Skip to content

Latest commit

 

History

History
49 lines (25 loc) · 3.04 KB

storage.md

File metadata and controls

49 lines (25 loc) · 3.04 KB
Raw

Storage in Kubernetes

You can also check this, if you want: https://kubernetes.io/docs/concepts/storage/

One very important part of an infrastructure, is its storage. Even in Kubernetes, storage is important.

Kubernetes offers storage as volumes from various providers.

You can define the volumes per individual deployment, or, group the properties about the storage under storage classes, and then just use the storage class names - which makes it easier to work with it.

For the sake of explaining our storage addons, we'll try to say that storage types in kubernetes are having the following group of properties:

  • a) ephemeral or not (data is removed or not after pod is stopped/moved somewhere else)

  • b) it forces or not a specific node to a pod when first created

  • c) it forces or not a specific node when pod is recreated

  • d) storage is local to a node or not

    Examples:

    Name Ephemeral? Forced node @first Forced node @later Local2Node
    awsElasticBlockStore, cephfs, glusterfs, nfs no no no no
    emptyDir yes no no yes
    hostPath no no (but you probably want it) no (but you probably want it) yes
    local no no yes yes
    Our addons:
    local-storage-class no no yes yes
    nfs-client-provisioner no no no yes, to master

Each type of storage has its Pro and Cons, of course.

With your local LXD clusters you don't have limits on what to use, but to make it easier for you to start working on what is more important to you, we provide some storage addons.

If you want to get access to the data stored in nodes ('Local2Node' == yes), you either:

  • enter the container and check it,
  • or, you can mount the directories from the nodes to your real host machine so that you don't need to run commands like lxc shell mysite-worker-2 This is what local-storage-class and nfs-client-provisioner are doing.