Function Compute：Asynchronous tasks

Limits

• Scenarios

  Although the asynchronous task feature provides more capabilities, more system overheads are incurred. We recommend that you disable the asynchronous task feature in the following scenarios:

  • Your business applications are sensitive to the latencies of request processing links and require the average latency to be less than 100 milliseconds.

  • You need to initiate at least thousands of asynchronous invocations per second.

• Supported regions

  The asynchronous task feature is supported in the following regions: China (Hangzhou), China (Shanghai), China (Qingdao), China (Beijing), China (Zhangjiakou), China (Shenzhen), China (Chengdu), China (Hong Kong), Singapore, UK (London), US (Silicon Valley), US (Virginia), Germany (Frankfurt), Malaysia (Kuala Lumpur), Indonesia (Jakarta), Thailand (Bangkok), Japan (Tokyo), and South Korea (Seoul).

• Validity period

  You can query the task status information only within the last seven days.

Comparison between asynchronous tasks and Kubernetes jobs

Both the asynchronous task feature of Function Compute and Kubernetes jobs can be used to execute asynchronous tasks. Function Compute provides a ready-to-use asynchronous task execution platform that is more efficient and easier to control and observe based on the asynchronous task feature. The following table compares the two methods:

Task management

States of an asynchronous task

For each task invocation, Function Compute records the information about the state changes of tasks and allows you to query the states of a task in real time. You can view the states of a task by using the Function Compute SDK or the Function Compute console. A task can be in one of the following states:

You can call the GetAsyncTask operation to query the details of an asynchronous task. You can also call the ListAsyncTasks operation to filter tasks in a specified state.

Manage tasks

{
  "asyncTask": true
}

Trigger tasks by using events

You can create triggers that support asynchronous invocations to trigger asynchronous tasks. The following items list the triggers that support asynchronous invocations.

• HTTP triggers

• Time triggers

• OSS triggers

• Simple Message Queue (formerly MNS) topic triggers

• EventBridge-based triggers:

  • Simple Message Queue (formerly MNS) queue triggers

  • ApsaraMQ for RocketMQ triggers

  • ApsaraMQ for RabbitMQ triggers

  • ApsaraMQ for Kafka triggers

  • ApsaraMQ for MQTT triggers

  • Data Transmission Service (DTS) triggers

  Important

  Make sure that HTTP triggers and EventBridge-based triggers are triggered in the asynchronous mode.

Task deduplication

To avoid repeated task execution, Function Compute allows you to set a globally unique ID for each submitted task. If an asynchronous operation, such as the asynchronous submission operation, times out, you can retry by submitting a task with the same ID. For more information, see Retry policy.

Function Compute allows you to configure TaskID, which is a globally unique task ID. We recommend that you specify a task ID each time you submit a task. You can use the task ID to retry the task in case of an error such as a timeout error. Function Compute verifies the tasks that you submit. If a task with the same ID enters the system, the request is rejected and error 409 is returned.

Set TaskID

You can submit a task for execution by using the Function Compute console, using Serverless Devs, or calling an API operation. If you need to configure TaskID, set the x-fc-async-task-id parameter when you trigger your function.

For more information about the sample code for setting TaskID, see Sample code.

Task monitoring

You can use the task monitoring metrics provided by Function Compute to monitor the status of asynchronous tasks in real time and detect exceptions such as errors and timeouts in a timely manner.

Prerequisites

A function is created. For more information, see Function creation.

Procedure

1. Log on to the Function Compute console. In the left-side navigation pane, click Functions.

2. In the top navigation bar, select a region. On the Functions page, click the desired function.

   On the details page of the function, click the Monitoring tab to view metrics related to asynchronous invocations and instance-level resource usage of asynchronous tasks. For more information about the metrics, see Monitoring metrics.

   • Click the Function Metrics tab to view metrics related to asynchronous invocations. The following figure shows an example:

     

   • Click the Instance Metrics tab to view instance-level resource usage of asynchronous tasks. The following figure shows an example:

     Note

     Before you enable instance-level metrics, you must enable the logging feature. For more information, see Configure the logging feature.

     

3. Configure metric-based alert rules.

   If you want to monitor a metric, such as memory usage and vCPU usage, click in the upper-right corner of the metric to go to the Create Metric-based Alert Rule panel. Configure alert conditions and thresholds based on your business requirements, and then click OK.

   The following figure provides an example on configuring an alert rule, in which an alert with Medium severity is reported to contacts by DingTalk, email, phone, and webhook around the clock if the backlog of asynchronous invocations of a function is greater than or equal to five.

   

Job orchestration

You can integrate Function Compute with CloudFlow so that asynchronous invocations of functions can be applied to task processes in large-scale and complex scenarios. You can orchestrate Function Compute asynchronous tasks based on the control logic of CloudFlow, such as Sequence, Choice, and Parallel. CloudFlow coordinates the execution of tasks according to the specified steps in a reliable manner, tracks state transitions of each task, and executes the defined retry logic when necessary to ensure smooth and successful task completion. For more information, see Integrate CloudFlow with Function Compute.

Prerequisites

• An asynchronous task is created. For more information, see Task management.

• Permissions to access CloudFlow are granted to Function Compute. For more information, see Grant Function Compute permissions to access other Alibaba Cloud services.

Procedure

1. Log on to the CloudFlow console. In the top navigation bar, select a region.

2. In the left-side navigation pane, click Workflows. On the Workflows page, click Create Workflow.

3. In the Create Workflow dialog box, select a method to create a workflow, and then select a mode for the workflow.

   CloudFlow allows you to create workflows in standard mode or express mode. For more information about the two modes, see Standard mode and express mode. For more information, see Create a workflow based on a blank canvas.

4. Edit the created flow.

   1. Edit the flow.

      You can use CloudFlow Studio or YAML to edit the flow. In this example, YAML is used.

      In the upper-right corner of the flow details page, click YAML. In the left-side code editor, enter code to be used by the flow and click Save.

      Sample code:

      Type: StateMachine
      Name: MyFlow
      SpecVersion: v1
      StartAt: InvokeFunction
      States:
        - Type: Task
          Name: InvokeFunction
          Action: FC:InvokeFunction
          TaskMode: RequestComplete
          Parameters:
            invocationType: Async
            resourceArn: acs:fc:us-west-1:103435468****:functions/func-i1****/LATEST
          End: true

   2. Configure a flow role.

      In the upper-right corner of the flow details page, click Configure Workflow, select a role from the Execution Role drop-down list in the Role Configuration section, and then click Save.

      Note

      Make sure that the role configured for the flow is attached the AliyunFCInvocationAccess policy.

5. On the flow details page, click Execute in the upper-right corner to trigger the flow and check whether the flow works as expected.