We received a notification from AWS at 6:27 PM EST on 1/7/15 about an unhealthy search node. search-prod1 became nonresponsive; attempts to connect to it via SSH, etc. failed.
I have tried to reboot the instance from the EC2 console panel, with no obvious success.
I have tried to stop the instance; it got stuck in the stopping state.
I tried to stop it again, which should have run a "force stop" state, but it remains stuck in the "stopping" state.
The search cluster otherwise appears healthy, and all shards have a replica.
@search-prod1@ finally responded to the stop command, and so I asked it to start up again. It's currently in the "pending" state.
@search-prod1@ is back online and has rejoined the cluster.
Note: EC2 console still says that search-prod1 is @pending@.
OK, back to @running@. Will monitor and close this later.