TaskOrchestrationContext.CallHttpAsync, isolated process, configure polling interval.

[DavidJFowler]

I'm calling an orchestration in an another durable function app using TaskOrchestrationContext.CallHttpAsync like this:

DurableHttpResponse response =
    await context.CallHttpAsync(HttpMethod.Post,
        claimStateUrl,
        content: null,
        retryOptions: null,
        asynchronousPatternEnabled: true);

I would like to change the default polling interval, but I can't work out how to do it. Is this possible?

[cgillum]

Looking at the code, it seems there are a couple options for configuring the polling interval.

azure-functions-durable-extension/src/Worker.Extensions.DurableTask/TaskOrchestrationContextExtensionMethods.cs

Lines 57 to 68 in ca4d724

	if (headersDictionary.TryGetValue("Retry-After", out StringValues retryAfterStr) && int.TryParse(retryAfterStr, out int retryAfter))
	{
	fireAt = context.CurrentUtcDateTime.AddSeconds(retryAfter);
	}
	else
	{
	// Gets configuration DefaultAsyncRequestSleepTimeMilliseconds from DurableTaskExtension.
	// If no value is provided, then use the default 30000 milliseconds.
	int asyncRequestSleepTimeMilliseconds = context.Properties.TryGetValue(PollingInterval, out var value) && value is double d
	? (int)d: DefaultPollingIntervalMilliseconds;
	fireAt = context.CurrentUtcDateTime.AddMilliseconds(asyncRequestSleepTimeMilliseconds);
	}

A Retry-After header on the server is one option. @nytian can you provide more information on the use of the DefaultAsyncRequestSleepTimeMilliseconds context property? I couldn't find documentation for this.

[DavidJFowler]

A Retry-After header on the server is one option.

That does not appear to work:

[Function(nameof(GetClaimState))]
public static async Task<HttpResponseData> GetClaimState(
    [HttpTrigger(AuthorizationLevel.Anonymous, "post", Route = "claimstate/claims/{claimId}")] HttpRequestData req,
    [DurableClient] DurableTaskClient client,
    FunctionContext executionContext,
    string claimId)
{
    ILogger logger = executionContext.GetLogger(nameof(GetClaimState));

    // Function input comes from the request content.
    string instanceId = await client.ScheduleNewOrchestrationInstanceAsync(
        nameof(GetStateOrchestration), claimId);

    logger.LogInformation("Started orchestration with ID = '{instanceId}'.", instanceId);

    // Returns an HTTP 202 response with an instance management payload.
    // See https://learn.microsoft.com/azure/azure-functions/durable/durable-functions-http-api#start-orchestration
    var response = await client.CreateCheckStatusResponseAsync(req, instanceId);
    response.Headers.Add("Retry-After", "10");  // Does not work, header is not returned with the response

    return response;
}

I found this post: https://dev.to/helasp/azure-functions-isolated-worker-http-trigger-custom-header-disappears-from-response-c32 but the workaround can't be made to work with CreateCheckStatusResponseAsync

[nytian]

To change the polling interval, update DefaultAsyncRequestSleepTimeMilliseconds value in HttpOptions section under DurableTaskOptions : https://github.com/Azure/azure-functions-durable-extension/blob/dev/src/WebJobs.Extensions.DurableTask/Options/HttpOptions.cs#L19

I will open a PR today to update our document for this HttpOptions section.

[DavidJFowler]

Hi,

Isn't that for in-process functions?

[nytian]

Hey, The isolated model is also set to use this same configuration.

[DavidJFowler]

Do you have an example of this?