Failed to heartbeat after deliberately killing cluster leader

3 consul cluster. If I kill the cluster leader, my Docker swarm using Consul for k/v store quits working. I thought that one of the other consul members would take over?

I've waited a few hours just to see how long it could take, but it kept trying to connect to the downed Docker container.

> ```
> 2016/02/04 07:53:46 [ERR] raft: Failed to heartbeat to 10.165.41.51:8300: dial tcp 10.165.41.51:8300: connection refused
> 2016/02/04 07:53:54 [ERR] raft: Failed to AppendEntries to 10.165.41.51:8300: dial tcp 10.165.41.51:8300: connection refused
> 2016/02/04 07:53:56 [ERR] raft: Failed to heartbeat to 10.165.41.51:8300: dial tcp 10.165.41.51:8300: connection refused
> 2016/02/04 07:54:04 [ERR] raft: Failed to AppendEntries to 10.165.41.51:8300: dial tcp 10.165.41.51:8300: connection refused
> 2016/02/04 07:54:07 [ERR] raft: Failed to heartbeat to 10.165.41.51:8300: dial tcp 10.165.41.51:8300: connection refused
> 2016/02/04 07:54:08 [INFO] serf: attempting reconnect to consul-doc8ucp1 10.165.41.51:8301
> 2016/02/04 07:54:15 [ERR] raft: Failed to AppendEntries to 10.165.41.51:8300: dial tcp 10.165.41.51:8300: connection refused
> 2016/02/04 07:54:17 [ERR] raft: Failed to heartbeat to 10.165.41.51:8300: dial tcp 10.165.41.51:8300: connection refused
> 2016/02/04 07:54:25 [ERR] raft: Failed to AppendEntries to 10.165.41.51:8300: dial tcp 10.165.41.51:8300: connection refused
> ```