All Products
Search

This Product

    Global Accelerator:Quotas and limits

    Document Center

    Global Accelerator:Quotas and limits

    Last Updated:Aug 27, 2026

    GA has three categories of quotas. Some can be raised on request; the rest are fixed product limits.

    • General quotas: cap the cloud resources a single Alibaba Cloud account can use.

    • API rate quotas: cap how frequently the account can call the GA OpenAPI.

    • Entitlement quotas: grant the right to use a feature after Alibaba Cloud authorizes it — a default of 0 means the feature is not enabled.

    The defaults listed on this page are for reference only. The value shown in the Quota Center is authoritative.

    Limits

    Configuration

    Standard

    Basic

    Supported acceleration areas

    Acceleration regions for Basic instances

    Acceleration IP types

    • Pay-as-you-go instances support both Elastic IP and Anycast EIP

    • Subscription instances support Elastic IP only

    Elastic IP only

    Acceleration IPs per instance

    • Pay-as-you-go:

      • EIP: 2 per acceleration region

      • Anycast EIP: 2 globally

    • Subscription: 1 EIP per acceleration region

    Controlled by gaplus_quota_basic_gaip_limit

    Minimum bandwidth per acceleration region

    • EIP: 2 Mbps

    • Anycast EIP: 200 Mbps

    EIP: 2 Mbps

    Supported backend types

    • Alibaba Cloud: ECS/ENI/ALB/NLB/CLB/OSS/Alibaba Cloud public IP/custom private IP/vSwitch

    • Non-Alibaba Cloud: custom public IP/custom domain

    ENI/CLB/NLB/ECS

    Routing types

    Intelligent routing, custom routing

    N/A — acceleration IP maps directly to endpoint

    Listener port range

    1–65499

    N/A

    Ports per listener

    • Intelligent routing

      • TCP/UDP: 30

      • HTTP/HTTPS: 1

    • Custom routing: 65499

    For TCP/UDP listeners on a pay-as-you-go instance, the total ports and port ranges across all listeners on one instance is also capped by gaplus_quota_listener_per_accelerator.

    N/A

    TCP/UDP fragmented packets

    Not supported

    Not supported

    GA runs on a distributed high-availability cluster. Per-connection throughput is bounded by the single-core performance of one data-plane node, while the bandwidth cap you configure is spread across multiple servers in the cluster. If your workload is dominated by high-throughput, long-lived connections, a single connection may not reach the instance's peak bandwidth. Split traffic into multiple concurrent connections to use the full bandwidth capacity.

    General quotas

    Adjustable quotas

    Quota name

    Description

    Default

    Maximum

    gaplus_quota_accelerator

    Maximum number of GA instances per user

    10

    40

    gaplus_quota_forwarding_rule_per_listener

    Forwarding rules per listener

    20

    150

    gaplus_quota_rule_condition_config_per_listener

    Total forwarding condition entries per listener

    40

    150

    gaplus_quota_vepg_per_listener

    Virtual endpoint groups per listener

    10

    150

    gaplus_quota_endpoint

    Maximum endpoints per endpoint group

    4

    30

    gaplus_quota_http_idle_timeout

    HTTP(S) connection idle timeout for pay-as-you-go Standard instances

    60s

    600s

    gaplus_quota_http_request_timeout

    HTTP(S) connection request timeout for pay-as-you-go Standard instances

    180s

    600s

    gaplus_quota_basic_gaip_limit

    Acceleration IPs per Basic instance

    20

    200

    Non-adjustable quotas

    Quota name

    Description

    Default

    gaplus_quota_listener_per_accelerator

    Maximum listeners per accelerator instance

    50

    gaplus_quota_epgs_per_listener

    Endpoint groups per listener

    10

    gaplus_quota_port_per_listener

    Maximum ports per listener

    30

    gaplus_quota_additional_certs_per_listener

    Maximum extended certificates per HTTPS listener

    25

    gaplus_quota_application_monitor

    Origin probes per GA instance

    5

    gaplus_quota_basic_gaip_idle_limit

    Idle acceleration IPs per Basic instance

    10

    gaplus_quota_basic_endpoint_limit

    Endpoints per Basic instance

    200

    gaplus_quota_listener_per_accelerator cannot be raised through the Quota Center. Contact your account manager to request an increase.

    API rate quotas

    API rate quotas cap how frequently a single Alibaba Cloud account can call the GA OpenAPI. The console displays the value as calls per 60 seconds — 300/60(s) means up to 300 calls every 60 seconds. To look up the limit for each operation, use either entry point:

    Most operations are capped globally; a few are unitized and counted per region. API rate quotas can also be raised from the Quota Center. An approved increase takes effect in about 3 minutes.

    Entitlement quotas

    Entitlement quotas require authorization from Alibaba Cloud before use. GA offers two:

    • Console downgrade ga_whitelist/instance_downgrade: allows downgrading the specification and bandwidth plan of a Standard instance.

    • Acceleration region allowlist ga_whitelist/region_whitelist: makes hidden acceleration regions visible and usable.

    View and raise quotas

    Manage quotas from the Alibaba Cloud Quota Center or from the Global Accelerator console quota management page. Both entry points expose the same set of quotas.

    • By default, only the Alibaba Cloud account (root account) can operate the Quota Center. RAM users must first be granted the AliyunQuotasFullAccess permission.

    • Each cloud product's technical support team reviews quota increase requests. A reasonable target value and a thorough justification improve the odds of approval. Results are delivered by SMS and email.

    • In resource directory scenarios, create a quota template to automatically submit increase requests for new member accounts as they join the resource directory. Templates require sign-in with the enterprise management account and an enabled resource directory.

    Console

    Sign in to the Quota management page, pick a quota category, and locate the target quota.

    • View a quota: review the name, description, and current usage of the quota.

    • Raise a quota: click Apply in the Actions column, fill in the desired value and justification, then submit. Track the review status under Application History — a status of Approved means the increase is live.

    API

    To cross-check actual GA resource usage, call GA resource-listing APIs such as ListAccelerators, ListListeners, and ListEndpointGroups.