Etcd restore. All the videos or learning platforms are showing different approaches to etcd restore. I want to understand why there are multiple procedures and clear confusion. can anyone please help?
The docs for doing an etcd restore are a bit complex, so videos, courses, or labs will often either simplify procedures for it. There’s also some variation due to etcdctl’s version. Key steps are:
- Create or find the restore archive created by etcdctl
- Restore a new etcd data directory with
etcdctl snapshot restore - If kube-apiserver is running, you’ll typically need to shut it down.
- Change the settings of etcd (in a kubeadm install, you’ll edit etcd.yaml) to point to the new data directory.
- Restart kube-apiserver.
Learn the procedures in the two etcd labs for starters, which will help you understand some of the issues.