Reduce thrown exceptions

src/Aspire.Hosting/Dcp/DcpExecutor.cs

@@ -56,7 +56,7 @@ internal sealed class DcpExecutor : IDcpExecutor, IAsyncDisposable
     private readonly ResourceSnapshotBuilder _snapshotBuilder;
 
     // Internal for testing.
-    internal ResiliencePipeline DeleteResourceRetryPipeline { get; set; }
+    internal ResiliencePipeline<bool> DeleteResourceRetryPipeline { get; set; }
     internal ResiliencePipeline CreateServiceRetryPipeline { get; set; }
     internal ResiliencePipeline WatchResourceRetryPipeline { get; set; }
 
@@ -1485,18 +1485,24 @@ async Task EnsureResourceDeletedAsync<T>(string resourceName) where T : CustomRe
             // before resorting to more extreme measures.
             if (!resourceNotFound)
             {
-                await DeleteResourceRetryPipeline.ExecuteAsync(async (state, attemptCancellationToken) =>
+                var result = await DeleteResourceRetryPipeline.ExecuteAsync<bool, string>(async (state, attemptCancellationToken) =>
                 {
                     try
                     {
                         await _kubernetesService.GetAsync<T>(state, cancellationToken: attemptCancellationToken).ConfigureAwait(false);
-                        throw new DistributedApplicationException($"Failed to delete '{state}' successfully before restart.");
+                        return false;
                     }
                     catch (HttpOperationException ex) when (ex.Response.StatusCode == System.Net.HttpStatusCode.NotFound)
                     {
                         // Success.
+                        return true;
                     }
                 }, resourceName, cancellationToken).ConfigureAwait(false);
+
+                if (!result)
+                {
+                    throw new DistributedApplicationException($"Failed to delete '{resourceName}' successfully before restart.");
+                }
             }
         }
     }

src/Aspire.Hosting/Dcp/DcpPipelineBuilder.cs

@@ -10,24 +10,24 @@ namespace Aspire.Hosting.Dcp;
 
 internal static class DcpPipelineBuilder
 {
-    public static ResiliencePipeline BuildDeleteRetryPipeline(ILogger logger)
+    public static ResiliencePipeline<bool> BuildDeleteRetryPipeline(ILogger logger)
     {
-        var ensureDeleteRetryStrategy = new RetryStrategyOptions()
+        var ensureDeleteRetryStrategy = new RetryStrategyOptions<bool>()
         {
             BackoffType = DelayBackoffType.Exponential,
             Delay = TimeSpan.FromMilliseconds(200),
             UseJitter = true,
             MaxRetryAttempts = 10, // Cumulative time for all attempts amounts to about 15 seconds
             MaxDelay = TimeSpan.FromSeconds(3),
-            ShouldHandle = new PredicateBuilder().Handle<Exception>(),
+            ShouldHandle = args => ValueTask.FromResult(!args.Outcome.Result),
             OnRetry = (retry) =>
             {
                 logger.LogDebug("Retrying check for deleted resource. Attempt: {Attempt}. Error message: {ErrorMessage}", retry.AttemptNumber, retry.Outcome.Exception?.Message);
                 return ValueTask.CompletedTask;
             }
         };
 
-        var execution = new ResiliencePipelineBuilder().AddRetry(ensureDeleteRetryStrategy).Build();
+        var execution = new ResiliencePipelineBuilder<bool>().AddRetry(ensureDeleteRetryStrategy).Build();
         return execution;
     }
 

src/Aspire.Hosting/Health/ResourceHealthCheckService.cs

@@ -293,7 +293,6 @@ internal class ResourceMonitorState
         private readonly object _lock = new object();
         private readonly string _resourceName;
         private TaskCompletionSource? _delayInterruptTcs;
-        private CancellationTokenSource? _delayCts;
 
         public ResourceMonitorState(ILogger logger, ResourceEvent initialEvent, CancellationToken serviceStoppingToken)
         {
@@ -334,6 +333,7 @@ public void SetLatestEvent(ResourceEvent resourceEvent)
 
         internal async Task<bool> DelayAsync(ResourceEvent? currentEvent, TimeSpan delay, CancellationToken cancellationToken)
         {
+            Task delayInterruptedTask;
             lock (_lock)
             {
                 // The event might have changed before delay was called. Interrupt immediately if required.
@@ -342,26 +342,16 @@ internal async Task<bool> DelayAsync(ResourceEvent? currentEvent, TimeSpan delay
                     _logger.LogTrace("Health monitoring delay interrupted for resource '{Resource}'.", _resourceName);
                     return true;
                 }
-                if (_delayCts == null || !_delayCts.TryReset())
-                {
-                    _delayCts = CancellationTokenSource.CreateLinkedTokenSource(cancellationToken);
-                }
                 _delayInterruptTcs = new TaskCompletionSource(TaskCreationOptions.RunContinuationsAsynchronously);
+                delayInterruptedTask = _delayInterruptTcs.Task;
             }
 
-            var completedTask = await Task.WhenAny(Task.Delay(delay, _delayCts.Token), _delayInterruptTcs.Task).ConfigureAwait(false);
+            // Don't throw to avoid writing the thrown exception to the debug console.
+            // See https://github.com/dotnet/aspire/issues/7486
+            await delayInterruptedTask.WaitAsync(delay, cancellationToken).ConfigureAwait(ConfigureAwaitOptions.SuppressThrowing);
+            var delayInterrupted = delayInterruptedTask.IsCompletedSuccessfully == true;
 
-            if (completedTask != _delayInterruptTcs.Task)
-            {
-                // Task.Delay won.
-                return false;
-            }
-            else
-            {
-                // Delay was interrupted. Cancel the delay task so it doesn't hang around when not needed.
-                _delayCts.Cancel();
-                return true;
-            }
+            return delayInterrupted;
         }
 
         private static bool ShouldInterrupt(ResourceEvent currentEvent, ResourceEvent previousEvent)

tests/Aspire.Hosting.Tests/Dcp/DcpExecutorTests.cs

@@ -1106,7 +1106,7 @@ public async Task ErrorIfResourceNotDeletedBeforeRestart()
         var appExecutor = CreateAppExecutor(distributedAppModel, kubernetesService: kubernetesService, events: dcpEvents);
 
         // Set a custom pipeline without retries or delays to avoid waiting.
-        appExecutor.DeleteResourceRetryPipeline = new ResiliencePipelineBuilder().Build();
+        appExecutor.DeleteResourceRetryPipeline = new ResiliencePipelineBuilder<bool>().Build();
 
         await appExecutor.RunApplicationAsync();