fix(grpc): count a worker fault once, not once per StreamEvents subscriber

WorkerClient.SetFaulted already records mxgateway.faults; EventStreamService added a
WorkerFaulted fault per subscriber, so one fault read as 1+K. The subscriber now records
only its WorkerFault disconnect reason. docs/Metrics.md updated.

src/ZB.MOM.WW.MxGateway.Server/Grpc/EventStreamService.cs

@@ -174,12 +174,13 @@ public sealed class EventStreamService(
                 {
                     // The distributor pump completes every subscriber channel with the source
                     // fault when the worker event stream terminates abnormally; that surfaces
-                    // here. Mirror the original ProduceEventsAsync behavior: fault the
-                    // session and record the metric, then propagate the terminal fault to the
-                    // gRPC client.
+                    // here. Fault the session and propagate the terminal fault to the gRPC
+                    // client. The fault itself is NOT counted here: WorkerClient.SetFaulted
+                    // already recorded it once, and every subscriber of the session sees the
+                    // same fault, so counting it per subscriber inflated mxgateway.faults to
+                    // 1+K. This subscriber's share is the WorkerFault disconnect reason.
                     disconnectReason = EventStreamDisconnectReasons.WorkerFault;
                     session.MarkFaulted(workerException.Message);
-                    metrics.Fault(WorkerClientErrorCode.WorkerFaulted.ToString());
                     throw;
                 }
                 catch (SessionManagerException sessionException)

src/ZB.MOM.WW.MxGateway.Tests/Gateway/Grpc/EventStreamServiceTests.cs

@@ -358,7 +358,63 @@ public sealed class EventStreamServiceTests
 
         Assert.Equal(WorkerClientErrorCode.WorkerFaulted, exception.ErrorCode);
         Assert.Equal(SessionState.Faulted, session.State);
+
+        // The worker fault is counted once, by the worker client that faulted (this fake does
+        // not count it); a subscriber observing it records a disconnect reason, not a fault.
+        Assert.Equal(0, metrics.GetSnapshot().Faults);
+    }
+
+    /// <summary>
+    ///     One worker fault is one <c>mxgateway.faults</c> increment however many
+    ///     <c>StreamEvents</c> subscribers observe it: <see cref="WorkerClient"/> records it when it
+    ///     faults, and each subscriber records only its <c>WorkerFault</c> disconnect reason.
+    ///     Counting it per subscriber as well inflated the metric to 1+K for K subscribers.
+    /// </summary>
+    /// <returns>A task that represents the asynchronous operation.</returns>
+    [Fact]
+    public async Task StreamEventsAsync_WhenWorkerFaultsUnderTwoSubscribers_CountsTheFaultOnce()
+    {
+        await using Workers.Fakes.FakeWorkerHarness fakeWorker = await Workers.Fakes.FakeWorkerHarness.CreateConnectedPairAsync();
+        using GatewayMetrics metrics = new();
+        using DisconnectReasonRecorder reasons = new(metrics);
+        WorkerClient client = fakeWorker.CreateClient(metrics: metrics);
+        Task startTask = client.StartAsync(CancellationToken.None);
+        await fakeWorker.CompleteStartupAsync();
+        await startTask.WaitAsync(TestTimeout);
+        GatewayOptions options = new()
+        {
+            Sessions = new SessionOptions { AllowMultipleEventSubscribers = true },
+        };
+        SessionManager manager = new(
+            new SessionRegistry(),
+            new SingleClientFactory(client),
+            Options.Create(options),
+            metrics);
+        GatewaySession session = await manager.OpenSessionAsync(
+            new SessionOpenRequest(RequestedBackend: null, ClientSessionName: "fault-count", ClientCorrelationId: null, CommandTimeout: null),
+            "client-1",
+            ownerKeyId: null,
+            CancellationToken.None);
+        EventStreamService service = new(manager, Options.Create(options), metrics);
+        await using IAsyncEnumerator<MxEvent> first = service
+            .StreamEventsAsync(CreateRequest(session.SessionId), callerKeyId: null, CancellationToken.None)
+            .GetAsyncEnumerator();
+        await using IAsyncEnumerator<MxEvent> second = service
+            .StreamEventsAsync(CreateRequest(session.SessionId), callerKeyId: null, CancellationToken.None)
+            .GetAsyncEnumerator();
+        Task<bool> firstMove = first.MoveNextAsync().AsTask();
+        Task<bool> secondMove = second.MoveNextAsync().AsTask();
+        await WaitUntilAsync(() => session.ActiveEventSubscriberCount == 2);
+
+        await fakeWorker.DisposeWorkerSideAsync();
+
+        await Assert.ThrowsAsync<WorkerClientException>(async () => await firstMove.WaitAsync(TestTimeout));
+        await Assert.ThrowsAsync<WorkerClientException>(async () => await secondMove.WaitAsync(TestTimeout));
+        Assert.Equal(WorkerClientState.Faulted, client.State);
         Assert.Equal(1, metrics.GetSnapshot().Faults);
+        Assert.Equal(
+            [EventStreamDisconnectReasons.WorkerFault, EventStreamDisconnectReasons.WorkerFault],
+            reasons.Reasons);
     }
 
     /// <summary>
@@ -1038,6 +1094,15 @@ public sealed class EventStreamServiceTests
     ///     measurement recorded on one <see cref="GatewayMetrics"/> instance (filtered by meter
     ///     reference, so parallel tests sharing the meter name cannot cross-talk).
     /// </summary>
+    private sealed class SingleClientFactory(IWorkerClient client) : ISessionWorkerClientFactory
+    {
+        /// <inheritdoc />
+        public Task<IWorkerClient> CreateAsync(GatewaySession session, CancellationToken cancellationToken)
+        {
+            return Task.FromResult(client);
+        }
+    }
+
     private sealed class DisconnectReasonRecorder : IDisposable
     {
         private readonly System.Diagnostics.Metrics.MeterListener _listener = new();