Integrate GA to Accelerate Applications

Key features

• Application acceleration: Improve the end user experience for global users through network acceleration.

• Simplified configuration: Set up an accelerator in the ALB console with one click to reduce configuration complexity.

Scenarios

This solution works best when your application must serve end users across multiple global regions:

• Game platform acceleration: Accelerate game platform services such as login or store systems to reduce slow logins and other negative user feedback.

• Enterprise application acceleration: Accelerate cross-border enterprise applications to improve global employee access and boost collaboration efficiency.

• Internet application acceleration: Accelerate internet-facing applications to enhance user experience and engagement, helping you build popular products.

Step 1: Enable application acceleration for your ALB instance

1. Log on to the Application Load Balancer (ALB) console.

2. In the top menu bar, select the region where your instance is deployed.

3. On the Instances page, find your target instance and click its instance ID.

4. On the Integrated Services tab, click Create GA.

   1. Activate GA: If you have never used GA, read the Global Accelerator Service Agreement and check the box to activate the service.

   2. Acceleration Area: Acceleration regions are the geographic areas where you want to accelerate access. Choose the region where your clients are located or the nearest region. Each acceleration region includes one or more Alibaba Cloud regions.

      Note

      • If your acceleration region includes the Chinese mainland or your backend servers are deployed in the Chinese mainland, you must complete ICP filing for your custom domain name.

      • If your acceleration region and origin server region involve cross-border access, read the Compliance Commitments Regarding Cross-border Data Transfers and check the box to agree. Premium bandwidth for cross-border acceleration is enabled by default.

5. After completing the configuration, click OK.

   Important

   The first time you enable acceleration, all listener configurations from your ALB instance are synchronized to GA. Subsequent changes to ALB listeners are not automatically synchronized. You must manually update them in the GA console.

Step 2: Configure domain name resolution for GA

After enabling GA acceleration for your ALB instance, the Integrated Services page displays information about the created GA instance, including its DNS name.

Copy the GA instance’s DNS name. Then add a CNAME record that points your custom domain to this DNS name.

1. On the Domain Name Resolution page, find your target domain and click Settings in the Actions column.

   Note

   For domains not registered with Alibaba Cloud, you must first add the domain to the Cloud DNS console before you can configure DNS records.

2. On the DNS settings page, click Add Record, configure the CNAME record, and then click OK.

   Setting	Description
   Record Type	Select CNAME from the drop-down list.
   Hostname	The prefix of your domain name. In this example, enter @. Note When creating a record for a root domain, set the host record to @.
   Query Source	Select Default.
   TTL	TTL (Time To Live) specifies how long DNS records are cached on DNS servers. Use the default value in this example.
   Record Value	Enter the CNAME address corresponding to the domain name, which is the DNS name of the GA instance that you copied.

Step 3: Verify acceleration performance

This test uses a public ALB instance with backend servers in US (Silicon Valley) and clients in the Hong Kong (China) acceleration region.

1. Test network latency after acceleration:

   1. Access http://<your-GA-domain> in a browser. The backend service loads normally. Refresh the browser multiple times to switch between ECS01 and ECS02.

   2. Run curl -o /dev/null -s -w "time_connect: %{time_connect}\ntime_starttransfer: %{time_starttransfer}\ntime_total: %{time_total}\n" "http[s]://<your-GA-domain>" to check packet latency after acceleration.

      After acceleration, the output looks like this:

      [root@xxx xxx ~]#  curl -o /dev/null -s -w "time_connect: %{time_connect}\ntime_starttransfer: %{time_starttransfer}\ntime_total: %{time_total}\n" "http://lxxx.cn"
      time_connect: 0.006
      time_starttransfer: 0.008
      time_total: 0.008

2. Test network latency before acceleration:

   1. Access http://<your-ALB-domain> in a browser. The backend service loads normally. Refresh the browser multiple times to switch between ECS01 and ECS02.

   2. Run curl -o /dev/null -s -w "time_connect: %{time_connect}\ntime_starttransfer: %{time_starttransfer}\ntime_total: %{time_total}\n" "http[s]://<your-ALB-domain>" to check packet latency before acceleration.

      Before acceleration, the output looks like this:

      [root@xxx ~]# curl -o /dev/null -s -w "time_connect: %{time_connect}\ntime_starttransfer: %{time_starttransfer}\ntime_total: %{time_total}\n" "http://bxxx.cn"
      time_connect: 0.162
      time_starttransfer: 0.320
      time_total: 0.321

3. Compare acceleration results:

   Parameter descriptions:

   • time_connect: Connection time—the time from request start until the TCP connection is established (in seconds).

   • time_starttransfer: Time to first byte—the time from request start until the first byte is received from the backend server (in seconds).

   • time_total: Total time—the time from request start until the full response is received (in seconds).

   Parameter	After acceleration (seconds)	Before acceleration (seconds)	Acceleration Data Reference (in Seconds)	Acceleration Reference (Percentage)
   time_connect	0.006	0.162	Improvement: 0.156	96.3% faster
   time_starttransfer	0.008	0.320	Improvement: 0.312	97.5% faster
   time_total	0.008	0.321	Improvement: 0.313	97.5% faster

   Note

   The example data shown here is for reference only. Actual acceleration results depend on your specific business and testing environment.

View acceleration status

On the Integrated Services page for your instance, the Global Accelerator card displays the GA instance ID, DNS name, instance status, and other details.

Disable application acceleration

To disable GA acceleration for your ALB instance, delete the GA instance in the GA console or remove the associated ALB listener and endpoint group.

What type of GA instance is created?

A pay-as-you-go standard GA instance.

How does enabling application acceleration affect billing?

Enabling application acceleration adds GA-related charges, including instance fees, CU fees, and traffic fees. For details, see Billing for pay-as-you-go Global Accelerator instances.

Why isn’t my updated ALB listener accelerated?

When you first enable GA integration, all ALB listener configurations are synchronized to GA. Subsequent changes to ALB listeners are not automatically synchronized. You must manually update them in the GA console.

Why did my ALB access control policy stop working after acceleration?

After GA acceleration, end users access your service through the GA DNS name. Therefore, the ALB access control policy no longer applies.

To enforce similar IP-based access control, configure an access control policy in GA. For more information, see GA access control.