I am trying to find a solution to make use of the same Amazon EFS for mounting multiple directories in the Kubernetes deployment. Here is my use case
We deploy the applications as a Kubernetes Pod in the Amazon EKS cluster. If i am using the same EFS for both the above mounts, i can see all the files from both the directories "/opt/templates" and "/var/logs" as i am using the same EFS.
How can i solve the problem of using same EFS for both the application without seeing app1 mounted files in app2 directory ? Is it even possible of using the same EFS ID for multiple applications ?
Here is the Kubernetes manifests i used for for one of the application which includes PersistentVolume, PVC and the Deployment
----
apiVersion: v1
kind: PersistentVolume
metadata:
name: efs-pv-1
spec:
capacity:
storage: 2Gi
volumeMode: Filesystem
accessModes:
- ReadWriteMany
persistentVolumeReclaimPolicy: Retain
storageClassName: efs-sc-report
csi:
driver: efs.csi.aws.com
volumeHandle: fs-XXXXX
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: efs-pvc-1
spec:
accessModes:
- ReadWriteMany
storageClassName: efs-sc
resources:
requests:
storage: 2Gi
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: deploy1
spec:
replicas: 1
selector:
matchLabels:
app: deploy1
template:
metadata:
labels:
app: deploy1
spec:
containers:
- name: app1
image: imageXXXX
ports:
- containerPort: 6455
volumeMounts:
- name: temp-data
mountPath: /opt/templates/
volumes:
- name: shared-data
emptyDir: {}
- name: temp-data
persistentVolumeClaim:
claimName: efs-pvc-1
It looks like you can do that by including the path as part of the volume handle.
A sub directory of EFS can be mounted inside container. This gives cluster operator the flexibility to restrict the amount of data being accessed from different containers on EFS.
For example:
volumeHandle: [FileSystemId]:[Path]
I think you will need to create two separate PVs and PVCs, one for /opt/templates, and the other for /var/logs, each pointing to a different path on your EFS.