Configure a retry policy

Prerequisites

• A cloud-native gateway instance is created.

• At least one route is configured for the gateway.

Route retry conditions

When a request to a backend service fails, the cloud-native gateway can automatically retry the request based on conditions that you configure. Retry conditions are protocol-specific. The following list describes the available conditions for HTTP and gRPC protocols.

• The following retry conditions are available for the HTTP protocol:

  • 5xx: The cloud-native gateway retries the failed request if the backend service returns any 5xx response, or if a disconnection, reset, or read timeout occurs.

    Note

    The 5xx condition includes the connect-failure and refused-stream conditions.

  • reset: The cloud-native gateway retries the failed request if a disconnection, reset, or read timeout occurs.

  • connect-failure: The cloud-native gateway retries the failed request if the request fails because of a disconnection.

  • refused-stream: The cloud-native gateway retries the failed request if the backend service resets the stream with the REFUSED_STREAM error code.

  • retriable-status-codes: The cloud-native gateway retries the request if the HTTP status code in the response from the backend service matches a retry status code that you specified.

    Note

    You can use retry status codes only if you specify retriable-status-codes as a retry condition.

• The following retry conditions are available for the gRPC protocol:

  • cancelled: The cloud-native gateway retries the request if the gRPC status code in the response header of the backend gRPC service is cancelled.

  • deadline-exceeded: The cloud-native gateway retries the request if the gRPC status code in the response header of the backend gRPC service is deadline-exceeded.

  • internal: The cloud-native gateway retries the request if the gRPC status code in the response header of the backend gRPC service is internal.

  • resource-exhausted: The cloud-native gateway retries the request if the gRPC status code in the response header of the backend gRPC service is resource-exhausted.

  • unavailable: The cloud-native gateway retries the request if the gRPC status code in the response header of the backend gRPC service is unavailable.

Configure a route retry policy

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

2. In the left-side navigation pane, choose Cloud-native Gateway > Gateways. On the Gateways page, click the ID of the gateway.

3. In the left-side navigation pane, click Routes. Then, click the Routes tab.

4. Find the route that you want to modify and click Policies in the Actions column.

5. On the Policies tab, click Retry.

6. Configure the parameters in the Retry rule and then click Save.

   Parameter	Description
   Retry Times	The maximum number of times a failed request is retried. Valid values: 0 to 10. A value of no more than 2 is recommended. Note A value of 0 disables retries.
   Retry Condition	Select the retry conditions that suit your workload. You can select multiple conditions, which are described in Route retry conditions.
   Retry Status Code	Specify the HTTP status codes whose responses are retried. You can configure multiple HTTP status codes. Note You must specify retriable-status-codes for the Retry Condition before you can configure the Retry Status Code.
   Enable	The enable/disable toggle of the route retry policy. Enable: The route retry policy of the gateway takes effect. Disable: The custom route retry policy does not take effect. Note After you disable the custom retry policy, the gateway falls back to its internal default retry configuration. The default number of retries is 2, and the default retry conditions are connect-failure, refused-stream, unavailable, cancelled, and retriable-status-codes.

   After you configure and enable the retry policy, send test requests to the route and verify that failed requests are retried according to the configured conditions and retry count.