Jan B janeczku

Here are the steps to enable and test the Kubernetes Volume Snaphot subsystem on Rancher RKE or RKE2.

Deploying the volume snapshot controller and CRDs is required for managing snapshots for storage providers such as Longhorn, Rook/Ceph via the native K8s API as well as for integrating Rancher-managed clusters with backup solutions like Kasten K10.

1. Clone the repository

export SNAPSHOT_CTRL_VER=v5.0.1
git clone -b $SNAPSHOT_CTRL_VER https://github.com/kubernetes-csi/external-snapshotter.git
cd external-snapshotter

worker_processes 4;
worker_processes 4;
worker_rlimit_nofile 40000;

events {
    worker_connections 8192;
}

http {

/home/USER/.docker/config.json

{
    "auths": {
        "https://index.docker.io/v1/": {
            "auth": "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx="
        },
        "quay.io": {
 "xxxx": "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"

To configure a new default registry that will be used to pull images that do not have fully qualified image paths, modify the containerd configuration (/etc/containerd/config.toml) as follows:


version = 2
[plugins]
  [plugins."io.containerd.grpc.v1.cri"]
    [plugins."io.containerd.grpc.v1.cri".registry.mirrors]
      [plugins."io.containerd.grpc.v1.cri".registry.mirrors."docker.io"]
 endpoint = ["https://private.registry.com:443"]

	apiVersion: apps/v1
	kind: Deployment
	metadata:
	name: test
	spec:
	replicas: 1
	selector:
	matchLabels:
	app: test
	template:

	apiVersion: argoproj.io/v1alpha1
	kind: EventSource
	metadata:
	name: resource
	spec:
	template:
	serviceAccountName: your-service-account
	resource:
	capi-cluster:
	namespace: fleet-default

	# Adding the following config stanza to all the Harvester create\|join configs will create
	# a custom cloud-config `/oem/95_user.yaml` during the (early) "initramfs" cloud-init stage
	# of the initial Harvester boot.
	# This cloud-config will be executed on each system (re-)boot during the (late) "boot" cloud-init
	# stage and may contain any cloud-init directives supported by the cOS Toolkit:
	# See https://rancher.github.io/elemental-toolkit/docs/reference/cloud_init/
	# Additionally, any files added to the /oem folder in day-2 are persistent and won’t be overwritten
	# during Harvester upgrades.

	write_files:

	# Filename: /oem/95_user.yaml
	# Ref: https://rancher.github.io/elemental-toolkit/docs/customizing/stages/
	name: "User Config"
	stages:
	initramfs:
	- name: "Drop unit file"
	files:
	- path: /etc/systemd/system/update-ca.service
	content: \|
	[Unit]

	# Replace private.registry.com with the host name of your private container registry
	# Save as /etc/rancher/rke2/registries.yaml on each Harvester server
	mirrors:
	docker.io:
	endpoint:
	- "https://private.registry.com"
	private.registry.com:
	endpoint:
	- "https://private.registry.com"
	configs:

	#!/usr/bin/env bash
	set -xe

	RANCHER_HOST="REDACTED.cloud"
	BEARER_TOKEN="token-hcjwf:REDACTED"

	if [[ $# -eq 0 ]] ; then
	echo "please specify cluster name"
	exit 1
	fi