Load balancing distributes traffic among origin servers according to traffic steering policies to reduce latency and improve service availability.
Availability
Load balancing is available in the Enterprise plan.
Benefits
A load balancer can be referenced by the domain names accelerated by Edge Security Acceleration (ESA) and TCP/UDP proxy applications. Load balancing provides the following features:
Allows you to configure multiple levels of primary and secondary origin servers and use a secondary origin server when a primary origin server is abnormal.
Distributes incoming requests across multiple origin servers based on a defined ratio to facilitate load balancing among the origin servers.
Allows you to probe origin servers by using protocols such as ICMP Ping, TCP, HTTP/HTTPS, UDP, and ICMP and remove faulty origin servers.
Allows you to configure custom rules that route user requests to different origin servers based on the values of fields such as HTTP request headers, URLs, and URIs.
Supports session persistence to ensure that requests from the same user are consistently routed to the same origin server.
How it works
For example, your website, www.example.com, is added to ESA and uses Server Groups A, B, and C.
If Server Group A fails, the system switches to Group B. If Group B also fails, it switches to Group C. If all servers are down, requests default to Group C.
The load balancer uses ICMP Ping requests to check server health and blocks any servers that fail the probe.
Scenarios
Different business scenarios require various load balancing solutions.
Finance, gaming, e-commerce, and pan-entertainment
These scenarios require high availability.
Network issues, loads on origin servers, and configuration errors may cause the origin server services to have fluctuating availability or even become unavailable. This results in significant business losses.
The load balancing service can add and configure multiple groups of origin servers for mutual disaster recovery. The service actively probes the health status of origin servers, isolates faulty servers in advance, and forwards business traffic to healthy origin servers.
Global news portals and online education platforms
These scenarios require routing based on the values of HTTP fields.
With load balancing, user requests are routed to different origin servers by configuring custom rules based on HTTP request headers, URLs, and URIs.
Gaming and co-streaming scenarios
These scenarios require session persistence.
Load balancing ensures the consistency of user status information by routing requests from the same user to the same origin server.
Supported features based on DNS setup
The load balancer you created has a corresponding custom domain, and the support features of this domain will vary based on your DNS setup.
NS setup
If you connect your website to ESA in NS setup, ESA manages DNS resolution for your domain. To prevent service interruptions, you need to add all DNS records of the domain to ESA and update the nameservers to ESA assigned ones at your registrar. The DNS records can be imported at a time.
If you select NS for DNS setup, the load balancer domain can be used for DNS resolution. The domain can be designated as the CDN domain or the origin for TCP/UDP proxy applications.
CNAME setup
If you connect your website to ESA in CNAME setup, you need to add a DNS record to map the domain that you want to proxy to the ESA assigned CNAME. This way, user requests destined for your domain can be forwarded to ESA points of presence (POPs). This enables content delivery acceleration, edge computing, and enhanced protection.
If you select CNAME for DNS setup, the load balancer domain cannot be directly used for DNS resolution. However, you still have the option to create a load balancer domain. The domain can function as a regular domain and be designated as the origin for TCP/UDP proxy applications.