As workloads on virtualized platforms continue to grow, network traffic in virtualized environments increases accordingly. Many applications are built on virtualized platforms, and service interruptions can cause significant losses, making application availability an increasingly critical requirement. At the same time, the backend server processing speed has fallen far behind the growth in network bandwidth. The rapid increase in bandwidth has driven steadily rising resource consumption on backend servers, making servers a bottleneck in the overall network architecture. Single points of failure on servers also seriously affect application stability and reliability. Meeting application performance and availability requirements in virtualized environments has become an urgent challenge.
Against this background, load balancing technology emerged. It distributes workloads across servers and computing resources in a network to achieve higher performance and availability.
Load balancing is typically implemented in two forms: hardware load balancers and software load balancers. Traditional hardware load balancers are expensive, inflexible, difficult to reconfigure, and slow to scale, leaving them ill-suited for enterprise cloud migration and digital transformation. Compared with hardware load balancers, software load balancers offer greater flexibility, higher resource utilization, and lower acquisition and maintenance costs, making them the preferred choice for improving application performance in virtualized environments.
In earlier versions of ANS load balancing, LB instances could only be deployed as a single group of load balancer virtual machines (2–4 virtual machines) within a single ACOS cluster. While this met basic traffic forwarding needs, it had limitations in multi-cluster or cross-cluster scenarios: traffic might need to be forwarded across clusters, a single group of load balancer virtual machines could not provide cluster-level disaster recovery, and high availability requirements could not be fully met. Deploying multiple ANS services to address this would significantly increase operations management costs.
ANS provides multi-cluster and cross-cluster load balancing capabilities, allowing load balancing services to span multiple ACOS clusters for greater flexibility and reliability. ANS also supports load balancing in VPC network scenarios, enabling load balancing for east-west traffic within a VPC and north-south traffic from external networks into the VPC.
