Restore cluster1_backup_to_restore.db deprecated etcdctl solution not working!

Kubernetes
,
#1

Hi,

I am currently trying to complete q3 of a kubernetes mock exam, but I am unable to. I really do need to practice this topic as this is one of the questions know I had issues with in the real exam. As you can see I have already ssh’d into the node and tried to restore based on the kubernetes.io documentation. I know this isn’t meant to be that hard. I do know that in the solution you provide you request we install the etcd utility using the following commands which I have done. However this is not practical in the exam as we would need to go to github to get the correct html address. None of this is in the documentation on kubernetes website. So I do not think it is practical to request us to memorise something that a) isn’t in the exam and b) does not make the snapshot restore command work.

install the etcd utility :

cluster1-controlplane ~ ➜ cd /tmp
cluster1-controlplane ~ ➜ export RELEASE=$(curl -s https://api.github.com/repos/etcd-io/etcd/releases/latest | grep tag_name | cut -d '"' -f 4)
cluster1-controlplane ~ ➜ wget https://github.com/etcd-io/etcd/releases/download/${RELEASE}/etcd-${RELEASE}-linux-amd64.tar.gz
cluster1-controlplane ~ ➜ tar xvf etcd-${RELEASE}-linux-amd64.tar.gz ; cd etcd-${RELEASE}-linux-amd64
cluster1-controlplane ~ ➜ mv etcd etcdctl  /usr/local/bin/
cluster1-controlplane ~ ➜ etcdctl snapshot restore --data-dir /root/default.etcd /opt/cluster1_backup_to_restore.db

cluster1-controlplane ~ ➜ ETCDCTL_API=3 etcdctl --data-dir /root/default.etcd snapshot restore /opt/cluster1_backup_to_restore.db
Deprecated: Use etcdutl snapshot restore instead.

2024-04-15T01:41:14Z info snapshot/v3_snapshot.go:260 restoring snapshot {“path”: “/opt/cluster1_backup_to_restore.db”, “wal-dir”: “/root/default.etcd/member/wal”, “data-dir”: “/root/default.etcd”, “snap-dir”: “/root/default.etcd/member/snap”}
2024-04-15T01:41:14Z info membership/store.go:141 Trimming membership information from the backend…
2024-04-15T01:41:14Z info membership/cluster.go:421 added member {“cluster-id”: “cdf818194e3a8c32”, “local-member-id”: “0”, “added-peer-id”: “8e9e05c52164694d”, “added-peer-peer-urls”: [“http://localhost:2380”]}
2024-04-15T01:41:14Z info snapshot/v3_snapshot.go:287 restored snapshot {“path”: “/opt/cluster1_backup_to_restore.db”, “wal-dir”: “/root/default.etcd/member/wal”, “data-dir”: “/root/default.etcd”, “snap-dir”: “/root/default.etcd/member/snap”}

If I am wrong, please tell me, but your solution which I copied after failing this question a dozen times does not work. Please either fix the lab or fix the solution.

#2

First of all, don’t worry about the deprecation message. etcdutl is pretty new, and it’s fine to use etcdctl to do a restore. Not a problem!

The rest of the messages are information, and are not errors either. I’ll need to check to see if I can find the question (a link to a lab is very welcome – there are 3 mock exams to check).

#3

OK, found your question; it’s in the UME-CKA exams. A couple of points:

  1. You don’t have to install etcdctl at all, it turns out: it’s already installed for you. In the real exam, this would also be the case.
  2. Doing a restore is also in the K8s docs, here.
  3. The question asks you to restore to the directory, which means that you don’t have to actually rewire /etc/kubernetes/manifests/etcd.yaml to use the new directory.