Linkerd: Up and Running

Chapter 2. Intro to Linkerd

The year 2015 was a very good one for cloud native computing: it brought us the first Kubernetes release, the creation of the Cloud Native Computing Foundation (CNCF), and the creation of Linkerd. Linkerd was one of the first five projects donated to the CNCF, and it was the project that coined the term “service mesh.”

In this chapter, you’ll learn more about Linkerd, where it comes from, what makes it special, and how it works. We’ll keep the history lesson short, useful, and interesting, but if you want to get right to the important information, feel free to skip ahead.

Where Does Linkerd Come From?

The Linkerd project was created in 2015 at Buoyant, Inc., by former Twitter engineers William Morgan and Oliver Gould. The first public release of Linkerd was in February 2016. You can see a brief summary of its history in Figure 2-1.

Figure 2-1. A brief timeline of Linkerd

Linkerd1

That first version of Linkerd, now called “Linkerd1,” was written mostly in Scala and was largely based on the Finagle RPC library created at Twitter. It was a multiplatform mesh that supported several different container schedulers and offered a number of powerful features. However, using Finagle required Linkerd1 to run on the Java Virtual Machine (JVM), and ultimately the JVM’s performance was simply too high a cost to bear.

Linkerd1 is at its end of life. Going forward, ...