Node management upon Manager failure

moby / swarmkit

A toolkit for orchestrating distributed systems at any scale. It includes primitives for node discovery, raft-based consensus, task scheduling and more.

Apache License 2.0

3.36k stars 615 forks source link

Node management upon Manager failure #232

Closed aluzzardi closed 8 years ago

aluzzardi commented 8 years ago

When the manager starts, it should:

Set all nodes (from state) status to UNKNOWN.
Start the heartbeat timer (with a "grace period" to give time for agents to connect)
If the agents come back online, they will be READY
If agents fail to come back within grace period, then the heartbeat will set them to DOWN, triggering the drainer.

This should handle all use cases, including leader re-election.

aluzzardi commented 8 years ago

Duplicate of #277