A Quick Introduction to Consul
For reasons that (hopefully) will become clear in the relatively near future, I decided I needed to take a look at Consul, a distributed service discovery tool and key value store. I know Consul’s description sounds like a mouthful of buzzwords, but it’s pretty accurate. This post provides a quick introduction to Consul, in which I’ll break down what Consul does and how it works (at a high level). I’ll then build on this introduction in later posts.
There’s a lot to Consul, so let’s start by breaking down the description of Consul, which I provided as “a distributed service discovery tool and key value store”. What does this mean, exactly?
- Consul is distributed. This means it runs as a cluster of systems so that there is no single point of failure. Consul uses a gossip protocol (known as Serf) to manage cluster membership, failure detection, and general orchestration. Managing cluster state via Serf is only part of the picture, though; the cluster must also manage consistency via a consensus protocol known as Raft. (Raft is the same consensus protocol used by etcd, for example.)
- Consul is also a service discovery tool. Applications can register with Continue reading
