Description of problem:
By default, etcdctl is invoked with ETCDCTL_ENDPOINTS set to all etcd members. However, when invoking snapshot backup, it should be invoked with a single endpoint. This is strictly enforced in the new etcd versions of 3.4.x.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. Login to one of the master hosts
2. Attempt to take a backup by invoking cluster-backup.sh
In 4.5, where etcd is bumped to use 3.4.7, it fails with an error:
Error: snapshot must be requested to one selected node, not multiple [https://10.0.145.58:2379 https://10.0.134.79:2379 https://10.0.137.186:2379 https://10.0.5.140:2379]
snapshot db and kube resources are successfully saved to backup directory specified.
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.
For information on the advisory, and where to find the updated
files, follow the link below.
If the solution does not work for you, open a new bug report.