s a good thing for developers who write containerized applications.
“In this industry, ‘boring’ really translates to ‘it’s ready for my business to depend on, I can build a road map around it and get support,'” Brockmeier said. “Kubernetes is boring in the dependable sense now.”
He added that Kubernetes is still exciting for its technical features. “It’s still certainly exciting technology in terms of what it can do,” But, like the Linux kernel or Apache HTTPD server, Kubernetes has now become so easy to work with that it is no longer the topic of hot conversation, he noted.
That’s a big change from just a couple of years ago, when Kubernetes was only one of several container orchestration options, and had not yet proven its mettle.
It’s worth noting, of course, that Red Hat took a bet on Kubernetes early on. Red Hat remains one of the main contributors to Kubernetes development. It is thus only natural for Red Hat to see Kubernetes as the go-to, dependable container orchestration solution.
Still, given how spectacularly Kubernetes has taken control of the container world and pushed Docker to the sidelines, it’s difficult to argue with Brockmeier. Almost everyone who uses containers seems to be using Kubernetes these days, and the platform has become boring.
Service Mesh is the Next Big Thing
The reason why is straightforward. “Running one container isn’t interesting or of great value. Running many, many containers (hundreds, thousands, and even more) is of great value,” Brockmeier said.
“Simplifying the networking between containers, enforcing policy, providing telemetry and reporting—those are of value,” and that’s where the service mesh comes in.
In the near future, Brockmeier expects to see intense interest in service meshes as they evolve. He did say, however, that they’ll likely become boring soon enough. “Service meshes aren’t boring yet, but I think we’ll get there in short time,” he said.