Network load balancer. Overview
Network load balancers (NLBs) are designed to build fault-tolerant and scalable applications and services in Yandex Cloud.
Fault tolerance and scalability are achieved by distributing the traffic entering the load balancer among the cloud resources behind it.
Network load balancers operate at Level 4 of the OSI network model. Yet they employ technologies operating at Level 3 to accelerate data packet processing.
Using a network load balancer
A network load balancer can be used together with the following resources and services in Yandex Cloud:
- VMs: individual VM instances added to the balancer's target group directly, without using integrations with other services.
- Instance groups (Instance Groups). The instance group tracks the number of VMs in the group and changes the composition of the balancer's target group as needed.
- Clusters (Managed Service for Kubernetes). The master node of the Kubernetes cluster tracks the number of VMs in the cluster node groups and changes the composition of the balancer's target group as needed.
A load balancer’s incoming traffic is distributed in a certain way among VMs located downstream of the load balancer.
The structure of the network load balancer includes the following key components:
- Traffic Listener accepts traffic on the balancer.
- Target Group: Set of resources that receive traffic from the listener.
- Resource health checker for target groups checks the availability of resources in the target group.
You can create a network load balancer by specifying only its name. The other parameters are optional. An external balancer is created by default. Traffic listeners and target groups can be created later and linked to the previously created network load balancer.
Deployment tools
A network load balancer can be deployed in Yandex Cloud using the following tools: