GA listeners - protocols, routing, and ports - Global Accelerator

Listener protocols

You can create up to 50 listeners per GA instance. The following listener protocols are supported: TCP, UDP, HTTP, and HTTPS.

Protocol	Description	Use scenario
TCP	Connection-oriented with high reliability. Requires connections before data transfer. Session persistence based on source IP addresses. Source IP addresses visible at the network layer. Slower transmission rate.	Scenarios requiring high reliability and data accuracy over speed, such as file transfers, email, and remote logons. General-purpose web applications.
UDP	Connectionless protocol. No three-way handshakes, error recovery, or data retransmission. High transmission rate.	Scenarios where real-time transmission outweighs reliability, such as video conferencing and real-time quotes.
HTTP	Connection-oriented with high reliability. Requires connections before data transfer. High transmission rate. Unencrypted data transmission.	Accelerating HTTP websites. Accelerating HTTP websites with specific domain names or paths.
HTTPS	Connection-oriented with high reliability. Requires connections before data transfer. Supports SSL certificate binding to servers for data security. Note For more information about SSL certificates, see What is Certificate Management Service? Encrypted data transmission. Configurable maximum HTTP version for HTTP/3 connections improves transmission stability, security, and efficiency.	Accelerating HTTPS websites with network security and efficiency for client access. Accelerating HTTPS websites with specific domain names or paths.

Routing listeners

GA supports intelligent routing and custom routing listeners. Select a routing type based on your business requirements.

Note

Custom routing listeners are in invitational preview. To use custom routing listeners, submit an application to your account manager. After your application is approved, you can use custom routing listeners.

Intelligent routing listeners

Intelligent routing is the default routing type of original listeners. After you configure an intelligent routing listener for a GA instance, GA selects a nearby healthy endpoint group based on network latency and forwards requests to the optimal endpoint. Latency varies by geographical location and network hops.

Ideal for fine-grained global traffic control. Configure traffic weights for endpoint groups and endpoints as needed. Use cases: blue-green deployment, A/B testing, multi-region deployment, service discontinuation or update in specific regions, and cross-region disaster recovery. For information about the use scenarios and examples, see Distribute traffic across endpoint groups in different scenarios and Examples on how to configure the traffic distribution feature for multiple endpoint groups.

Intelligent routing listeners support TCP, UDP, HTTP, and HTTPS.

Custom routing listeners

After you configure a custom routing listener for a GA instance, GA generates a port mapping table from the listener port range, endpoint group port ranges, and endpoint IP addresses (vSwitches). The listener forwards requests to specific vSwitch IP addresses and ports based on this mapping table.

Custom routing listeners forward traffic to specified backend servers. Use cases: multiplayer online games, video conferencing, and virtual classrooms. For information about how custom routing listeners work and how to use custom routing listeners, see How custom routing listeners work.

Custom routing listeners support only TCP and UDP. You can specify protocols when you configure endpoint groups. Specify TCP, UDP, or both for the destination port range of each endpoint group.

Note

All listeners on the same GA instance must use the same routing type, which cannot be changed after configuration. For information about the differences between the two types of routing listeners, see the "Comparison between intelligent routing listeners and custom routing listeners" section of the How custom routing listeners work topic.

Listener ports

Listener ports receive and forward requests to endpoints.

Intelligent routing listener ports

Note

The following limits apply to intelligent routing listeners on the same GA instance:

The ports of TCP, HTTP, and HTTPS listeners cannot be the same.
The port of a UDP listener cannot be the same as the port of an HTTPS listener for which HTTP/3 is enabled.

Port quotas by protocol:

Listener protocol	Listener port range	Port quota
TCP	1~65499	30. Separate multiple listener ports with commas (,). Example: `80,90,8080`. You can use a hyphen (-) to specify a port range. For example, you can enter `80-83` to specify ports 80, 81, 82, and 83.
UDP	1~65499	30. Separate multiple listener ports with commas (,). Example: `80,90,8080`. You can use a hyphen (-) to specify a port range. For example, you can enter `80-83` to specify ports 80, 81, 82, and 83.
HTTP	1~65499	1.
HTTPS	1~65499	1.

The following limits apply to TCP or UDP listeners based on the billing method of the GA instance:

Pay-as-you-go

Specify multiple port ranges with consecutive ports for a single listener. Valid ports: 1 to 65499. Port ranges must not overlap.
The number of ports supported by a single listener is determined by the value of the gaplus_quota_port_per_listener quota. Each port or port range consumes one quota.

The number of ports, including ports and port ranges, supported by all listeners on a single GA instance is determined by the value of the gaplus_quota_listener_per_accelerator quota, which caps the total across all listeners on the GA instance. This total cannot exceed the gaplus_quota_listener_per_accelerator quota.

Note
The default value of the gaplus_quota_listener_per_accelerator quota is 50. To increase it, contact your account manager.

For example, if you specify 1,3-350,400-800 for a TCP listener of your GA instance, three quotas of gaplus_quota_port_per_listener are consumed.

In this case, you can add at most 27 ports or port ranges to the TCP listener. If the number of ports or port ranges of the TCP listener reaches 30, you can add at most 20 ports or port ranges to other TCP or UDP listeners on the GA instance.

Warning

Changing the port or listener protocol may interrupt data transmission. Modify listener configurations during off-peak hours or after switching traffic to another GA instance.

Subscription

The number of ports supported by a single listener is determined by the value of the gaplus_quota_port_per_listener quota. Each port consumes one quota.

For example, if you specify 1,3-5 for a TCP listener of your GA instance, four quotas of gaplus_quota_port_per_listener are consumed. In this case, you can add at most 26 ports to the TCP listener.
By default, you can specify port ranges with more than 300 consecutive ports. A listener with more than 300 consecutive ports is an advanced listener with the following limits:
- If you cannot configure a port range that contains more than 300 consecutive ports for a TCP or UDP listener, your GA instance may be using an earlier version. Contact your account manager to upgrade it.
- An advanced listener requires more than 300 ports, up to 65,499.
- Each subscription GA instance supports only one advanced listener.
- An advanced port range consumes one quota of gaplus_quota_port_per_listener.
For example, if you specify TCP ports 1–400, TCP port 443, UDP ports 200–210, and UDP ports 230–240, then TCP ports 1–400 is the advanced port range.

Warning

Changing the port or listener protocol may interrupt data transmission. Modify listener configurations during off-peak hours or after switching traffic to another GA instance.

Custom routing listener ports

The following limits apply to custom routing listener ports:

Custom routing listeners use ports 1–65499. Ports 25, 250, 4789, and 4790 are system-reserved and excluded from port mapping tables. Pay-as-you-go GA instances also exclude port 6081.
The listener port range must accommodate the endpoint group ports and vSwitch IP addresses. The following formula must be satisfied: Number of listening ports (minus the number of system-reserved ports) ≥ Number of all ports in the endpoint group × Number of IP addresses of all vSwitches in the endpoint. Specify a large port range for the listener.

For example, if you set the destination port range of endpoint groups to 81-85 and all vSwitches that are used as endpoints provide 16 IP addresses, the listener port range must contain at least 80 ports. You can set the port range of the listener to 101-180. If you set the port range of the listener to 101-179, the listener cannot be created.
After configuration, you cannot remove ports that exist in the port mapping table when you modify the port range.

For example, the original listening port range is 100-10000, and a mapping is established between port 199 and the destination port 80 of the endpoint in the port mapping table. You can expand the listening port range to 20-10000. You cannot narrow the listening port range to 200-10000.
Listener ports on the same GA instance must not overlap.