-
Notifications
You must be signed in to change notification settings - Fork 9.8k
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
Rejoining the cluster with a different client/raft url breaks etcd #1430
Comments
Adding |
Running
|
If I stop etcd0, things start working again. |
We don't support peer url update now. if you restart with
, it works. We would come up with warning message printed out for this case. |
@unihorn I've tried to follow your suggestion and it made etcd panic:
|
The panic happens because etcd0 uses the log that belongs to a different cluster from etcd1 and etcd2.
etcd1
etcd2:
The SYNC committed displays different time, so these logs belong to two different cluster. They cannot be run together. We have introduced We need to build command to automate the process to locate this kind of bugs. |
@kelseyhightower @unihorn
|
I guess the confusion comes in because I'm not sure how I created a new cluster. I just stopped and started the same node using the same data dir. |
Ok, this all makes sense now. Maybe along the way I ran bootstrap static again. |
further plan in #1440 . Thanks for reporting! |
Steps to reproduce:
Start a three node cluster:
etcd0
etcd1
etcd2
At this point everything works fine.
Stop etcd0 and add it back with a different client url
Logs from etcd0 etcd1 etcd2
etcd0
Initial cluster bootstrap
Restart with a different set of ips and ports for both raft and client urls
etcd1
etcd2
The text was updated successfully, but these errors were encountered: