HOW TO TAKE PVC BACKUP IN LINODE?

Install velero + restic

• Download velero binary (https://github.com/vmware-tanzu/velero/releases)

• untar velero binary using tar -zvxf filename

• sudo mv velero-v1.8.1-linux-arm64/velero /usr/local/bin/ or set path to the folder which you have untar on second step

• velero help (to check if velero is installed or not)

• install velero + restic to the cluster

• velero install \
  --provider velero.io/aws \
  --bucket LINODE_BUCKET_NAME \
  --plugins velero/velero-plugin-for-aws:v1.0.1\
  --backup-location-config s3Url=https://us-southeast-1.linodeobjects.com,region=us-southeast-1 \
  --use-volume-snapshots=true \
  --secret-file=./cloud-credentials --image velero/velero:v1.4.0 --use-restic
  (Dont forget to change the region in url and region paramaeter )

• as to take backup with restic we need to annotate pod and statefulset or deployment and dont use pvc or pv name but use disk name declared in your manifest yaml file

• kubectl -n <namespace name> annotate pod/mysql-dev-0 [backup.velero.io/backup-volumes=<disk_name>](http://backup.velero.io/backup-volumes=<Disk name>)

• velero backup create **BACKUP_NAME** --include-namespaces <namespace> --wait

• velero restore create --from-backup **BACKUP_NAME**

IMPORTANT POINT

• restic is demon set to it will run on each and every node so suppose if you have tainted some node then remove taint . and taint again

• after restore kindly make sure restic init container got initiated ( this will help to restore data from s3 or object storage)

• if your statefulset contain taint and toleration make sure same is already created on new cluster or same cluster ( kindly don’t edit statefulset using kubect edit)

• cloud credential file containts

  aws_access_key_id=<linode_access_key></linode_access_key>

  aws_secret_access_key=<linode_secret_access_key></linode_secret_access_key>