panic when deathnode iterates over already terminated instance

[frimik]

instance.go:

	return &InstanceMonitor{
		autoscalingGroupID:  autoscalingGroupID,
		ipAddress:           *response.PrivateIpAddress,
		instanceID:          instanceID,
		lifecycleState:      lifecycleState,
		isProtected:         isProtected,
		ctx:                 ctx,
		tagRemovalTimestamp: tagRemovalTimestamp,
	}, nil

in this case, the response.PrivateIpAddress will be nil <*string> and you will panic.

Handle it and finish the lifecycle on this instance if you find that is needed. Otherwise you will continually panic until the instance lifecycle timeout expires (3600 seconds)