Sometimes an ec2 instance will fail to launch, switching from pending to 'terminated' , currently this locks the worker daemon in the loop as it is waiting for instances to enter the "running" state. We should detect this state change in the launch loop, delete the node, and replace with another instance of the same type.
Sometimes an ec2 instance will fail to launch, switching from pending to 'terminated' , currently this locks the worker daemon in the loop as it is waiting for instances to enter the "running" state. We should detect this state change in the launch loop, delete the node, and replace with another instance of the same type.