We are seeing an issue with our AWS cluster. If we lose a node, the entire cluster becomes unresponsive. I set this - index.unassigned.node_left.delayed_timeout": "0" - but still doesn't help.
Is there something else that I need to set? If we lose a node, our cluster should not go unresponsive for several seconds (Or at least I wouldn't think it should. This was not the behavior in 1.5