Etcd backup query


while I understand that etcd is running as a POD inside cluster. I was hoping any command to interact with etcd, we will be using kubectl exec followed by the command we need to run inside the POD.

But in the lab we are executing etcdctl command directly on control node. How that command is taking backup of DB from inside a POD and restroing same too. Also do we need to have etcdctl binary available mandatory inside master node?


Hello @06ashishrawat,

etcdctl is a client tool made to communicate with etcd and all you need to do is providing the endpoint of the etcd.

Yes, you need to have etcdctl available inside the node.

Hello @KodeKloud,

OK. Got it. So bydefault if we don’t provide any endpoint in our command like below :

then by default it will go to 2379 port of localhost, but in case if etcd cluster is running somewhere else, in any other node then we need to provide --endpoints=:2379.

Is it a correct understanding?