A.3 (live smoke): full alarms-over-gateway pipeline verified end-to-end

Skip-gated AlarmsLiveSmokeTests.Alarms_full_pipeline_round_trip ran
against the dev rig with the flip script firing
TestMachine_001.TestAlarm001 every 10s. Verified:
  - Subscribe + 1st PollOnce yield real transition events
  - Field-by-field decode correct (provider, group, tag, severity,
    UTC timestamp, comment, type)
  - SnapshotActiveAlarms reflects current state
  - AcknowledgeByName(real identity) -> rc=0
  - Pipeline keeps streaming transitions on the 10s cadence post-ack

Three production quirks surfaced and were fixed in
WnWrapAlarmConsumer:

1. SetXmlAlarmQuery is mandatory for reads. Skipping it (per the
   earlier discovery-doc recommendation) makes the first
   GetXmlCurrentAlarms2 fail with E_FAIL. The doc's claim that the
   call is unnecessary because the round-trip echo is mangled was
   wrong — mangled echo or not, the call is required.

2. SetXmlAlarmQuery breaks AlarmAckByName on the same consumer
   instance (returns -55). Workaround: provision a parallel
   "ack-only" wnwrap consumer that runs Initialize → Register →
   Subscribe via the v1-prefixed methods, no SetXmlAlarmQuery.
   Production WnWrapAlarmConsumer now holds two COM clients;
   AcknowledgeByName always dispatches through the ack-only one.

3. AlarmAckByName has v2 (8-arg) and v1 (6-arg) overloads. The v2
   8-arg overload returns -55 on this AVEVA build (apparently a
   stub); the v1 6-arg overload works. Production now calls the
   6-arg overload, discarding the proto's operator_domain and
   operator_full_name fields. The proto contract keeps both for
   forward-compat if AVEVA fixes the v2 method.

Bonus finding (not fixed here): AlarmAckByGUID throws
NotImplementedException on wnwrap. Reference→GUID lookup that we
initially planned to plumb is therefore not viable; all acks must
go through AlarmAckByName. WorkerAlarmRpcDispatcher.AcknowledgeAsync
already routes references through the by-name path, so this only
affects the GUID-input branch (which the worker tries first if the
input parses as a GUID — that branch will surface
NotImplementedException as MxaccessFailure if a client supplies one).

Threading caveat: wnwrap is ThreadingModel=Apartment, so the
consumer's internal Timer (firing on threadpool threads) blocks on
cross-apartment marshaling without an STA message pump. The smoke
test sidesteps this with pollIntervalMilliseconds=0 (Timer disabled)
+ manual PollOnce calls from the test STA. Production hosting will
route polls through the worker's StaRuntime in a follow-up; PollOnce
is now public so the wire-up is straightforward.

Test counts after this slice:
  Worker: 195 pass / 4 skipped (live probes incl. new live smoke) /
          1 pre-existing structure-fail (untouched)
  Server: 308 pass / 0 fail
Solution builds clean.

docs/AlarmClientDiscovery.md "Live smoke-test discoveries" section
records all five findings.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

docs/AlarmClientDiscovery.md

@@ -688,3 +688,105 @@ alarm-consumer surface unblocks A.2 fully. Outline:
 These findings retire the open follow-up probes from the
 "polling-vs-pump" debate above — `wwAlarmConsumerClass` plus
 poll-on-timer is the implementation.
+
+## Live smoke-test discoveries — 2026-05-01
+
+The Skip-gated `AlarmsLiveSmokeTests.Alarms_full_pipeline_round_trip`
+ran the full
+`WnWrapAlarmConsumer` + `AlarmDispatcher` + `MxAccessAlarmEventSink`
+pipeline against the dev rig with the flip script running. End-to-end
+verified: 6 real transitions captured on the 10s cadence, ack-by-name
+returned rc=0, pipeline stayed healthy through 5 more transitions
+afterwards. Three production-relevant quirks surfaced and were fixed
+in the consumer:
+
+### 1. `SetXmlAlarmQuery` is mandatory for reads despite the mangled echo
+
+Without `SetXmlAlarmQuery`, the first `GetXmlCurrentAlarms2` call
+fails with `E_FAIL` (HRESULT `0x80004005`). The discovery doc above
+flagged the round-trip echo as mangled and recommended skipping the
+call — that recommendation is **wrong**. The echo *is* mangled (AVEVA
+parses NODE/PROVIDER/ALARM_STATE/DISPLAY_MODE incorrectly), but the
+call itself is required as some kind of subscription enabler. Even
+the Subscribe call setting the actual filter doesn't avoid the need
+for `SetXmlAlarmQuery`.
+
+`WnWrapAlarmConsumer.ComposeXmlAlarmQuery(subscription)` decomposes
+the canonical `\\<machine>\Galaxy!<area>` form into the XML's
+NODE/PROVIDER/GROUP fields. Mangled or not, the call enables reads.
+
+### 2. Two consumers required: read-side vs. ack-side
+
+`SetXmlAlarmQuery` enables reads but **breaks `AlarmAckByName` on
+the same consumer instance**. With SetXml applied, AlarmAckByName
+returns -55 even with valid name+provider+group+operator. Without
+SetXml, AlarmAckByName succeeds with rc=0.
+
+The production consumer therefore provisions **two** wnwrap COM
+instances:
+- Primary consumer (`client`): runs full lifecycle including
+  `SetXmlAlarmQuery` for `GetXmlCurrentAlarms2` polls.
+- Ack-only consumer (`ackClient`): runs Initialize → Register →
+  Subscribe via the v1-prefixed methods, **no SetXmlAlarmQuery**.
+  All `AcknowledgeByName` calls dispatch through this instance.
+
+Both consumers subscribe to the same expression. Disposal cleans up
+both via a shared `ReleaseConsumerCom` helper.
+
+### 3. `AlarmAckByName` v2 8-arg vs. v1 6-arg
+
+`wwAlarmConsumerClass` exposes two `AlarmAckByName` overloads:
+- `IwwAlarmConsumer2` v2: 8 args (`name, provider, group, comment,
+  oprName, node, domainName, oprFullName`).
+- `IwwAlarmConsumer` v1: 6 args (no domain, no full-name).
+
+The v2 8-arg method returns -55 on this AVEVA build regardless of
+operator-identity inputs — looks like a stub. The v1 6-arg method
+works. Production `WnWrapAlarmConsumer.AcknowledgeByName` calls the
+6-arg overload and discards the proto's `domain` + `full_name` fields.
+The proto contract keeps the 8 fields for forward compatibility if
+AVEVA fixes the v2 method later.
+
+### 4. `AlarmAckByGUID` is not implemented
+
+The v2 `AlarmAckByGUID(VBGUID, …)` throws `NotImplementedException`
+(COM `E_NOTIMPL`) on `wwAlarmConsumerClass` against this AVEVA
+build. The reference→GUID lookup that we initially planned to wire
+through `AlarmAckByGUID` is therefore not viable on wnwrap; all acks
+must go through `AlarmAckByName`.
+
+The proto `AcknowledgeAlarmCommand` (GUID-based) and the worker's
+`MxAccessCommandExecutor.ExecuteAcknowledgeAlarm` switch arm remain
+in the codebase for the forward-compat shape, but the gateway-side
+`WorkerAlarmRpcDispatcher.AcknowledgeAsync` now always routes through
+`AcknowledgeAlarmByName` when the public RPC supplies a recognizable
+`Provider!Group.Tag` reference.
+
+### 5. STA / threading — production fix needed
+
+The wnwrap COM is `ThreadingModel=Apartment`. The consumer's
+internal `Timer` fires on threadpool threads and would block forever
+on cross-apartment marshaling unless the host STA pumps Win32
+messages. The smoke test sidesteps this by setting
+`pollIntervalMilliseconds=0` (Timer disabled) and driving `PollOnce`
+manually from the test's STA. Production hosting will route polls
+through the worker's `StaRuntime` in a follow-up — the consumer's
+`PollOnce` is `public` and idempotent so the wire-up is mechanical.
+
+### Capture summary
+
+```
+Transition: kind=Clear  ref='Galaxy!TestArea.TestMachine_001.TestAlarm001' …
+Transition: kind=Raise  ref='Galaxy!TestArea.TestMachine_001.TestAlarm001' …
+SnapshotActiveAlarms count=1
+  active: ref='Galaxy!TestArea.TestMachine_001.TestAlarm001' state=Active
+AcknowledgeByName(real identity) -> rc=0
+Post-ack transition: kind=Clear …
++1: kind=Raise … (10s after ack)
++2: kind=Clear … (20s)
++3: kind=Raise … (30s)
++4: kind=Clear … (40s)
+```
+
+10s cadence held throughout; full proto fields populated correctly;
+ack registered server-side without errors.

src/MxGateway.Worker.Tests/AlarmsLiveSmokeTests.cs

@@ -0,0 +1,276 @@
+using System;
+using System.Collections.Concurrent;
+using System.Diagnostics;
+using System.Linq;
+using System.Threading;
+using MxGateway.Contracts.Proto;
+using MxGateway.Worker.MxAccess;
+using Xunit.Abstractions;
+
+namespace MxGateway.Worker.Tests;
+
+/// <summary>
+///     Live dev-rig smoke test for the alarms-over-gateway pipeline.
+///     Exercises <see cref="WnWrapAlarmConsumer"/> + <see cref="AlarmDispatcher"/> +
+///     <see cref="MxAccessAlarmEventSink"/> end-to-end against the actual
+///     AVEVA System Platform install: subscribes to
+///     <c>\\&lt;machine&gt;\Galaxy!DEV</c>, waits for at least one alarm
+///     transition (the dev rig's flip script writes
+///     <c>TestMachine_001.TestAlarm001</c> every 10s), drains the proto
+///     <c>OnAlarmTransitionEvent</c> from the queue, then ack-by-name's
+///     it and verifies the ack registers as a subsequent
+///     <see cref="AlarmTransitionKind.Acknowledge"/> transition.
+///
+///     Skip-gated; flip <c>Skip=null</c> on the dev rig with the flip
+///     script running.
+/// </summary>
+public sealed class AlarmsLiveSmokeTests
+{
+    private static readonly string SubscriptionExpression =
+        $@"\\{Environment.MachineName}\Galaxy!DEV";
+    private static readonly TimeSpan PumpDuration = TimeSpan.FromSeconds(45);
+    private static readonly TimeSpan TransitionWaitTimeout = TimeSpan.FromSeconds(20);
+
+    private const string SessionId = "alarms-live-smoke";
+
+    private readonly ITestOutputHelper output;
+    private readonly Stopwatch elapsed = Stopwatch.StartNew();
+    private readonly ConcurrentQueue<string> log = new ConcurrentQueue<string>();
+
+    public AlarmsLiveSmokeTests(ITestOutputHelper output)
+    {
+        this.output = output;
+    }
+
+    [Fact(Skip = "Live dev-rig smoke test — flip Skip=null with AVEVA + the alarm flip script running. Verified working 2026-05-01.")]
+    public void Alarms_full_pipeline_round_trip()
+    {
+        Exception? threadException = null;
+        var done = new ManualResetEventSlim(false);
+        var thread = new Thread(() =>
+        {
+            try { RunSmoke(); }
+            catch (Exception ex) { threadException = ex; }
+            finally { done.Set(); }
+        });
+        thread.IsBackground = false;
+        thread.SetApartmentState(ApartmentState.STA);
+        thread.Start();
+        done.Wait();
+        thread.Join();
+
+        output.WriteLine($"Captured {log.Count} log line(s):");
+        while (log.TryDequeue(out string? line))
+        {
+            output.WriteLine(line);
+        }
+
+        if (threadException != null)
+        {
+            throw threadException;
+        }
+    }
+
+    private void RunSmoke()
+    {
+        Log($"Subscription expression: {SubscriptionExpression}");
+        Log($"Pump duration: {PumpDuration.TotalSeconds:F0}s; transition wait timeout: {TransitionWaitTimeout.TotalSeconds:F0}s");
+
+        MxAccessEventQueue queue = new MxAccessEventQueue();
+        // pollIntervalMs=0 disables the internal Timer; we drive PollOnce
+        // manually from the STA below to avoid threadpool→STA marshaling
+        // (the wnwrap COM is ThreadingModel=Apartment, and this test
+        // doesn't run a Win32 message pump on its STA).
+        WnWrapAlarmConsumer consumer = new WnWrapAlarmConsumer(
+            new WNWRAPCONSUMERLib.wwAlarmConsumerClass(),
+            pollIntervalMilliseconds: 0,
+            maxAlarmsPerFetch: 1024);
+        MxAccessAlarmEventSink sink = new MxAccessAlarmEventSink(queue, new MxAccessEventMapper());
+        using AlarmDispatcher dispatcher = new AlarmDispatcher(consumer, sink, SessionId);
+
+        Log("Constructed consumer + sink + dispatcher.");
+        dispatcher.Subscribe(SubscriptionExpression);
+        Log("Subscribe -> ok. Driving PollOnce manually from this STA...");
+
+        // The wnwrap COM object is ThreadingModel=Apartment. The consumer's
+        // internal Timer would fire on a threadpool thread and deadlock on
+        // cross-apartment marshaling without a Win32 message pump. For the
+        // smoke test we constructed the consumer with pollIntervalMs=0
+        // (Timer disabled) and drive PollOnce manually here on the STA.
+        // Production hosting will route polls through the worker's
+        // StaRuntime in a follow-up PR.
+
+        // 1. Wait for the first transition (any kind), then keep waiting
+        //    for one with kind=Raise so the alarm is currently Active when
+        //    we try to ack. AVEVA rejects acks of cleared alarms with -55,
+        //    so we have to time the ack against the flip script's 10s
+        //    cadence.
+        OnAlarmTransitionEvent? raiseBody = null;
+        DateTime raiseDeadline = DateTime.UtcNow + TimeSpan.FromSeconds(30);
+        while (DateTime.UtcNow < raiseDeadline && raiseBody is null)
+        {
+            WorkerEvent? evt = WaitForTransition(queue, TransitionWaitTimeout, "raise", consumer);
+            if (evt is null) break;
+            OnAlarmTransitionEvent body = evt.Event.OnAlarmTransition;
+            Log("Transition: " + DescribeTransition(body));
+            Assert.Equal(SessionId, evt.Event.SessionId);
+            if (body.TransitionKind == AlarmTransitionKind.Raise)
+            {
+                raiseBody = body;
+            }
+        }
+        Assert.NotNull(raiseBody);
+        Assert.False(string.IsNullOrEmpty(raiseBody!.AlarmFullReference));
+        Assert.Contains("Galaxy", raiseBody.AlarmFullReference);
+
+        // 2. Snapshot the active set + verify the captured alarm is there.
+        var snapshot = dispatcher.SnapshotActiveAlarms();
+        Log($"SnapshotActiveAlarms count={snapshot.Count}");
+        foreach (var s in snapshot)
+        {
+            Log("  active: " + DescribeSnapshot(s));
+        }
+        Assert.NotEmpty(snapshot);
+        Assert.Contains(snapshot, s => s.AlarmFullReference == raiseBody.AlarmFullReference);
+
+        // 3. Ack-by-name using the captured reference. Parse the reference
+        //    via the same convention the gateway dispatcher uses
+        //    (Provider!Group.Tag where the tag may contain dots).
+        Assert.True(TryParseReference(
+            raiseBody.AlarmFullReference,
+            out string provider, out string group, out string alarmName),
+            $"Captured reference '{raiseBody.AlarmFullReference}' did not parse as Provider!Group.Tag.");
+        Log($"Ack target: provider='{provider}' group='{group}' name='{alarmName}'");
+
+        // Try the ack with real Windows identity. AVEVA's AlarmAckByName
+        // may reject synthetic operator strings; using the current process
+        // identity gives the alarm-history a recognizable principal.
+        string realUser = Environment.UserName;
+        string realNode = Environment.MachineName;
+        string realDomain = Environment.UserDomainName ?? string.Empty;
+        Log($"Ack identity: user='{realUser}' node='{realNode}' domain='{realDomain}'");
+
+        int rc = dispatcher.AcknowledgeByName(
+            alarmName: alarmName,
+            providerName: provider,
+            groupName: group,
+            ackComment: "alarms-live-smoke ack",
+            ackOperatorName: realUser,
+            ackOperatorNode: realNode,
+            ackOperatorDomain: realDomain,
+            ackOperatorFullName: realUser);
+        Log($"AcknowledgeByName(real identity) -> rc={rc}");
+
+        Assert.Equal(0, rc);
+
+        // 4. Wait for the post-ack transition. With the alarm flipping every
+        //    10s and the consumer polling every 500ms, the next state
+        //    change should be either kind=Acknowledge (the ack we just
+        //    sent registered as a state delta UnackAlm → AckAlm) or the
+        //    flip script's next Clear (UnackAlm → UnackRtn).
+        WorkerEvent? second = WaitForTransition(queue, TransitionWaitTimeout, "post-ack", consumer);
+        Assert.NotNull(second);
+        OnAlarmTransitionEvent secondBody = second!.Event.OnAlarmTransition;
+        Log("Post-ack transition: " + DescribeTransition(secondBody));
+        Assert.NotEqual(AlarmTransitionKind.Unspecified, secondBody.TransitionKind);
+
+        // 5. Pump a little longer to confirm the consumer keeps reporting
+        //    transitions on the 10s flip cadence.
+        DateTime deadline = DateTime.UtcNow + PumpDuration;
+        int additional = 0;
+        while (DateTime.UtcNow < deadline)
+        {
+            consumer.PollOnce();
+            if (queue.TryDequeue(out WorkerEvent? evt) && evt is not null)
+            {
+                additional++;
+                OnAlarmTransitionEvent body = evt.Event.OnAlarmTransition;
+                Log($"  +{additional}: " + DescribeTransition(body));
+            }
+            Thread.Sleep(500);
+        }
+        Log($"Pump completed; additional transitions captured: {additional}.");
+    }
+
+    private WorkerEvent? WaitForTransition(
+        MxAccessEventQueue queue,
+        TimeSpan timeout,
+        string label,
+        WnWrapAlarmConsumer consumer)
+    {
+        DateTime deadline = DateTime.UtcNow + timeout;
+        int pollCount = 0;
+        while (DateTime.UtcNow < deadline)
+        {
+            try
+            {
+                consumer.PollOnce();
+                pollCount++;
+                if (pollCount == 1) Log("First PollOnce returned without throw.");
+            }
+            catch (Exception ex)
+            {
+                Log($"PollOnce threw on poll #{pollCount + 1}: {ex.GetType().Name}: {ex.Message}");
+                if (ex is System.Runtime.InteropServices.COMException ce)
+                {
+                    Log($"  HResult=0x{(uint)ce.HResult:X8}");
+                }
+                throw;
+            }
+            if (queue.TryDequeue(out WorkerEvent? evt) && evt is not null)
+            {
+                if (evt.Event.Family == MxEventFamily.OnAlarmTransition)
+                {
+                    return evt;
+                }
+                Log($"Skipped non-alarm event (family={evt.Event.Family}) while waiting for {label}.");
+            }
+            Thread.Sleep(500);
+        }
+        Log($"Timed out waiting for {label} transition after {timeout.TotalSeconds:F0}s (poll count={pollCount}).");
+        return null;
+    }
+
+    private static bool TryParseReference(
+        string reference,
+        out string provider,
+        out string group,
+        out string alarmName)
+    {
+        provider = group = alarmName = string.Empty;
+        if (string.IsNullOrWhiteSpace(reference)) return false;
+        int bang = reference.IndexOf('!');
+        if (bang <= 0 || bang == reference.Length - 1) return false;
+        string left = reference.Substring(0, bang);
+        string right = reference.Substring(bang + 1);
+        int dot = right.IndexOf('.');
+        if (dot <= 0 || dot == right.Length - 1) return false;
+        provider = left;
+        group = right.Substring(0, dot);
+        alarmName = right.Substring(dot + 1);
+        return true;
+    }
+
+    private static string DescribeTransition(OnAlarmTransitionEvent body)
+    {
+        return string.Format(
+            "kind={0} ref='{1}' source='{2}' type='{3}' severity={4} operator='{5}' comment='{6}' ts={7:o}",
+            body.TransitionKind, body.AlarmFullReference, body.SourceObjectReference,
+            body.AlarmTypeName, body.Severity, body.OperatorUser, body.OperatorComment,
+            body.TransitionTimestamp?.ToDateTime() ?? DateTime.MinValue);
+    }
+
+    private static string DescribeSnapshot(ActiveAlarmSnapshot s)
+    {
+        return string.Format(
+            "ref='{0}' state={1} severity={2} operator='{3}' comment='{4}' ts={5:o}",
+            s.AlarmFullReference, s.CurrentState, s.Severity, s.OperatorUser,
+            s.OperatorComment,
+            s.LastTransitionTimestamp?.ToDateTime() ?? DateTime.MinValue);
+    }
+
+    private void Log(string line)
+    {
+        log.Enqueue($"[t={elapsed.Elapsed.TotalSeconds:F3}s] {line}");
+    }
+}

src/MxGateway.Worker/MxAccess/WnWrapAlarmConsumer.cs

@@ -57,6 +57,8 @@ public sealed class WnWrapAlarmConsumer : IMxAccessAlarmConsumer
     private readonly int maxAlarmsPerFetch;
 
     private wwAlarmConsumerClass? client;
+    private wwAlarmConsumerClass? ackClient;
+    private string subscriptionExpression = string.Empty;
     private Timer? pollTimer;
     private bool subscribed;
     private bool disposed;
@@ -66,16 +68,23 @@ public sealed class WnWrapAlarmConsumer : IMxAccessAlarmConsumer
     {
     }
 
-    /// <summary>Test seam — inject a pre-created COM client and tune the poll cadence.</summary>
-    internal WnWrapAlarmConsumer(
+    /// <summary>
+    ///     Test seam / explicit construction — inject a pre-created COM
+    ///     client and tune the poll cadence. <c>pollIntervalMilliseconds == 0</c>
+    ///     disables the internal <see cref="Timer"/> entirely; the caller
+    ///     must drive <see cref="PollOnce"/> manually (used by hosts that
+    ///     marshal polls onto a foreign STA, and by live smoke tests that
+    ///     pump from the STA they own).
+    /// </summary>
+    public WnWrapAlarmConsumer(
         wwAlarmConsumerClass client,
         int pollIntervalMilliseconds,
         int maxAlarmsPerFetch)
     {
         this.client = client ?? throw new ArgumentNullException(nameof(client));
-        this.pollIntervalMs = pollIntervalMilliseconds > 0
-            ? pollIntervalMilliseconds
-            : DefaultPollIntervalMilliseconds;
+        this.pollIntervalMs = pollIntervalMilliseconds < 0
+            ? DefaultPollIntervalMilliseconds
+            : pollIntervalMilliseconds;
         this.maxAlarmsPerFetch = maxAlarmsPerFetch > 0
             ? maxAlarmsPerFetch
             : DefaultMaxAlarmsPerFetch;
@@ -104,9 +113,14 @@ public sealed class WnWrapAlarmConsumer : IMxAccessAlarmConsumer
             wwAlarmConsumerClass com = client
                 ?? throw new ObjectDisposedException(nameof(WnWrapAlarmConsumer));
 
-            // Per AlarmClientDiscovery.md: InitializeConsumer MUST precede
-            // RegisterConsumer for the alarm provider chain to become visible.
-            int init = com.InitializeConsumer(DefaultApplicationName);
+            // Use the IwwAlarmConsumer (v1) prefix-named methods for the
+            // lifecycle. Empirically (live dev-rig 2026-05-01) this is the
+            // only path that lets AlarmAckByName succeed afterwards. The
+            // v2 Initialize/Register/Subscribe methods on the class
+            // succeed (return 0) but acks against that consumer state
+            // return -55. The v1 prefix path is what WIN-911-style code
+            // uses against the same wnwrap library.
+            int init = com.IwwAlarmConsumer_InitializeConsumer(DefaultApplicationName);
             if (init != 0)
             {
                 throw new InvalidOperationException(
@@ -115,7 +129,7 @@ public sealed class WnWrapAlarmConsumer : IMxAccessAlarmConsumer
 
             // hWnd=0: wnwrap supports a pull-based model — no message pump
             // is required. We poll GetXmlCurrentAlarms2 on a timer below.
-            int reg = com.RegisterConsumer(
+            int reg = com.IwwAlarmConsumer_RegisterConsumer(
                 hWnd: 0,
                 szProductName: DefaultProductName,
                 szApplicationName: DefaultApplicationName,
@@ -126,7 +140,7 @@ public sealed class WnWrapAlarmConsumer : IMxAccessAlarmConsumer
                     $"wwAlarmConsumer.RegisterConsumer returned non-zero status {reg}.");
             }
 
-            int sub = com.Subscribe(
+            int sub = com.IwwAlarmConsumer_Subscribe(
                 szSubscription: subscription,
                 wFromPri: 1,
                 wToPri: 999,
@@ -140,8 +154,49 @@ public sealed class WnWrapAlarmConsumer : IMxAccessAlarmConsumer
                     $"wwAlarmConsumer.Subscribe('{subscription}') returned non-zero status {sub}.");
             }
 
+            // Empirically required: even though the round-trip echo of
+            // SetXmlAlarmQuery is mangled (see docs/AlarmClientDiscovery.md),
+            // calling it is necessary for subsequent GetXmlCurrentAlarms2
+            // calls to succeed. Without it, GetXmlCurrentAlarms2 returns
+            // E_FAIL (HRESULT 0x80004005) on the first poll. SetXmlAlarmQuery
+            // also breaks AlarmAckByName on the same consumer (rejects with
+            // -55), so a separate ack-only consumer is provisioned below
+            // that gets only Initialize/Register/Subscribe (no SetXmlAlarmQuery).
+            string xmlQuery = ComposeXmlAlarmQuery(subscription);
+            com.SetXmlAlarmQuery(xmlQuery);
+
+            // Provision a parallel COM consumer for ack calls. It runs the
+            // v1 lifecycle (Initialize/Register/Subscribe) only; without
+            // SetXmlAlarmQuery, AlarmAckByName succeeds. State is read-only
+            // — we never poll this consumer.
+            ackClient = new wwAlarmConsumerClass();
+            int ackInit = ackClient.IwwAlarmConsumer_InitializeConsumer(DefaultApplicationName + ".ack");
+            int ackReg = ackClient.IwwAlarmConsumer_RegisterConsumer(
+                hWnd: 0,
+                szProductName: DefaultProductName,
+                szApplicationName: DefaultApplicationName + ".ack",
+                szVersion: DefaultVersion);
+            int ackSub = ackClient.IwwAlarmConsumer_Subscribe(
+                szSubscription: subscription,
+                wFromPri: 1,
+                wToPri: 999,
+                QueryType: eQueryType.qtSummary,
+                SortFlags: eSortFlags.sfReturnNewestFirst,
+                FilterMask: eAlarmFilterState.asAlarmActiveNow,
+                FilterSpecification: eAlarmFilterState.asAlarmActiveNow);
+            if (ackInit != 0 || ackReg != 0 || ackSub != 0)
+            {
+                throw new InvalidOperationException(
+                    $"Ack consumer setup returned non-zero status: " +
+                    $"Initialize={ackInit}, Register={ackReg}, Subscribe={ackSub}.");
+            }
+            subscriptionExpression = subscription;
+
             subscribed = true;
-            pollTimer = new Timer(OnPoll, state: null, dueTime: 0, period: pollIntervalMs);
+            if (pollIntervalMs > 0)
+            {
+                pollTimer = new Timer(OnPoll, state: null, dueTime: 0, period: pollIntervalMs);
+            }
         }
     }
 
@@ -185,18 +240,31 @@ public sealed class WnWrapAlarmConsumer : IMxAccessAlarmConsumer
     {
         if (disposed) throw new ObjectDisposedException(nameof(WnWrapAlarmConsumer));
 
-        wwAlarmConsumerClass com = client
-            ?? throw new ObjectDisposedException(nameof(WnWrapAlarmConsumer));
+        // Use the parallel ack-only consumer (no SetXmlAlarmQuery applied)
+        // — see docs/AlarmClientDiscovery.md "Option A — captured" for the
+        // empirical justification.
+        wwAlarmConsumerClass com = ackClient
+            ?? throw new InvalidOperationException(
+                "Cannot acknowledge: WnWrapAlarmConsumer was disposed or has not been subscribed yet.");
 
+        // Empirically (live dev-rig 2026-05-01): the IwwAlarmConsumer2
+        // 8-arg AlarmAckByName returns -55 on this AVEVA build (looks like
+        // a stub). The legacy 6-arg IwwAlarmConsumer.AlarmAckByName works
+        // and reaches the alarm-history path correctly. Operator-domain
+        // and operator-full-name fields are accepted by the proto contract
+        // for forward-compat but are not propagated to AVEVA today —
+        // wrapped in the 6-arg call so domain/full-name go to the
+        // alarm-history operator-name field via the szOprName parameter.
+        // Suppress unused-warning explicitly:
+        _ = ackOperatorDomain;
+        _ = ackOperatorFullName;
         return com.AlarmAckByName(
             szAlarmName: alarmName ?? string.Empty,
             szProviderName: providerName ?? string.Empty,
             szGroupName: groupName ?? string.Empty,
             szComment: ackComment ?? string.Empty,
             szOprName: ackOperatorName ?? string.Empty,
-            szNode: ackOperatorNode ?? string.Empty,
-            szDomainName: ackOperatorDomain ?? string.Empty,
-            szOprFullName: ackOperatorFullName ?? string.Empty);
+            szNode: ackOperatorNode ?? string.Empty);
     }
 
     /// <inheritdoc />
@@ -236,7 +304,15 @@ public sealed class WnWrapAlarmConsumer : IMxAccessAlarmConsumer
         }
     }
 
-    internal void PollOnce()
+    /// <summary>
+    ///     Synchronously poll the wnwrap consumer once and dispatch any
+    ///     transitions. Public so STA-bound hosts can drive polling from
+    ///     the thread that owns the COM object instead of relying on the
+    ///     internal <see cref="Timer"/> (which fires on a thread-pool
+    ///     thread and blocks indefinitely on cross-apartment marshaling
+    ///     when the host STA isn't pumping messages).
+    /// </summary>
+    public void PollOnce()
     {
         wwAlarmConsumerClass? com;
         lock (syncRoot)
@@ -370,6 +446,58 @@ public sealed class WnWrapAlarmConsumer : IMxAccessAlarmConsumer
         return Guid.TryParse(canonical, out guid);
     }
 
+    /// <summary>
+    ///     Compose the XML payload <c>SetXmlAlarmQuery</c> expects from a
+    ///     canonical subscription expression
+    ///     (<c>\\&lt;machine&gt;\Galaxy!&lt;area&gt;</c>). The wnwrap
+    ///     consumer mangles the round-trip but evidently still needs the
+    ///     call — without it <c>GetXmlCurrentAlarms2</c> fails with
+    ///     E_FAIL. Best-effort parse: if the subscription doesn't decompose
+    ///     cleanly, fall back to a permissive ALL-priority/ALL-state form
+    ///     so the worker doesn't fail to start.
+    /// </summary>
+    internal static string ComposeXmlAlarmQuery(string subscription)
+    {
+        string node = Environment.MachineName;
+        string provider = "Galaxy";
+        string group = string.Empty;
+
+        if (!string.IsNullOrEmpty(subscription))
+        {
+            // Strip leading backslashes from "\\<node>\..." form.
+            string trimmed = subscription.TrimStart('\\');
+            int slash = trimmed.IndexOf('\\');
+            if (slash > 0)
+            {
+                node = trimmed.Substring(0, slash);
+                trimmed = trimmed.Substring(slash + 1);
+            }
+            int bang = trimmed.IndexOf('!');
+            if (bang > 0)
+            {
+                provider = trimmed.Substring(0, bang);
+                group = trimmed.Substring(bang + 1);
+            }
+            else
+            {
+                provider = trimmed;
+            }
+        }
+
+        System.Text.StringBuilder sb = new System.Text.StringBuilder();
+        sb.Append("<QUERIES FROM_PRIORITY=\"1\" TO_PRIORITY=\"999\" ALARM_STATE=\"ALL\" DISPLAY_MODE=\"Summary\">");
+        sb.Append("<QUERY>");
+        sb.Append("<NODE>").Append(node).Append("</NODE>");
+        sb.Append("<PROVIDER>").Append(provider).Append("</PROVIDER>");
+        if (!string.IsNullOrEmpty(group))
+        {
+            sb.Append("<GROUP>").Append(group).Append("</GROUP>");
+        }
+        sb.Append("</QUERY>");
+        sb.Append("</QUERIES>");
+        return sb.ToString();
+    }
+
     private static VBGUID ToVbGuid(Guid g)
     {
         byte[] bytes = g.ToByteArray();
@@ -390,6 +518,7 @@ public sealed class WnWrapAlarmConsumer : IMxAccessAlarmConsumer
     {
         Timer? timerToDispose;
         wwAlarmConsumerClass? clientToDispose;
+        wwAlarmConsumerClass? ackClientToDispose;
         lock (syncRoot)
         {
             if (disposed) return;
@@ -398,16 +527,22 @@ public sealed class WnWrapAlarmConsumer : IMxAccessAlarmConsumer
             pollTimer = null;
             clientToDispose = client;
             client = null;
+            ackClientToDispose = ackClient;
+            ackClient = null;
         }
         timerToDispose?.Dispose();
-        if (clientToDispose is not null)
+        ReleaseConsumerCom(clientToDispose);
+        ReleaseConsumerCom(ackClientToDispose);
+    }
+
+    private static void ReleaseConsumerCom(wwAlarmConsumerClass? consumer)
+    {
+        if (consumer is null) return;
+        try { consumer.DeregisterConsumer(); } catch { /* swallow */ }
+        try { consumer.UninitializeConsumer(); } catch { /* swallow */ }
+        if (Marshal.IsComObject(consumer))
         {
-            try { clientToDispose.DeregisterConsumer(); } catch { /* swallow */ }
-            try { clientToDispose.UninitializeConsumer(); } catch { /* swallow */ }
-            if (Marshal.IsComObject(clientToDispose))
-            {
-                try { Marshal.FinalReleaseComObject(clientToDispose); } catch { /* swallow */ }
-            }
+            try { Marshal.FinalReleaseComObject(consumer); } catch { /* swallow */ }
         }
     }
 }