Phase 2 Task 3. Central-side end of the boundary: routes every central->node
command out over DPS or ClusterClient per MeshTransport:Mode, and forwards
inbound ApplyAcks to the deploy coordinator singleton (Forward, not Tell, so
the coordinator sees the node rather than this relay).
Three things here are load-bearing and easy to get wrong later:
SendToAll, never Send. Today's DPS publish reaches EVERY DriverHostActor and
the node side has no ClusterId or node filter to compensate -- scoping happens
later, inside the artifact. ClusterClient.Send delivers to exactly ONE
registered actor, so it would deploy to a single node while every other node
silently kept its old config, and the deployment could still seal green.
Sabotage-verified: swapping to Send reddens exactly that one test.
Exactly ONE ClusterClient, fleet-wide. A receptionist serves its whole cluster,
so SendToAll reaches every registered node-comm actor in the mesh regardless of
which contact point was dialled. One client per application Cluster -- the
shape Phase 6 wants -- would fan each command out once per cluster while the
fleet is still one mesh: N x duplicate DispatchDeployment and ApplyAck. Marked
TODO(Phase 6).
The contact set does NOT scope delivery. Excluding a maintenance-mode node from
the contacts does not stop it receiving commands; it still receives them, as it
does under today's broadcast. The filter is about which receptionists are worth
dialling, not about who gets the message. Stated in the code because the
opposite is the natural assumption.
Also carries the sister project's shipped fixes: per-row address parsing so one
malformed ClusterNode cannot abort the refresh and leave the contact set
half-built (regression test orders the bad row FIRST), the cache entry cleared
before recreate so a failed create cannot leave commands routing into a
stopping actor, and a Status.Failure handler so a DB outage is a Warning rather
than a silent debug-level unhandled message.
Two test-harness bugs found and fixed while writing this, both mine: SubscribeAck
goes to the SENDER (so the mediator subscribe needed an explicit sender), and
EventFilter matches case-INSENSITIVELY, so a "was NOT" filter also caught this
actor's own "was not created" warning.
Claude-Session: https://claude.ai/code/session_01GASWkNEi68FSCtvr6rLoEW
Phase 2 Task 2. Two small pieces the comm actors are built on.
MeshCommand carries a command plus the DPS topic it would have been published
on, so the admin singletons stop knowing which transport is in force. Without
it the dark switch would have to be threaded through five publish sites across
two actors, each free to drift. Central-side only -- it never crosses the wire,
so it carries no serialization contract.
DefaultMeshClusterClientFactory appends a generation counter to the actor name.
Context.Stop is asynchronous and the name stays reserved until termination
completes, so recreating the same name inside one message handler throws
InvalidActorNameException -- which is exactly what a contact-set change does
(stop the old client, create the replacement, one handler). Ported from the
sister project, where it was a shipped bug fix rather than foresight.
Sabotage-verified: freezing the name to a constant reddens both name tests with
InvalidActorNameException.
Dropped a third test I had written -- it asserted ClusterClientSettings
directly and never touched the factory, so it was coverage theatre. Replaced
with a comment saying where contact propagation IS covered (the Task 8 boundary
test, the only place it can actually fail).
Claude-Session: https://claude.ai/code/session_01GASWkNEi68FSCtvr6rLoEW
RedundancyStateActor derived the driver Primary from ClusterState.RoleLeader("driver").
Akka offers two different notions of a "first" member and they are not the same
one: role leader is the lowest-ADDRESSED Up member (host, then port), while
ClusterSingletonManager places singletons on the OLDEST — lowest up-number.
They agree on a freshly-formed cluster, which is why every existing test passed.
They diverge after any restart: the restarted node re-joins as the youngest while
keeping its address, so if it holds the lower address it becomes role leader while
the singletons stay put. The snapshot would then name a Primary that is not
hosting the work, and every Primary-gated surface follows it — inbound device
writes, native-alarm acks, the fleet-wide alerts emit, and the alarm-history
drain would all enable on the wrong node while the node actually running the
singletons stayed gated off.
BuildSnapshot now selects the oldest Up member carrying the driver role, matching
singleton placement. Leaving members are excluded: a node handing its singletons
over must not be named Primary.
NodeRedundancyState.IsRoleLeaderForDriver is renamed IsDriverPrimary, and
NodeHealthInputs.IsDriverRoleLeader likewise. Keeping the old names would have
left the wire contract asserting a derivation the code no longer uses — the same
drift that made this defect invisible.
Proven by a real two-node cluster rather than a mock. RedundancyPrimaryElectionTests
binds the first-joining node to the HIGHER port, so oldest and lowest-address name
different nodes, and includes a fixture assertion that the divergence actually
occurred — without it the real assertion could pass for the wrong reason. Positive
control: restoring the RoleLeader derivation turns exactly the two election tests
red while the fixture check stays green.
Runtime.Tests 440 passed, ControlPlane.Tests 82, Cluster.Tests 36.
Claude-Session: https://claude.ai/code/session_01GASWkNEi68FSCtvr6rLoEW
Materialize each native alarm ONCE at the raw tag (ConditionId = RawPath, Raw
realm); wire the single condition as an SDK event notifier of each referencing
equipment's UNS folder so one ReportEvent fans to every root without
re-reporting per root (which would break Server-object dedup + Part 9 ack
correlation).
- New sink method WireAlarmNotifiers(alarmNodeId, alarmRealm,
notifierFolderNodeIds, notifierFolderRealm) on IOpcUaAddressSpaceSink,
forwarded through DeferredAddressSpaceSink + SdkAddressSpaceSink +
NullOpcUaAddressSpaceSink (the forwarding-trap guard); auto-covered by the
DeferredSinkForwardingReflectionTests realm + forwarding guards + a
hand-written forward test.
- OtOpcUaNodeManager: the normative AddNotifier(isInverse) pair + idempotent
EnsureFolderIsEventNotifier per equipment folder; tracked per condition in
_alarmNotifierWiring. Teardown symmetry: RemoveNotifier(bidirectional:true) on
RebuildAddressSpace, RemoveAlarmConditionNode, and RemoveEquipmentSubtree so no
inverse-notifier entry leaks across redeploys.
- AddressSpaceApplier.MaterialiseRawSubtree wires notifiers for each native alarm
tag, resolving its ReferencingEquipmentPaths (Area/Line/Equipment) to the
EquipmentId folder NodeIds via BuildEquipmentIdByFolderPath.
- AlarmTransitionEvent gains ReferencingEquipmentPaths (empty default); /alerts
renders the referencing-equipment list as display metadata.
- Un-skipped + rewrote the native-alarm dark tests
(DriverHostActorNativeAlarmTests x6, DriverHostActorNativeAlarmAckRoutingTests
x1) for the v3 raw-condition model; new NodeManagerMultiNotifierAlarmTests
proves multi-notifier wiring + teardown symmetry (no leaked duplicates after a
re-trip) + applier wiring test.
Claude-Session: https://claude.ai/code/session_01LVneM3eh1UtJxEisFXgmox
T21: add an AdminUI path for acknowledging/shelving alarms that routes
through the admin-pinned AdminOperationsActor cluster singleton, which
republishes onto the same 'alarm-commands' DPS topic the OPC UA method
path (T18) and the engine subscriber (T19) use. The broadcast + the
ScriptedAlarmHostActor ownership filter handle cross-node routing, so
the singleton needs no knowledge of which node owns the alarm.
- Commons: AcknowledgeAlarmCommand/ShelveAlarmCommand (+ result records)
and a shared AlarmCommandsTopic const; ScriptedAlarmHostActor now
re-exports that const (mirrors the DriverControlTopic pattern).
- AdminOperationsActor: two handlers map the control-plane messages to
AlarmCommand (Acknowledge / OneShotShelve / TimedShelve / Unshelve,
threading User/Comment/UnshelveAtUtc) and publish via the DPS mediator.
- IAdminOperationsClient + AdminOperationsClient: typed Acknowledge/Shelve
ask wrappers mirroring StartDeploymentAsync.
- Alerts.razor: per-row DriverOperator-gated Ack/Shelve/Unshelve controls;
operator name from AuthenticationState. Timed-shelve datetime UI deferred.
- 5 TestKit tests (mediator-probe subscribed to alarm-commands) verifying
each kind's mapping + reply; 56/56 ControlPlane tests green.
Deep-adopt the shared audit record. Deletes the bespoke 8-field positional
Commons AuditEvent and repoints the writer path at ZB.MOM.WW.Audit.AuditEvent
(0.1.0, feed-mapped via dohertj2-gitea). Adds the package reference to both
Commons and ControlPlane.
- AuditWriterActor now implements IAuditWriter: WriteAsync(evt, ct) is a
best-effort, never-throwing entry point that Self.Tell()s the event onto the
same batching/dedup/flush pipeline and returns Task.CompletedTask. Existing
Receive<AuditEvent> + 500/5s batching + two-layer dedup unchanged.
- Flush mapping updated for the canonical field types: OccurredAtUtc is now
DateTimeOffset (.UtcDateTime into the datetime2 column), SourceNode is string?
(was NodeId.Value), CorrelationId is Guid? (stored null when null). Outcome is
NOT yet persisted (column lands in Task 2.2).
- New AuditOutcomeMapper.FromAction maps the OtOpcUa action vocabulary to the
required canonical Outcome: OpcUaAccessDenied / CrossClusterNamespaceAttempt ->
Denied; config verbs (DraftCreated/Edited, Published, RolledBack, NodeApplied,
ClusterCreated, NodeAdded, CredentialAdded/Disabled, ExternalIdReleased) ->
Success. OtOpcUa emits no Failure events.
The Akka message shape changed, but the structured audit path is dormant (zero
production emit/Tell sites; all live audit flows through the bespoke SP path),
so there is no rolling-deploy wire-compat concern. Tested-not-exercised by
design.
ControlPlane.Tests: 44/44 green (AuditWriterActor suite rewritten to construct
the canonical record + assert the Outcome derivation table + the WriteAsync
best-effort/mailbox-routing contract + null SourceNode/CorrelationId handling).
- Topic-name drift fix: DriverHealthChanged.TopicName and
DriverControlTopic.Name now live on the message contracts in
Commons. AkkaDriverHealthPublisher, DriverStatusSignalRBridge,
DriverHostActor, and AdminOperationsActor all delegate to the
single constant so a rename can't silently desynchronise
publisher and subscriber.
- DriverStatusPanel._opResultClearTimer switched from
System.Timers.Timer to System.Threading.Timer + awaited
DisposeAsync. Prevents an in-flight 8s clear-callback from
invoking StateHasChanged on a component whose hub has already
been released.
- PublishHealthSnapshot deduplicates against the last published
(state, lastSuccess, lastError, errorCount) fingerprint. The
30s heartbeat no longer floods the SignalR layer with identical
Healthy snapshots — newly-joined clients still warm up via the
snapshot store on JoinDriver.
- RestartDriver / ReconnectDriver messages + AdminOperationsActor
handlers (broadcast via driver-control DPS topic; audited via
ConfigEdits).
- DriverHostActor subscribes to driver-control; locates the
matching child DriverInstanceActor and stops+respawns it
(Restart) or sends it a ForceReconnect internal message
(Reconnect — re-enters Reconnecting state without full stop).
DriverInstanceSpec constructor call uses named args to handle
the full 6-parameter signature.
- New DriverOperator authorization policy mapped to DriverOperator
or FleetAdmin role; documented in docs/security.md. Map LDAP
group via GroupToRole (e.g. "ot-driver-operator": "DriverOperator").
- DriverStatusPanel renders Reconnect + Restart buttons when the
user holds the DriverOperator policy (hidden otherwise). Restart
requires an in-page Razor confirm block (no JS confirm, keeps
SignalR event loop unblocked). Both buttons show a spinner and
are disabled during in-flight; result chip auto-clears after 8s.
Username sourced from AuthenticationStateProvider.
Reconnect resolves to "ForceReconnect" (re-enter Reconnecting,
not full stop+respawn) — transport drops and retries while actor
and in-memory state are preserved. All DriverInstanceActor states
handle ForceReconnect safely (no-op when already in transition).
- IDriverProbe abstraction in Core.Abstractions; one impl per driver
type, resolved by DriverType string. Phase 7.3 + 7.4 add concrete
probes for the 9 supported driver types.
- TestDriverConnect / TestDriverConnectResult messages.
- AdminOperationsActor.HandleTestDriverConnectAsync looks up the probe
by DriverType, runs it with a [1,60]s clamped timeout, and returns
success/latency or failure/message. Probes that throw or time out
surface as soft failures.
Final F15 batch wires up the SignalR-backed live pages, ports the bulk
equipment importer, and progressively enhances the Script source editor
with Monaco.
Message contracts:
- Commons.Messages.Alerts.AlarmTransitionEvent — fires on every alarm
state transition; published on the `alerts` DPS topic by future
ScriptedAlarmActor (F9) emits.
- Commons.Messages.Logging.ScriptLogEntry — one log line emitted by a
hosted script; published on the `script-logs` DPS topic by future
VirtualTagActor (F8) + ScriptedAlarmActor (F9) emits.
(Folder named "Logging" to dodge .gitignore's "logs/" rule.)
SignalR plumbing:
- AlertHub gains MethodName + bridge actor (AlertSignalRBridge)
- ScriptLogHub introduced; ScriptLogSignalRBridge follows the same
DPS-subscribe → IHubContext fan-out pattern as FleetStatusSignalRBridge
- WithOtOpcUaSignalRBridges now spawns all three bridges
- MapOtOpcUaHubs maps /hubs/script-log alongside the existing hubs
Pages:
- /alerts live alarm tail, 200-row capacity
- /script-log live script-log tail with level + script
filter, 500-row capacity
- /clusters/{id}/equipment/import — CSV bulk Equipment add with preview
(Name/MachineCode/UnsLineId/Driver +
optional ZTag/SAPID/Manufacturer/Model;
skips rows whose MachineCode already
exists in the fleet)
- ScriptEdit progressively enhanced with Monaco editor via JSInterop —
the textarea remains Blazor's source of truth and Monaco syncs into it
on every keystroke so @bind keeps working; falls back gracefully if
the CDN is unreachable.
MainLayout nav gains a "Live" section (Deployments, Alerts, Alarms
historian) and a "Scripts" link under Scripting. ClusterEquipment
surfaces the new Import CSV button.
Tally: F15 ships ~42 razor pages + 3 SignalR hubs + 3 bridge actors.
Microsoft.AspNetCore.SignalR.Client added (was already in central PM).
All 104 v2 tests remain green.
- New Commons.Messages.Fleet.GetDiagnostics request record.
- DriverHostActor handles GetDiagnostics in all three states (Steady, Applying,
Stale); replies with a NodeDiagnosticsSnapshot built from _currentRevision
+ the local NodeId. Drivers list is empty until F7 wires the per-instance
children.
- FleetDiagnosticsClient now resolves the target via ActorSelection at
akka.tcp://{system}@{nodeId}/user/driver-host and Asks with a 3s timeout.
On timeout/peer-down it returns an empty snapshot so the UI degrades
gracefully rather than throwing.
Two new integration tests in Host.IntegrationTests:
- GetDiagnostics_returns_snapshot_with_target_NodeId verifies the
cross-node Ask/Reply works.
- GetDiagnostics_after_deploy_reports_current_revision exercises the
end-to-end path: AdminOps starts a deployment, both DriverHostActors
apply, then diagnostics reports the new revision on both nodes.
All 98 v2 tests pass (was 96 + 2 new).
Adds the empty project skeletons that subsequent v2 tasks fill in:
src/Core/ZB.MOM.WW.OtOpcUa.Commons (types, interfaces, message contracts)
src/Core/ZB.MOM.WW.OtOpcUa.Cluster (Akka.Hosting + cluster wiring)
src/Server/ZB.MOM.WW.OtOpcUa.Security (cookie+JWT auth, LDAP)
src/Server/ZB.MOM.WW.OtOpcUa.ControlPlane (admin-role cluster singletons)
src/Server/ZB.MOM.WW.OtOpcUa.Runtime (per-node driver actors)
src/Server/ZB.MOM.WW.OtOpcUa.OpcUaServer (OPC UA SDK application host)
src/Server/ZB.MOM.WW.OtOpcUa.AdminUI (Razor class library)
src/Server/ZB.MOM.WW.OtOpcUa.Host (single fused web binary)
Each project sets TreatWarningsAsErrors=true in its own csproj (per the
Directory.Build.props deviation note in the previous commit). NuGetAuditSuppress
entries cover transitive vulnerability advisories the new strictness surfaces:
- GHSA-g94r-2vxg-569j (OpenTelemetry.Api 1.9.0 via Akka.Cluster.Hosting/Tools)
- GHSA-h958-fxgg-g7w3 (Opc.Ua.Core 1.5.374.126 via OpcUaServer)
- GHSA-37gx-xxp4-5rgx + GHSA-w3x6-4m5h-cxqf (legacy advisories already accepted)
OpcUaServer pins OPCFoundation.NetStandard.Opc.Ua.Configuration to 1.5.374.126
via VersionOverride to match Opc.Ua.Server's transitive Opc.Ua.Core (same
constraint as the legacy Server project).
Runtime does NOT project-reference any concrete Driver.* assemblies; drivers
load reflectively at runtime (Phase 6). Runtime gets the IDriver contract
through Core.Abstractions instead.
Host's Microsoft.Extensions.Hosting.WindowsServices is conditional on the
Windows OS so the project builds on macOS dev machines.
Build verification: dotnet build -> 438 warnings (all pre-existing xUnit1051
in legacy Server.Tests/Admin.Tests), 0 errors. Closes Task 9 (build green
smoke check, no separate commit).
Joseph Doherty