DocsICP FilingConsole
官方文档
首页

Origin probing

更新时间：
复制 MD 格式
产品详情
我的收藏

Global Accelerator provides the origin probing feature. You can create origin probing tasks for a listener to monitor the end-to-end network quality from detection points to an origin server through Global Accelerator. This helps you quickly identify network faults and optimize your network.

Origin probing overview

You can create an origin probing task for a listener of a Global Accelerator instance. The task sends probing requests that simulate real user access from detection points distributed across multiple regions to monitor access to your origin server and measure its availability.

By default, an origin probing task runs every 5 minutes. You can also use the following features to monitor availability anomalies in real time:

  • Network diagnostics

    • Diagnose now: If your origin server's availability drops, you can use this feature to troubleshoot network link issues.

    • Set automatic diagnostics: When availability falls below a specified threshold, a diagnostic is automatically triggered. This captures data about the anomaly to help with future troubleshooting.

  • Create an alert rule

    You can create an alert rule for your origin probing task. When availability drops below a specified threshold, an alert is triggered, enabling real-time monitoring of the end-to-end network quality.

image

Billing

The origin probing feature is free. If you set an alert rule for an origin probing task, you may be charged for alert notifications sent by SMS or phone calls. For more information, see Billing overview.

Limitations

  • You can create origin probing tasks only for standard Global Accelerator instances of Large I specifications or higher.

  • You cannot create origin probing tasks for UDP listeners.

  • The port specified for the monitored host must be within the port range of the listener.

  • Detection points support probing only over HTTP or HTTPS and use IPv4.

  • For each Global Accelerator instance, you can create a maximum of five origin probing tasks.

Prerequisites

  • Before you create an origin probing task, make sure that the Global Accelerator instance is in the Available state.

  • If your GA instance uses the subscription billing method, you must have purchased and associated a basic bandwidth plan.

  • You have configured an acceleration area, a listener, and an endpoint group for the Global Accelerator instance. For more information, see the following topics:

Procedure

image

Create an origin probing task

  1. Log on to the GA console.

  2. In the left-side navigation pane, click Origin Probing.

  3. On the Origin Probing page, click Create Origin Probing Task. In the Create Origin Probing Task dialog box, configure the following parameters and click OK.

    Parameter

    Description

    Task Name

    Enter a name for the origin probing task.

    IP Protocol

    The IP protocol used by the origin probing task.

    Only IPv4 is supported.

    GA Instance

    Select the Global Accelerator instance that you want to monitor.

    Only standard Global Accelerator instances are supported.

    Listener

    Select the listener that you want to monitor.

    UDP listeners are not supported.

    Domain Name

    Enter the domain name or accelerated IP address that you want to monitor.

    You can also specify a custom port for monitoring. If you do not specify a port, the default port is 80 for HTTP and 443 for HTTPS. The port must be within the port range of the listener.

    • Enter a domain name

      • If you enter a domain name in the <domain>:<port> (port is optional) or http://<domain>:<port> (port is optional) format, you can specify a custom port. The default port is 80.

        Examples: example.com, http://example.com, and http://example.com:8080.

      • If you enter a domain name in the https://<domain>:<port> (port is optional) format, you can specify a custom port. The default port is 443.

        Examples: https://example.com and https://example.com:8443.

    • Enter an accelerated IP address

      • If you enter an accelerated IP address in the <accelerated IP address>:<port> (port is optional) or http://<accelerated IP address>:<port> (port is optional) format, you can specify a custom port. The default port is 80.

        Examples: 47.254.XX.XX, http://47.254.XX.XX, and http://47.0.XX.XX:22.

      • If you enter an accelerated IP address in the https://<accelerated IP address>:<port> (port is optional) format, you can specify a custom port. The default port is 443.

        Examples: https://47.254.XX.XX and https://47.254.XX.XX:8443.

    You can enter multiple monitoring addresses at once. Separate each address with a line break. Global Accelerator automatically creates a separate origin probing task for each address.

    Frequency

    The interval between probes to the origin server.

    The default value is 5 Minutes. Detection points monitor the target address once every 5 minutes.

    Automatic Diagnostics

    Specifies whether to enable automatic diagnostics.

    If you enable this feature, the system automatically triggers network diagnostics when the specified conditions are met.

    Trigger Condition

    The condition that triggers automatic diagnostics, based on an availability threshold.

    You can configure the Trigger Condition parameter only after you enable Automatic Diagnostics.

    For example, Availability <90%Trigger Network Link Diagnostics means that when the availability of the origin server is less than 90%, the system automatically triggers a network link diagnosis from the detection point to the origin server.

    Advanced Settings

    HTTP Method

    Configure the request method for the origin probing task.

    1. Select a request method. Valid values are GET, POST, and HEAD.

    2. Configure the following parameters based on the selected request method:

      • Request Content: The content of the request sent by the origin probing task. The format can be key-value pairs such as key1=value1;key2=value2, or a JSON string such as {"test":"testValue"}.

        This parameter is required only when HTTP Method is set to POST.

      • HTTP Header: The header information for the request. The format is key1:value1. Use line breaks to separate multiple headers.

      • Acceptable HTTP Status Codes >=400: HTTP status codes of 400 or greater that should be considered successful. While these codes typically indicate an error, you can specify any that represent a healthy status for your application. Separate multiple codes with commas (,).

    Enable SNI

    Specifies whether to enable Server Name Indication (SNI). This feature is disabled by default.

    This feature allows a server to present multiple SSL certificates from a single IP address.

  4. In the Create Origin Probing Task dialog box, click Close.

    To monitor for anomalies in real time, wait 3 to 5 minutes, and then click View How to Create Alert to navigate to the CloudMonitor console and set up the relevant alert rules. For more information, see Create an alert rule.

Network diagnostics

  1. Log on to the GA console.

  2. In the left-side navigation pane, click Origin Probing.

  3. On the Origin Probing page, find the origin probing task and click its ID.

  4. On the origin probing task details page, click the Network Diagnostics tab, and then select a diagnostic method.

    • Diagnose: Click Diagnose.

    • Set Automatic Diagnostics: Click Set Automatic Diagnostics. In the Set Automatic Diagnostics dialog box, set the Status and Trigger Condition, and then click OK.

    Note

    After you select a diagnostic method, the system immediately starts the diagnosis. The process takes about one minute to complete.

  5. On the Network Diagnostics tab, view the diagnostic results.

    By default, the system displays diagnostic results from the last hour. You can also specify a custom time range to view results from that period. The system can display diagnostic results from the last 30 days.诊断结果

    Item

    Description

    Diagnostic Status

    • Successful: The diagnosis completed without finding any issues.

    • Failed: The diagnosis found an issue. To troubleshoot, check the HTTP Diagnostic Code, HTTP Diagnostic Content, and Diagnostic Result.

    HTTP Diagnostic Code

    Records the value returned by the origin server.

    • 2xx: Success status code. The request was successfully processed.

    • 3xx: Redirection status code. Further action is needed to complete the request.

    • 4xx: Client error status code. The server cannot process the request.

    • 5xx: Server error status code. The server failed to process the request.

    • 600: Internal error. For example, a listener is not configured. See the corresponding diagnostic result for details.

    For more information about the meanings of 2xx to 5xx status codes, see Common HTTP status codes.

    HTTP Diagnostic Content

    Records the response information obtained from the network diagnosis.

    The information varies based on the cause of the issue. The response may come from the origin server. For example, if the origin server cannot find the requested page, it returns a diagnostic content that contains 404 Not Found. The response may also come from a detection point. For example, if the origin server port is not open, the response contains diagnostic content such as Get http://www.aliyun.com/: dial tcp 106.11.XX.XX:80: connect: connection refused.

    Diagnostic Result

    A description of the diagnostic result for the origin probing task.

    • All forward nodes work well.: No issues were found.

    • Endpoint network error.: An issue occurred on the origin server. Check if your application on the origin server is running correctly.

    • Public network error.: A public network error occurred. This usually refers to a network error between the client and the acceleration area.

    • GA internal error.: An internal error occurred. For example, the Global Accelerator instance encountered an exception while processing a request.

    • GA has been deleted.: The current Global Accelerator instance has been deleted.

    • GA state is not stable: The current Global Accelerator instance is in an unstable state, such as Configuring.

    • GA has no listener configuration.: No listener is configured for the current Global Accelerator instance.

    • Missing endpoint configuration.: An endpoint is not configured.

    • Missing acceleration region configuration.: An acceleration area is not configured.

    • Missing endpoint group configuration.: An endpoint group is not configured.

Create an alert rule

  1. Log on to the GA console.

  2. In the left-side navigation pane, click Origin Probing.

  3. On the Origin Probing page, click Create Alert Rule.

  4. On the Alert Rule page of the CloudMonitor console, click Create Alert Rule.

  5. In the Create Alert Rule panel, configure the parameters to create an alert rule, and then click Confirm.

    The following table lists only the parameters that are highly relevant to this topic. For information about other parameters, see Create an alert rule.

    Parameter

    Description

    Service

    Select Global Accelerator.

    Resource Range

    Select the scope of resources to which the alert rule applies.

    • All Resources: The alert rule applies to all standard Global Accelerator instances within the current Alibaba Cloud account.

    • Application Groups: The alert rule applies to all resources within a specified application group in the current Alibaba Cloud account.

    • Instance: The alert rule applies to specified standard Global Accelerator instances within the current Alibaba Cloud account.

    If you select Application Groups or Instance, you must also specify the Associate Resource. Select the application group that contains the target standard Global Accelerator instance, or select the target standard Global Accelerator instance itself.

    Description

    The body of the alert rule. An alert is triggered when the monitoring data meets the alert condition. To set up the rule description:

    1. Click Create Rule.

    2. In the Add Rule Description panel, set parameters such as Rule Name, Metric Type, Metrics, and Threshold and Alert Level.

      This topic describes how to set an alert rule for origin server availability. Take note of the following parameters:

      • Metrics: Select standard_task > GaApplicationMonitorAvailability.

      • Dimension: Select the ID of the target origin probing task.

    3. Click OK.

View the origin probing results

  1. Log on to the GA console.

  2. In the left-side navigation pane, click Origin Probing.

  3. On the Origin Probing page, find the origin probing task and click its ID.

  4. On the task details page, click the Monitor tab to view the results.

    探测结果

    • The Overview section shows information such as origin server availability and alerts.

      To view alert rules and alert history, you must first create a threshold-based alert rule for the origin probing task. For more information, see Create an alert rule.

    • In the Availability Analysis section, you can select a time range to view availability statistics for that period.

    • In the Probing Result (Last 6 Hours) section, you can filter results by internet service provider (ISP) and detection point to view specific details. You can call the DescribeApplicationMonitor API operation to view the detection points that correspond to an origin probing task.

More operations

Actions

Description

Modify an origin probing task

  1. On the Origin Probing page, find the target task and click Edit in the Actions column.

  2. In the Modify Origin Probing Task dialog box, you can modify the task name, listener, monitored host, and advanced settings.

    For more information, see Create an origin probing task.

Enable or disable an origin probing task

On the Origin Probing page, find the target task and enable or disable it by using the toggle in the Status column.

Note

If you delete a listener that is associated with an origin probing task, the task is also disabled.

Delete an origin probing task

  1. On the Origin Probing page, find the target task and click Delete in the Actions column.

  2. In the confirmation dialog box, click OK.

Examples

Suppose a website, example.com, uses Global Accelerator to improve its user access experience. You can use the instance monitoring feature of Global Accelerator to view bandwidth and connectivity data between acceleration areas and endpoint regions.

To get more complete and timely visibility into end-to-end network quality from clients to the origin server, the website also configures origin probing.

Default probing

By default, an origin probing task runs every 5 minutes to provide basic monitoring of origin server availability. The following figure shows the origin probing configuration for this example.源站探测

Real-time monitoring

The website enables automatic diagnostics and creates an alert rule to monitor origin server issues in real time.

  • The following figure shows the automatic diagnostics configuration in this example. When availability drops below 95%, the system automatically triggers network link diagnostics and retains data about the anomaly.自动诊断

  • The following figure shows the alert rule configuration in this example. When availability drops below the configured threshold, the system sends alert notifications based on the alert level. For example, if availability stays below 80% for three consecutive periods, it triggers an urgent alert notification.报警规则

References

Previous：Instance diagnosisNext：Anti-DDoS Basic