fix(communication): resolve Communication-002/003 — gRPC reconnect stream cleanup and subscription map safety

src/ScadaLink.Communication/Actors/DebugStreamBridgeActor.cs

@@ -183,6 +183,15 @@ public class DebugStreamBridgeActor : ReceiveActor, IWithTimers
             return;
         }
 
+        // Unsubscribe the failed stream on the *previous* endpoint before reconnecting.
+        // This cancels the local subscription CTS and -- where the channel is still
+        // alive -- propagates gRPC cancellation to the site so its SiteStreamGrpcServer
+        // stops the StreamRelayActor for this correlation ID, rather than leaving a
+        // zombie relay actor until TCP RST / keepalive eventually detects the loss.
+        var previousEndpoint = _useNodeA ? _grpcNodeAAddress : _grpcNodeBAddress;
+        var previousClient = _grpcFactory.GetOrCreate(_siteIdentifier, previousEndpoint);
+        previousClient.Unsubscribe(_correlationId);
+
         // Flip to the other node
         _useNodeA = !_useNodeA;
 

src/ScadaLink.Communication/Grpc/SiteStreamGrpcClient.cs

@@ -57,6 +57,32 @@ public class SiteStreamGrpcClient : IAsyncDisposable
         _subscriptions[correlationId] = cts;
     }
 
+    /// <summary>
+    /// Registers a subscription's CancellationTokenSource for a correlation ID.
+    /// If an entry already exists for that correlation ID (a reconnect race where two
+    /// <see cref="SubscribeAsync"/> calls briefly share an ID), the prior CTS is
+    /// cancelled and disposed so it cannot leak. Internal for testability.
+    /// </summary>
+    internal void RegisterSubscription(string correlationId, CancellationTokenSource cts)
+    {
+        if (_subscriptions.TryGetValue(correlationId, out var prior) && !ReferenceEquals(prior, cts))
+        {
+            prior.Cancel();
+            prior.Dispose();
+        }
+        _subscriptions[correlationId] = cts;
+    }
+
+    /// <summary>
+    /// Removes the subscription entry for a correlation ID only if the stored CTS is
+    /// exactly the one supplied. A racing replacement stream may already own the slot,
+    /// in which case this is a no-op. Internal for testability.
+    /// </summary>
+    internal void RemoveSubscription(string correlationId, CancellationTokenSource cts)
+    {
+        _subscriptions.TryRemove(new KeyValuePair<string, CancellationTokenSource>(correlationId, cts));
+    }
+
     /// <summary>
     /// Opens a server-streaming subscription for a specific instance.
     /// This is a long-running async method; the caller launches it as a background task.
@@ -74,7 +100,7 @@ public class SiteStreamGrpcClient : IAsyncDisposable
             throw new InvalidOperationException("Cannot subscribe on a test-only client.");
 
         var cts = CancellationTokenSource.CreateLinkedTokenSource(ct);
-        _subscriptions[correlationId] = cts;
+        RegisterSubscription(correlationId, cts);
 
         var request = new InstanceStreamRequest
         {
@@ -103,7 +129,8 @@ public class SiteStreamGrpcClient : IAsyncDisposable
         }
         finally
         {
-            _subscriptions.TryRemove(correlationId, out _);
+            // Remove only our own entry -- a racing reconnect may already own the slot.
+            RemoveSubscription(correlationId, cts);
         }
     }