Terms

Term

Description

Global Traffic Manager

A DNS-based service that routes users to the nearest application endpoint, distributes high-concurrency loads evenly, and switches traffic based on health checks for disaster recovery.

CNAME access domain

When you purchase Global Traffic Manager, the system automatically assigns a domain record to each instance. You can use a CNAME record to point your domain to the Global Traffic Manager instance.

Geographic location-based access policy

Uses intelligent DNS resolution to assign different IPAM pools to users from different networks or regions, routing global users to the nearest service endpoint.

Latency-based access policy

Measures network latency from each visitor to application address regions and routes the visitor to the server cluster with the lowest latency.

Primary address pool collection

A collection of same-type address pools grouped for different business scenarios within an access policy. User traffic is routed through the primary address pool collection by default.

Secondary address pool collection

When an application service fails, traffic is automatically switched from the primary address pool collection to the secondary address pool collection to minimize service downtime.

IPAM pool

A group of IP addresses that provide the same service, used to distribute user traffic and reduce the load on individual IP addresses. An address pool can contain IP addresses or domain names.

Return all addresses

A load balancing mode in which a domain resolves to all available IP addresses, with each address handling an equal share of traffic.

Return addresses by weight

A load balancing mode in which a domain resolves to multiple addresses, with each address assigned a different weight to handle a proportional share of traffic.

Health check

Monitors the availability of application services on IP addresses in real time by using ping, TCP, or HTTP probes from a group of monitoring nodes, and determines service health based on configured alert rules.

Detection task

When health checks are enabled in a public network scenario, each target IP address requires one monitoring node, which counts as one detection task.

Detection task quantity calculation formula:

Used detection tasks = Number of addresses in address pool A × Number of monitoring nodes in address pool A health check configuration + Number of addresses in address pool B × Number of monitoring nodes in address pool B health check configuration

Traffic rerouting

Distributes requests from users to different destination servers by analyzing network traffic characteristics in real time.

Maximum number of returned addresses

When the access policy type is

Latency-based access policy and the application service has multiple IP addresses, the system returns one or more addresses with the lowest latency. Valid values: 1 to 8.

Latency resolution scheduling optimization

When the maximum number of returned addresses is greater than 1, you can enable this mode to let the system intelligently select the optimal addresses to return.

SNI

Server Name Indication (SNI), defined in RFC 4366, is an extension to SSL/TLS available in SSLv3/TLSv1. It allows clients to include the requested hostname during the SSL handshake (in the ClientHello phase), enabling the server to select the correct domain and return the appropriate certificate.