Replies: 3 comments 6 replies
-
Purge the member data again and start it with |
Beta Was this translation helpful? Give feedback.
-
I purged the member by removing the
Nodes 1 & 2 are not logging anything so it looks like node 3 doesn't try to connect to them. I checked that the cluster infromation is available in the discovery cluster with
|
Beta Was this translation helpful? Give feedback.
-
After a reset of the etcd cluster, I could not reproduce this exact issue. However, this is still the case currently: once I remove a member from the cluster with
This isn't a huge problem though, I just have to make sure not to |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
I'm working on setting up an etcd cluster using cluster discovery (with an existing etcd cluster).
I tested the following scenario: the server
etcd-testing3
had to be reinstalled due to failure.After server reset, the node
etcd-testing3
was started with the same configuration used before:When I list the cluster members, node 3 is still listed as member:
However, it looks like the node 3 has started a cluster and a life on its own, as the node 1 & 2 says they can't communicate with node 3 and they
failed to find remote peer in cluster
.The node 3 has logged the following:
But the node 1 & 2 are spamming the following error:
I also tried to remove the node 3 from the cluster with
etcdctl
.However, it couldn't join the cluster as the member
6fa6da62f2b4add9
was "removed from the cluster", so node 1 & 2 refused it.Resetting the node 3 again did not help as it did not change member ID (it seems the ID is generated from the member name and it's URLs, so it won't change unless I change the member name).
Is there a way to make my server member of the cluster again?
Beta Was this translation helpful? Give feedback.
All reactions