Skip to content

Used watches for better resiliency upon failure #2

Description

@behemphi

Leverage watches in etcd to account for failures of the controller, docker, machine and network on restart.

If a machine in the leadership ring goes down, this should be updated within 1 second so that any machine subsequently coming up replaces it in the leadership ring.

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

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions