Product Description Fault Tolerant Load Balancer
A fail-safe load balancer is a reverse proxy that distributes Internet traffic between different Selectel services located in different regions and availability zones:
Balancer can be used to build fault-tolerant services reserved in multiple regions and hybrid infrastructures based on dedicated servers, cloud servers and VMware-based public cloud. It also provides DDoS protection for all inbound traffic at the L3 and L4 levels.
Principle of operation
To connect the balancer, you combine the infrastructure into a private network via a Selectel global router You connect the balancer to the same network. We provide you with an external IP address of the balancer to receive traffic, the address is protected from DDoS attacks and announced to the Internet via BGP Anycast protocol. Traffic coming to the balancer is distributed in the infrastructure behind it according to the specified rules — balancing algorithm, protocol, IP addresses and ports of target servers.
The balancer is reserved in two independent regions, Moscow and St. Petersburg, with a pair of High-Avaliability balancers operating in both regions. If a router or balancer in one region becomes unavailable, traffic is automatically switched to the other region.
Selectel is responsible for configuring the balancer, providing fault tolerance and monitoring.
Balancing levels
According to the OSI model, balancing is possible at layers:
- L4 over TCP protocol (transport layer) — the balancer redirects requests to servers according to specified algorithms;
- L7 over HTTP (application layer) — the balancer analyzes the contents of the request and chooses which server to send the request to.
Balancing algorithms
Different balancing algorithms are available at both balancing levels (L4, L7):
- Round Robin is a round robin algorithm in which requests are passed to each server in turn;
- Weighted Round Robin is a weighted round robin algorithm in which each server is assigned a weighting factor corresponding to its performance and capacity. Servers with a higher weighting factor receive more requests;
- Source IP hash is an algorithm that, depending on the HTTP header or IP address, selects a preferred server to receive the request;
- Least Connections — an algorithm that sends the request to the least loaded server.
Comparison of balancing levels
Limitations
The product has no restrictions on the number of target servers and the amount of traffic to be transmitted.