You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Describe the bug:
When an etcd cluster is scaled up, there is a rare case where the third pod cannot join the cluster. Hence, we have only 2 members working is expected ways. What this results in is a cluster is size 2 with a learner that cannot join.
Although this means that the quorum is maintained, it also means that we lose a bit of resilience as one we are now just pod away from quorum loss.
The third pod complains of a cluster ID mismatch error and hence refuses to join the cluster. A possible cause of this is incorrect configuration passed to the etcd container before it is ready to join the cluster. The root cause however, is not very clear.
Expected behavior:
Cluster scale up to always succeed
How To Reproduce (as minimally and precisely as possible):
It is not yet clear how to consistently reproduce this. It is seen quite sporadically.
The text was updated successfully, but these errors were encountered:
Describe the bug:
When an etcd cluster is scaled up, there is a rare case where the third pod cannot join the cluster. Hence, we have only 2 members working is expected ways. What this results in is a cluster is size 2 with a learner that cannot join.
Although this means that the quorum is maintained, it also means that we lose a bit of resilience as one we are now just pod away from quorum loss.
The third pod complains of a
cluster ID mismatch
error and hence refuses to join the cluster. A possible cause of this is incorrect configuration passed to the etcd container before it is ready to join the cluster. The root cause however, is not very clear.Expected behavior:
Cluster scale up to always succeed
How To Reproduce (as minimally and precisely as possible):
It is not yet clear how to consistently reproduce this. It is seen quite sporadically.
The text was updated successfully, but these errors were encountered: