The following considerations should be taken into account while implementing this pattern:

• Route optimization: Optimizing a route is a complex process. Considering that load handled by an instance of a Microservice can change quickly, the records about previous response time may not lead to the most optimal route. Constant performance monitoring need to happen to analyze the load on virtual machines hosting Microservices.
• Health checks: Microservices can be hosted on commodity hardware which are susceptible to failure. Load balancer should be cognizant about the health of the virtual machines hosting Microservices.
• Point of failure: Load balancer can act like a single point of failure. It is critical to make load balancing service and its associated state highly available. Geo-replicating instances of the load balancer is advisable to achieve high availability.