Recover etcd
·
etcd
k8s
Table of Contents
| Control Panel |
|---|
| svr1 |
| svr2 |
| svr3 |
svr1>
svr1 #
Stop k8s>
Stop k8s #
systemctl stop kubelet
crictl rm -f $(crictl ps -q)
start the etcd in single node>
start the etcd in single node #
vim /etc/kubernetes/manifests/etcd.yaml
- --force-new-cluster=true
Add svr2 etcd to cluster>
Add svr2 etcd to cluster #
kubectl exec -it etcd-svr1.hhuge9.com -nkube-system -- sh -c 'ETCDCTL_API=3 etcdctl --endpoints 127.0.0.1:2379 --cert=/etc/kubernetes/pki/etcd/server.crt --key=/etc/kubernetes/pki/etcd/server.key --cacert=/etc/kubernetes/pki/etcd/ca.crt member add svr2.hhuge9.com --peer-urls="https://192.168.0.102:2380"'
Save the output>
Save the output #
export ETCD_NAME="svr2.hhuge9.com"
export ETCD_INITIAL_CLUSTER="svr1.hhuge9.com=http://192.168.0.101:2380,svr2.hhuge9.com=http://192.168.0.102:2380"
export ETCD_INITIAL_CLUSTER_STATE=existing
systemctl start kubelet
svr2>
svr2 #
Stop k8s and remove etcd data>
Stop k8s and remove etcd data #
systemctl stop kubelet
crictl rm -f $(crictl ps -q)
rm -rf /var/lib/etcd
Update etcd config (with the output in svr1)>
Update etcd config (with the output in svr1) #
vim /etc/kubernetes/manifests/etcd.yaml
- --initial-cluster=svr1.hhuge9.com=https://192.168.0.101:2380,svr2.hhuge9.com=https://192.168.0.102:2380
- --initial-cluster-state=existing
systemctl start kubelet