dohertj2/lmxopcua
1
0
Fork 0
Code Issues 2 Pull Requests 1 Actions Packages Projects Releases Wiki Activity

feat(alarms): Primary-gated AlarmTransitionEvent fan-out for native alarms (Phase B WS-5)

Browse Source
This commit is contained in:
Joseph Doherty
2026-06-14 03:48:41 -04:00
parent b50ef9fc2d
commit 8736fcc37c
2 changed files with 177 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/Server/ZB.MOM.WW.OtOpcUa.Runtime/Drivers/DriverHostActor.cs
+62 -5
View File
@@ -8,6 +8,7 @@ using Microsoft.Extensions.Logging.Abstractions;
using ZB.MOM.WW.OtOpcUa.Commons.Engines;
using ZB.MOM.WW.OtOpcUa.Commons.Interfaces;
using ZB.MOM.WW.OtOpcUa.Commons.Messages.Admin;
using ZB.MOM.WW.OtOpcUa.Commons.Messages.Alerts;
using ZB.MOM.WW.OtOpcUa.Commons.Messages.Deploy;
using ZB.MOM.WW.OtOpcUa.Commons.Messages.Fleet;
using ZB.MOM.WW.OtOpcUa.Commons.Messages.Redundancy;
@@ -116,6 +117,11 @@ public sealed class DriverHostActor : ReceiveActor, IWithTimers
/// value variables, so they are kept OUT of the value maps + value-subscription set.</summary>
private readonly Dictionary<(string DriverInstanceId, string FullName), HashSet<string>> _alarmNodeIdByDriverRef = new();
/// <summary>Condition NodeId → (EquipmentId, tag Name, OPC UA alarm type) for building the
/// AlarmTransitionEvent fan-out. Built in the same PushDesiredSubscriptions alarm branch.</summary>
private readonly Dictionary<string, (string EquipmentId, string Name, string AlarmType)> _alarmMetaByNodeId =
new(StringComparer.Ordinal);
/// <summary>Derives a full Part 9 condition snapshot from each native alarm transition delta,
/// tracking per-condition-NodeId prior state. <see cref="Clear"/>'d on every apply alongside the
/// value maps so stale condition state never leaks across redeploys.</summary>
@@ -131,6 +137,12 @@ public sealed class DriverHostActor : ReceiveActor, IWithTimers
/// </summary>
private RedundancyRole? _localRole;
/// <summary>Cached cluster DistributedPubSub mediator, resolved once in <see cref="PreStart"/> (on the
/// actor thread) and reused for the Primary-gated native-alarm <c>alerts</c> fan-out in
/// <see cref="ForwardNativeAlarm"/> instead of re-resolving it per-publish. Mirrors
/// <c>ScriptedAlarmHostActor._mediator</c>.</summary>
private IActorRef _mediator = null!;
/// <summary>
/// Routes an inbound operator write (resolved from the OPC UA node-manager side, Task 11) to the
/// owning driver child. <see cref="HandleRouteNodeWrite"/> gates on the local node being the
@@ -265,14 +277,17 @@ public sealed class DriverHostActor : ReceiveActor, IWithTimers
/// <inheritdoc />
protected override void PreStart()
{
// Resolve the cluster DPS mediator once, on the actor thread, and reuse it for every subscribe +
// the Primary-gated native-alarm alerts fan-out in ForwardNativeAlarm (mirrors ScriptedAlarmHostActor).
_mediator = DistributedPubSub.Get(Context.System).Mediator;
// Subscribe to deployments topic so the coordinator's broadcast lands here.
DistributedPubSub.Get(Context.System).Mediator.Tell(new Subscribe(DeploymentsTopic, Self));
_mediator.Tell(new Subscribe(DeploymentsTopic, Self));
// Subscribe to driver-control topic so AdminUI Reconnect/Restart commands land here.
DistributedPubSub.Get(Context.System).Mediator.Tell(new Subscribe(DriverControlTopic, Self));
_mediator.Tell(new Subscribe(DriverControlTopic, Self));
// Subscribe to the redundancy-state topic so cluster role changes cache this node's role — the
// inbound-write gate in HandleRouteNodeWrite reuses the SAME signal the scripted-alarm emit gate
// uses so only the Primary services operator writes (the secondary keeps state warm for failover).
DistributedPubSub.Get(Context.System).Mediator.Tell(
_mediator.Tell(
new Subscribe(ZB.MOM.WW.OtOpcUa.Runtime.OpcUa.OpcUaPublishActor.RedundancyStateTopic, Self));
// Spawn the VirtualTag host BEFORE Bootstrap so the bootstrap-restore path (which routes
// through PushDesiredSubscriptions and Tells ApplyVirtualTags) has a live host to target.
@@ -489,8 +504,16 @@ public sealed class DriverHostActor : ReceiveActor, IWithTimers
/// For each node the <see cref="_nativeAlarmProjector"/> projects the transition delta into a full
/// <c>AlarmConditionSnapshot</c>, then this Tells <see cref="ZB.MOM.WW.OtOpcUa.Runtime.OpcUa.OpcUaPublishActor.AlarmStateUpdate"/>
/// — the SAME message scripted alarms use, so it routes through <c>WriteAlarmCondition</c>. An
/// unknown ref is Debug-logged and dropped (mirrors the value drop). The <c>/alerts</c> fan-out is a
/// separate concern (Task 7) and is NOT emitted here.
/// unknown ref is Debug-logged and dropped (mirrors the value drop).
///
/// <para>
/// Each transition is ALSO published as an <see cref="AlarmTransitionEvent"/> to the cluster
/// <c>alerts</c> topic — the single historization + live <c>/alerts</c> fan-out path, exactly as
/// scripted alarms do (<c>ScriptedAlarmHostActor.OnEngineEmission</c>). The OPC UA condition
/// write above stays UNGATED (a Secondary keeps its address space warm for failover), but the
/// cluster-wide alerts publish is Primary-gated on the same <see cref="_localRole"/> redundancy
/// signal the inbound-write gate uses — only the Primary publishes the single fleet-wide copy.
/// </para>
/// </summary>
private void ForwardNativeAlarm(DriverInstanceActor.AttributeAlarmPublished msg)
{
@@ -506,6 +529,33 @@ public sealed class DriverHostActor : ReceiveActor, IWithTimers
var snapshot = _nativeAlarmProjector.Project(nodeId, msg.Args);
_opcUaPublishActor.Tell(new ZB.MOM.WW.OtOpcUa.Runtime.OpcUa.OpcUaPublishActor.AlarmStateUpdate(
nodeId, snapshot, msg.Args.SourceTimestampUtc));
// Warm-standby dedup: the OPC UA condition write above is UNGATED (the secondary keeps its
// address space warm), but only the Primary publishes the cluster-wide alerts transition.
// Default-emit until told we are Secondary/Detached so single-node deploys + the boot window
// never drop transitions — the SAME signal HandleRouteNodeWrite gates writes on.
if (_localRole is RedundancyRole.Secondary or RedundancyRole.Detached) continue;
var meta = _alarmMetaByNodeId.TryGetValue(nodeId, out var m)
? m : (EquipmentId: nodeId, Name: nodeId, AlarmType: "AlarmCondition");
_mediator.Tell(new Publish(ScriptedAlarmHostActor.AlertsTopic, new AlarmTransitionEvent(
AlarmId: nodeId,
EquipmentPath: meta.EquipmentId,
AlarmName: meta.Name,
TransitionKind: msg.Args.Kind.ToString(),
// The projector mapped the four-bucket AlarmSeverity onto the OPC UA 1..1000 scale already;
// reuse its ushort so the condition node + the alerts row agree on severity.
Severity: snapshot.Severity,
Message: msg.Args.Message,
// Native transitions are device-driven; an operator comment only rides along on an
// acknowledge from the upstream alarm system. "device" marks a non-operator origin.
User: msg.Args.OperatorComment is null ? string.Empty : "device",
TimestampUtc: msg.Args.SourceTimestampUtc,
AlarmTypeName: meta.AlarmType,
Comment: msg.Args.OperatorComment,
// Native alarms always historize (no per-condition opt-out surface yet — that is a
// scripted-alarm plan flag); pass a concrete true so the historian's null-default isn't relied on.
HistorizeToAveva: true)));
}
}
@@ -805,6 +855,9 @@ public sealed class DriverHostActor : ReceiveActor, IWithTimers
// every apply; the projector is Clear()'d too so stale per-condition state never leaks across
// redeploys (renames/removals/address-space rebuilds).
_alarmNodeIdByDriverRef.Clear();
// Per-condition metadata (EquipmentId / Name / OPC UA alarm type) for the alerts fan-out, built in
// the SAME alarm branch as the node map so a redeploy can't leave it out of sync. Cleared alongside it.
_alarmMetaByNodeId.Clear();
_nativeAlarmProjector.Clear();
foreach (var t in composition.EquipmentTags)
{
@@ -818,6 +871,10 @@ public sealed class DriverHostActor : ReceiveActor, IWithTimers
if (!_alarmNodeIdByDriverRef.TryGetValue(key, out var aset))
_alarmNodeIdByDriverRef[key] = aset = new HashSet<string>(StringComparer.Ordinal);
aset.Add(nodeId);
// Capture the per-condition metadata the alerts fan-out (ForwardNativeAlarm) needs to build
// the AlarmTransitionEvent: the equipment path, the operator-visible alarm name, and the
// OPC UA Part 9 subtype. Keyed by the condition NodeId (the projection's own key).
_alarmMetaByNodeId[nodeId] = (t.EquipmentId, t.Name, t.Alarm.AlarmType);
continue;
}
if (!_nodeIdByDriverRef.TryGetValue(key, out var set))