fix(worker): handle AckMsg add/advise/write COM failures and server death

Claude-Session: https://claude.ai/code/session_016J6isfcgqy4SB2RK3ZvFvT

src/ZB.MOM.WW.MxGateway.Worker.Tests/MxAccess/AlarmAckWriterTests.cs

@@ -93,6 +93,68 @@ public sealed class AlarmAckWriterTests
         Assert.Equal(1, server.Operations.Count(op => op == "Advise"));
     }
 
+    [Fact]
+    public void Acknowledge_WhenAdviseThrows_RejectsCleansUpAndReAddsNextTime()
+    {
+        RecordingAckServer server = new RecordingAckServer { AuthenticateResult = 5, AdviseException = new COMException("x", unchecked((int)0x80004005)) };
+        using AlarmAckWriter writer = new AlarmAckWriter(server, new MxAccessWriteCompletionCache(), () => { }, TimeSpan.FromMilliseconds(10), "MxGateway");
+
+        AlarmAckResult first = writer.Acknowledge("T.A.AckMsg", "one", Alice());
+        server.AdviseException = null;
+        writer.Acknowledge("T.A.AckMsg", "two", Alice());
+
+        Assert.Equal(AlarmAckOutcome.WriteRejected, first.Outcome);
+        Assert.Equal(unchecked((int)0x80004005), first.NativeStatus);
+        Assert.DoesNotContain("pw-secret", first.Diagnostic, StringComparison.Ordinal);
+        Assert.Contains("RemoveItem", server.Operations);
+        Assert.Equal(2, server.Operations.Count(op => op.StartsWith("AddItem", StringComparison.Ordinal)));
+    }
+
+    [Fact]
+    public void Acknowledge_WhenCachedWriteThrows_RejectsAndReAddsNextTime()
+    {
+        RecordingAckServer server = new RecordingAckServer { AuthenticateResult = 5 };
+        using AlarmAckWriter writer = new AlarmAckWriter(server, new MxAccessWriteCompletionCache(), () => { }, TimeSpan.FromMilliseconds(10), "MxGateway");
+        writer.Acknowledge("T.A.AckMsg", "one", Alice());
+        server.WriteException = new COMException("x", unchecked((int)0x80004005));
+
+        AlarmAckResult failed = writer.Acknowledge("T.A.AckMsg", "two", Alice());
+        server.WriteException = null;
+        writer.Acknowledge("T.A.AckMsg", "three", Alice());
+
+        Assert.Equal(AlarmAckOutcome.WriteRejected, failed.Outcome);
+        Assert.Equal(2, server.Operations.Count(op => op.StartsWith("AddItem", StringComparison.Ordinal)));
+    }
+
+    [Fact]
+    public void Acknowledge_WhenWriteDisconnected_TearsDownServer()
+    {
+        // The test seam has no COM factory, so re-registration itself is not
+        // exercised here; this verifies the teardown that makes it possible.
+        RecordingAckServer server = new RecordingAckServer { AuthenticateResult = 5, WriteException = new COMException("x", unchecked((int)0x80010108)) };
+        using AlarmAckWriter writer = new AlarmAckWriter(server, new MxAccessWriteCompletionCache(), () => { }, TimeSpan.FromMilliseconds(10), "MxGateway");
+
+        AlarmAckResult result = writer.Acknowledge("T.A.AckMsg", "one", Alice());
+
+        Assert.Equal(AlarmAckOutcome.WriteRejected, result.Outcome);
+        Assert.Equal(unchecked((int)0x80010108), result.NativeStatus);
+        Assert.Contains("Unregister", server.Operations);
+    }
+
+    [Fact]
+    public void Acknowledge_WhenNonOkCompletionHasNoDetail_ReturnsMinusOne()
+    {
+        RecordingAckServer server = new RecordingAckServer { AuthenticateResult = 5 };
+        MxAccessWriteCompletionCache cache = new MxAccessWriteCompletionCache();
+        server.OnWrite = (handle, item) => cache.Record(handle, item, Rows(MxStatusCategory.SecurityError, 0));
+        using AlarmAckWriter writer = new AlarmAckWriter(server, cache, () => { }, TimeSpan.FromSeconds(1), "MxGateway");
+
+        AlarmAckResult result = writer.Acknowledge("T.A.AckMsg", "checked", Alice());
+
+        Assert.Equal(AlarmAckOutcome.WriteRejected, result.Outcome);
+        Assert.Equal(-1, result.NativeStatus);
+    }
+
     [Fact]
     public void AlarmAckOperator_ToString_OmitsPassword()
     {
@@ -110,6 +172,10 @@ internal sealed class RecordingAckServer : IMxAccessServer
 
     public Exception? AuthenticateException { get; set; }
 
+    public Exception? AdviseException { get; set; }
+
+    public Exception? WriteException { get; set; }
+
     public string? LastPassword { get; private set; }
 
     public Action<int, int>? OnWrite { get; set; }
@@ -132,7 +198,14 @@ internal sealed class RecordingAckServer : IMxAccessServer
 
     public void RemoveItem(int serverHandle, int itemHandle) => Operations.Add("RemoveItem");
 
-    public void Advise(int serverHandle, int itemHandle) => Operations.Add("Advise");
+    public void Advise(int serverHandle, int itemHandle)
+    {
+        Operations.Add("Advise");
+        if (AdviseException is not null)
+        {
+            throw AdviseException;
+        }
+    }
 
     public void UnAdvise(int serverHandle, int itemHandle) => Operations.Add("UnAdvise");
 
@@ -143,6 +216,11 @@ internal sealed class RecordingAckServer : IMxAccessServer
     public void Write(int serverHandle, int itemHandle, object? value, int userId)
     {
         Operations.Add("Write:" + value + ":" + userId);
+        if (WriteException is not null)
+        {
+            throw WriteException;
+        }
+
         OnWrite?.Invoke(serverHandle, itemHandle);
     }
 

src/ZB.MOM.WW.MxGateway.Worker/MxAccess/AlarmAckWriter.cs

@@ -111,17 +111,48 @@ public sealed class AlarmAckWriter : IAlarmAckWriter
                 $"Operator authentication failed for '{ackOperator.VerifyUser}' (no user id returned).");
         }
 
-        if (!itemHandles.TryGetValue(itemAddress, out int itemHandle))
+        ulong baselineVersion = 0;
+        int itemHandle = 0;
+        bool cached = itemHandles.TryGetValue(itemAddress, out itemHandle);
+        string step = "AddItem";
+        try
         {
-            itemHandle = mx.AddItem(serverHandle, itemAddress);
+            if (!cached)
-            // MXAccess rejects writes to added-but-unadvised items with E_INVALIDARG.
+            {
-            mx.Advise(serverHandle, itemHandle);
+                itemHandle = mx.AddItem(serverHandle, itemAddress);
-            itemHandles[itemAddress] = itemHandle;
+                itemHandles[itemAddress] = itemHandle;
+                step = "Advise";
+                // MXAccess rejects writes to added-but-unadvised items with E_INVALIDARG.
+                mx.Advise(serverHandle, itemHandle);
+            }
+
+            pumpStep();
+            step = "Write";
+            baselineVersion = completionCache.CurrentVersion(serverHandle, itemHandle);
+            mx.Write(serverHandle, itemHandle, comment ?? string.Empty, userId);
+        }
+        catch (COMException ex)
+        {
+            int hresult = ex.HResult;
+            if (itemHandles.Remove(itemAddress))
+            {
+                try { mx.UnAdvise(serverHandle, itemHandle); } catch { /* best effort */ }
+                try { mx.RemoveItem(serverHandle, itemHandle); } catch { /* best effort */ }
+            }
+
+            if (IsServerGone(hresult))
+            {
+                TearDownServer();
+            }
+
+            // Diagnostic names the step and HRESULT only: exception text can echo the item or operator.
+            return AlarmAckResult.Create(
+                AlarmAckOutcome.WriteRejected,
+                hresult,
+                $"AckMsg {step} failed (HRESULT 0x{hresult:X8}).");
         }
 
-        pumpStep();
+        ulong baseline = baselineVersion;
-        ulong baseline = completionCache.CurrentVersion(serverHandle, itemHandle);
-        mx.Write(serverHandle, itemHandle, comment ?? string.Empty, userId);
 
         if (!completionCache.TryWaitForCompletion(
                 serverHandle,
@@ -158,6 +189,17 @@ public sealed class AlarmAckWriter : IAlarmAckWriter
         if (disposed) return;
         disposed = true;
 
+        TearDownServer();
+    }
+
+    private static bool IsServerGone(int hresult) =>
+        hresult == unchecked((int)0x80010108) // RPC_E_DISCONNECTED
+        || hresult == unchecked((int)0x800706BA) // RPC_S_SERVER_UNAVAILABLE
+        || hresult == unchecked((int)0x80010105); // RPC_E_SERVERFAULT
+
+    /// <summary>Releases the MXAccess registration so the next call re-registers (also used by Dispose).</summary>
+    private void TearDownServer()
+    {
         if (comEventSource is not null)
         {
             try { comEventSource.OnWriteComplete -= OnWriteComplete; } catch { /* best effort */ }
@@ -190,7 +232,12 @@ public sealed class AlarmAckWriter : IAlarmAckWriter
     {
         if (server is not null) return server;
 
-        object created = factory!.Create() ?? throw new InvalidOperationException("MXAccess COM factory returned null.");
+        if (factory is null)
+        {
+            throw new InvalidOperationException("MXAccess server is gone and no COM factory is available to re-register.");
+        }
+
+        object created = factory.Create() ?? throw new InvalidOperationException("MXAccess COM factory returned null.");
         try
         {
             comObject = created;