Load Balancing Across Microservices.

With the advent of cloud container technology, enterprises develop applications through microservices, breaking monolithic software into a suite of small services whose instances run independently in containers. User requests are served by a series of microservices forming a chain, and the chains often share microservices. Existing load balancing strategies either incur significant networking overhead or ignore the competition for shared microservices across chains. Furthermore, tpical load balancing solutions leverage a hybrid technique by combining IITTP with message queue to support microservice communications. bringing additional operational complexity. To address these challenges, we propose a chain -oriented load balancing algorithm (('OLBA) based solely on message queues, which balances load based on microservice requirements of chains to minimize response time. We model the load balancing problem as a non -cooperative game, and leverage Nash bargaining to coordinate microservice allocation across chains. Emplo mg convex optimization with rounding, we efficiently solve the problem that is proven NP -hard. Extensive trace -driven simulations demonstrate that COLBA reduces the overall average response time at least by 13c4 compared with existing load balancing strategies.