Based on my understanding of load-balancing, requests are forwarded to multiple instances of a service so that if one service is down, it will be redirected to one that is available.
I have created a microservice that implements load-balancing like this using Ribbon and Hystrix to handle fallback. However, this only works if the load-balancing service is working properly.
What if the load-balancing service goes down? Is there a solution to keep the microservice running so that the instances of the service can still be reached?