Skip to content

Peer does not recover after repeated replication stream teardowns (to be confirmed) #453

Description

@JoshuaChi

Problem

After several consecutive stream teardowns caused by a backlog flush, the peer was
still at ~2% applied when observed, ~10s after the load ended. It is not known
whether it is permanently stuck or catching up slowly.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    raft-clusterCluster-wide operations and coordination issuesraft-membershipDynamic membership changes (node addition/removal from cluster)

    Type

    Projects

    No projects

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions