Question

You are developing a nightly data aggregation pipeline using Azure Durable Functions (.NET Isolated process). The orchestrator function is triggered daily to collect sales reports from multiple regional API endpoints, consolidate them, and write the summary to a database.

The orchestrator code is defined as follows:

csharp
[Function("AggregateDailySalesOrchestrator")]
public static async Task<SalesSummary> Run(
[OrchestrationTrigger] TaskOrchestrationContext context)
{
var regions = await context.CallActivityAsync<List<string>>("GetActiveRegions", null);

var tasks = new List<Task<RegionReport>>();
foreach (var region in regions)
{
tasks.Add(context.CallActivityAsync<RegionReport>("FetchRegionReport", region));
}

// Fan-out: wait for all parallel fetch operations to complete
var reports = await Task.WhenAll(tasks);

// Fan-in processing
var summary = new SalesSummary { RunId = Guid.NewGuid().ToString() };
foreach (var report in reports)
{
if (report.Timestamp >= DateTime.UtcNow.AddHours(-24))
{
summary.TotalRevenue += report.Revenue;
}
}

await context.CallActivityAsync("SaveSummaryToDatabase", summary);
return summary;
}

During testing under heavy loads, you observe that the `FetchRegionReport` activity function occasionally takes up to $12$ minutes to complete due to slow external APIs, resulting in execution failures. The Function App is currently deployed to an Azure Functions Consumption plan.

Which action must you take to ensure that the orchestration runs successfully, executes without timeouts, and produces deterministic results?