Tuesday, 4 August 2020

Kubernetes: Storage Options

If you store any data in a container/Pod the data is available until the container/Pod live. Once Container/Pod deleted, crashed, restarted you will lose the data.

 

If a Pod encapsulates multiple containers, there may be a need to share data across containers.

 

To solve this problem, Kubernetes come up with different storage options.

a.   Pod Volumes: Used to allocate storage outside of the container

b.   Persistent Volume (PV): Using this, Pods can connect to external storage. Persistent Volume claims (PVC) are used to request access to a specified storage.

c.    ConfigMaps: Using this Pods can connect to Configuration files and variables.

 

How to define a Volume that is available to Pod?

Define the volume in spec.volumes.

 

How to mount a volume to the container?

Containers can mount the volume using spec.containers.volumeMounts property.

 

Volume Types

At the time of writing this post, Kubernetes support the following volume types.

 

a.   awsElasticBlockStore

b.   azureDisk

c.    azureFile

d.   cephfs

e.   cinder

f.     configMap

g.   csi

h.   downwardAPI

i.     emptyDir

j.     fc (fibre channel)

k.    flexVolume

l.     flocker

m. gcePersistentDisk

n.   gitRepo (deprecated)

o.   glusterfs

p.   hostPath

q.   iscsi

r.     local

s.    nfs

t.     persistentVolumeClaim

u.   projected

v.    portworxVolume

w.  quobyte

x.    rbd

y.    scaleIO

z.    secret, storageos and vsphereVolume.

 

In this post, I am going to show the demo using ‘emptyDir’ volume type.

 

An emptyDir volume is first created when a Pod is assigned to a Node, and exists as long as that Pod is running on that node. Containers in the Pod can read and write the same files in the emptyDir volume, though that volume can be mounted at the same or different paths in each Container.

 

What will happen when a Pod deleted from the node?

Data in emptyDir will be deleted forever.

 

Let’s create a Pod with two containers and mount a volume that is shared across the containers.

 

Step 1: Create a Pod.

 

podVolumesDemo.yml

apiVersion: v1
kind: Pod
metadata:
  name: my-pod
  labels:
    app: hello-world
    author: krishna
    serviceType: user-service
spec:
  containers:
    - name: container1
      image: busybox
      command:
        - sleep 
        - "300"
      volumeMounts:
        - mountPath: /container1-mount
          name: my-volume
    - name: container2
      image: busybox
      command:
        - sleep
        - "300"
      volumeMounts:
        - mountPath: /container2-mount
          name: my-volume
  volumes:
  - name: my-volume
    emptyDir: {}

Create Pod by executing the below command.

$kubectl create -f podVolumesDemo.yml 
pod/my-pod created

Query all the pods.

$kubectl get pods
NAME     READY   STATUS    RESTARTS   AGE
my-pod   2/2     Running   0          77s

Step 2: Create some files in the mounted volume of container1 and make sure that these are available to container2.

 

Let’s create two files (hello.txt, demo.ppt) in the mounted directory.

 

kubectl exec -it my-pod -c container1 -- touch /container1-mount/hello.txt

kubectl exec -it my-pod -c container1 -- touch /container1-mount/demo.ppt

 

Let's confirm that the two files are available to container2

kubectl exec -it my-pod -c container2 -- ls -l /container2-mount

$kubectl exec -it my-pod -c container1 -- touch /container1-mount/hello.txt
$
$kubectl exec -it my-pod -c container1 -- touch /container1-mount/demo.ppt
$
$kubectl exec -it my-pod -c container2 -- ls -l /container2-mount
total 0
-rw-r--r--    1 root     root             0 Jun  8 09:15 demo.ppt
-rw-r--r--    1 root     root             0 Jun  8 09:15 hello.txt

You can delete the pod by executing the below command.

kubectl delete pod my-pod

$kubectl delete pod my-pod
pod "my-pod" deleted


Previous                                                    Next                                                    Home

No comments:

Post a Comment