Volumes

============

everything in k8 is ephemeral. pods, nodes, master node all are ephemeral. So it is not good to store DB inside K8

Storage administration --> manage all the disks related to servers

Backup every day

Restore testing

N/w to storage

EBS -> Elastic block storage

EFS -> Elastic file system

If server is in AZ us-east-1b, EBS also should be here. EBS is less latency. EBS is well suitable for DB and OS

1. node and EBS volume should be in same AZ

2. drivers install

3. EC2 worker nodes should have permission to work with EBS volumes

static and dynamic provisioning

=================================

EBS static

==========

we have to create disk manually. vol-06237d8b3183eefc9

drivers install

kubectl apply -k "github.com/kubernetes-sigs/aws-ebs-csi-driver/deploy/kubernetes/overlays/stable/?ref=release-1.40"

everything in k8 is resource/object. if you can treat or manage the volumes from k8 objects any k8 engineer can work on this..

PV --> Persistant volume, represents or wrapper of physical volume

PVC --> Claiming the volume

Storage class

kid --> mother --> father --> wallet

Pod --> PVC --> PV --> volume

access modes

============

ReadWriteOnce, ReadOnlyMany, ReadWriteMany, or ReadWriteOncePod,

Lifecycle policies

=================

if pod schedules in 1a and volume is in 1b, pod status will be in pending for continously.

scheduler --> schedules your pods on to appropriate nodes