-
Notifications
You must be signed in to change notification settings - Fork 275
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Cannot join nodes back after cluster-reset when not restoring from snapshot #2991
Comments
@brandond Do you know when we could plan to have this done? Or should it be put in the Backlog? |
This appears to be an issue with etcd, but we've been unable to reproduce it in a simple configuration such that upstream can diagnose and fix it. I would like QA to validate that it is still an issue with newer releases of etcd (such as the version shipped with 1.26), and if so we can put another spike of work into coming up with a simple reproducer for upstream. |
note I did diff the etcdctl get k8s output and they were different but they are a lot of noise in a comment. I have two stashed if you want to diff them yourself... Reproduced using VERSION=v1.23.6+rke2r2 $ get_etcd
ec2-user@ip-172-31-19-252:~$ kgn
$ get_etcd
ec2-user@ip-172-31-20-246:~$ kgn
$ get_figs rke2
Attempting to reproduce on v1.24.11+rke2r1 was more problematic first node
second node
first node
second node
Attempting to reproduce on v1.26.2+rke2r1 first node $ kgn
ec2-user@ip-172-31-20-246:~$ kgp -A
ec2-user@ip-172-31-20-246:~$ get_etcd
second node ec2-user@ip-172-31-19-252:~$ kgn
ec2-user@ip-172-31-19-252:~$ kgp -A
ec2-user@ip-172-31-19-252:~$ get_etcd
|
Yes, this seems to be fairly easy to reproduce using embedded etcd. I need to be able to provide the etcd maintainers with a set of steps to reproduce on vanilla etcd, without using Kubernetes, in order for them to diagnose it. |
As of RKE2 v1.27.4+rke2r1 I'm no longer seeing this behavior and am electing to close this issue as stale and no longer relevant. |
Forwardport for: #2857
The text was updated successfully, but these errors were encountered: