{"id":4848,"date":"2018-03-19T09:17:21","date_gmt":"2018-03-19T16:17:21","guid":{"rendered":"https:\/\/www.couchbase.com\/blog\/?p=4848"},"modified":"2020-11-12T11:37:08","modified_gmt":"2020-11-12T19:37:08","slug":"kubernetes-operators-game-changer","status":"publish","type":"post","link":"https:\/\/www.couchbase.com\/blog\/kubernetes-operators-game-changer\/","title":{"rendered":"Why Kubernetes Operators Are a Game Changer"},"content":{"rendered":"

The whole web developer community is\u00a0excited about Kubernetes (K8s). No wonder it is the hottest topic at the conferences and developer events that I have been to the last year.<\/span><\/p>\n

It is not just a tool for managing containers, in fact, K8s allows you to easily add <\/span>load balancing <\/span><\/i>and avoids the need of a <\/span>service discovery layer<\/span><\/i> (you no longer need <\/span>eureka<\/span><\/a> for instance). \u00a0K8s also automate application deployments and updates, and most importantly, let you plug\/write custom controller for your infrastructure.<\/span><\/p>\n

Fantastic, right? But managing stateless containers isn\u2019t that complicated, after all, they are essentially ephemeral and you can kill and spin up instances as you wish without any considerable side effects. <\/span><\/p>\n

But this is half of the story, in general, applications can\u2019t be entirely stateless, and in most cases, we simply push the state to the layers down below. So, how do we deal with stateful applications in K8s? Luckily, since version 1.5 there is something called StatefulSets.<\/span><\/p>\n

 <\/p>\n

Stateful Containers<\/b><\/h2>\n

Kubernetes StatefulSets gives you a set of resources to deal with stateful containers, such as: volumes, stable network ids, ordinal indexes from 0 to N, etc. Volumes are one of the key features which allow us to run stateful applications on top of Kubernetes, let\u2019s see the two main types currently supported:<\/span><\/p>\n

 <\/p>\n

Ephemeral storages<\/b>\u00a0v<\/strong>olumes<\/b><\/h3>\n

The behavior of ephemeral storages is different than what you are used to in Docker, in Kubernetes the volume outlives any containers that run within the Pod and the data is preserved across container restarts. But if the Pod gets killed, the volume is automatically removed. <\/span><\/p>\n

 <\/p>\n

Persistent storage volumes<\/b><\/h3>\n

In a persistent storage, as the name suggests, the data lifetime is independent of the Pod\u2019s lifetime. So, even when the Pod dies or is moved to another node, that data still persists until it is explicitly deleted by the user. In those kinds of volumes, the data is stored typically <\/span>remotely<\/b>.<\/span><\/p>\n

 <\/p>\n

We are looking forward to Kubernetes support to <\/span>Local Persistent Storages<\/span><\/a> as it will definitely be the best fit for running databases, but in the meantime, we use ephemeral storages by default. At this point, you might wonder why we are using ephemeral storages instead of the persistent ones. Not surprisingly, there are many reasons for that:<\/span><\/p>\n