Eliminates the services.BuildServiceProvider() captive-provider antipattern
(ASP0000) inside AddJwtBearer. The new ConfigureJwtBearerFromTokenService
resolves JwtTokenService from the real DI container at runtime and stays
in lock-step with JwtTokenService.BuildValidationParameters.
All 27 Security.Tests stay green, including the F1 integration tests that
exercise /auth/token through the real bearer pipeline.
Injects AuthenticationStateProvider and reads the current user's identity
name on Deploy click, replacing the "(current user)" placeholder.
Anonymous case falls back to "(anonymous)" — should never hit in practice
since the page requires FleetAdmin/ConfigEditor.
Mirrors the publisher-injection pattern from FleetStatusBroadcaster and
PeerOpcUaProbeActor: Props accepts an optional Action<object> override so
tests can use a TestProbe sink instead of bootstrapping DistributedPubSub
(unreliable single-node in TestKit).
Un-skips the two RedundancyStateActor tests deferred under F6.
Mirrors WithOtOpcUaControlPlaneSingletons for the driver role. Spawns
DriverHostActor + DbHealthProbeActor on the host's ActorSystem and
registers both under marker keys. Host's Program.cs now calls it when
the node carries the driver role, so driver-only and admin+driver
deployments both auto-bootstrap the per-node actors.
Integration test covers the registration round-trip via Microsoft.Extensions.Hosting
+ Akka.Hosting AddAkka.
Task 56: removes the legacy in-process Server + Admin Web project + their test
projects (Server.Tests, Admin.Tests, Admin.E2ETests). The fused OtOpcUa.Host
binary built across Phases 1-9 is now the sole production entry point.
What happened to the 47 legacy Admin Blazor pages: per follow-up F15, the
v1 architecture's draft/publish UX is replaced by v2's live-edit + snapshot-
deploy model, so a 1:1 migration is not meaningful. The mechanical move via
git mv preserves the history; service classes + page bodies that referenced
removed v1 types (ConfigGeneration, RedundancyRole, GenerationId) were
deleted. AdminUI now ships a minimal Home page + the v2 Deployments page.
Per-page rebuild against the v2 surface is tracked as F15. The v2 Deployments
page (Task 52) is the only first-party UI shipping in this PR.
Task 57: solution build green; 84+ tests green across active v2 + legacy
driver test projects.
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).
Adds Directory.Packages.props (ManagePackageVersionsCentrally) and
Directory.Build.props (net10.0/nullable/implicit usings/LangVersion latest).
Strips Version attributes from every csproj PackageReference and consolidates
versions into the central file.
Side fixes (necessary to keep the build green on .NET SDK 10.0.105 on macOS):
- Microsoft.CodeAnalysis.CSharp{,.Workspaces}: 5.3.0 -> 5.0.0. The 5.3.0
analyzer DLL references compiler 5.3.0.0 and the local SDK ships compiler
5.0.0.0, producing CS9057 on every project that loaded the Analyzers
output. Master itself was broken on this machine pre-change.
- Server + Server.Tests pin OPCFoundation.NetStandard.Opc.Ua.{Configuration,
Client} to 1.5.374.126 via VersionOverride, matching Opc.Ua.Server's
pin. Mixing 1.5.378.106 Opc.Ua.Core transitively with 1.5.374.126
Opc.Ua.Server breaks CustomNodeManager2 override signatures
(CS0115 on LoadPredefinedNodes/Browse/HistoryRead*) and CS7069 in
the tests. The pin disappears when the legacy Server project is
deleted in Task 56.
- Client.UI + Client.UI.Tests: NuGetAuditSuppress for
GHSA-xrw6-gwf8-vvr9 (Tmds.DBus.Protocol 0.20.0 reaches both projects
transitively from Avalonia.Desktop on Linux/macOS only).
Deviation from the plan: TreatWarningsAsErrors=true is NOT set in
Directory.Build.props because the pre-v2 Admin/Server test projects carry
~240 xUnit1051 analyzer warnings that would fail the build. New v2 projects
opt in via their own csproj; the global flag can return once the legacy
projects are deleted in Task 56.
- Server-004: pass the role-derived display name to UserIdentity's base
ctor (the SDK's DisplayName has no public setter) and drop the dead
Display property; make RoleBasedIdentity internal sealed.
- Server-006: derive a bounded CancellationToken from the SDK's
OperationContext.OperationDeadline in OnReadValue / OnWriteValue so a
stalled driver call can no longer pin the request thread.
- Server-008: mark handled slots via CallMethodRequest.Processed = true
in RouteScriptedAlarmMethodCalls (the SDK skips on Processed, not on a
Good error slot).
- Server-012: PeerHttpProbeLoop.ProbeAsync stops mutating client.Timeout
per call; uses a per-request CancellationTokenSource linked to the
shutdown token instead.
- Server-014: wire SealedBootstrap into Program.cs via AddSealedBootstrap
+ OpcUaServerService so the generation-sealed cache + stale-config flag
+ resilient reader actually run; /healthz now reflects cache-fallback
state.
- Server-015: replace the stale 'PR 16 / PR 17 minimum-viable scope'
class summaries on OtOpcUaServer and OpcUaServerOptions with the
shipped LDAP + anonymous-role + configurable security-profile prose.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- Admin-010: vendor Bootstrap 5.3.3 (CSS + JS bundle + maps + provenance
README) under wwwroot/lib/bootstrap and reference local paths from
App.razor — Admin no longer pulls Bootstrap from jsDelivr.
- Admin-011: swap FleetStatusPoller's three plain dictionaries for
ConcurrentDictionary so ResetCache can't race a poll tick.
- Admin-012: drop the EquipmentId column from EquipmentCsvImporter (per
admin-ui.md — equipment id is system-derived from EquipmentUuid);
EquipmentImportBatchService and the textarea placeholder updated to
match.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The Admin-003 fix gated every SignalR hub with [Authorize], but the server-side
Blazor HubConnection clients had no way to authenticate: the browser's HttpOnly
auth cookie is not reachable from the interactive circuit, so every hub negotiate
returned 401 and the Admin live-update feature was non-functional app-wide
(silently degraded on Hosts/ScriptLog, fatal on the cluster pages).
Introduce a token-based hub auth path:
- HubTokenService mints/validates short-lived tokens using ASP.NET Core Data
Protection (the same primitive that protects the auth cookie — no signing-key
management, no new packages). Tokens carry the user's name + roles.
- HubTokenAuthenticationHandler is a custom "HubToken" auth scheme that reads the
token from the Authorization: Bearer header (negotiate) or the access_token
query parameter (WebSocket upgrade).
- The "HubClients" authorization policy runs both the cookie and HubToken
schemes; the hub endpoints use RequireAuthorization("HubClients").
- AdminHubConnectionFactory builds hub connections with an AccessTokenProvider
that mints a fresh token for the circuit's authenticated user on every
(re)connect. All six hub-consuming pages now resolve connections through it.
Hub negotiate now returns 200 and the WebSocket upgrades (101); live updates
work. The best-effort try/catch guards added previously are kept as defence.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The Admin-003 fix gated every SignalR hub with [Authorize]/RequireAuthorization,
but the server-side HubConnection clients on ClusterDetail, AclsTab, RedundancyTab
and RoleGrants cannot forward the browser's HttpOnly auth cookie — so the hub
negotiate returns 401. Those four pages called HubConnection.StartAsync()
unguarded, so the 401 surfaced as an unhandled exception (a 500 page for the
prerendered ClusterDetail, a broken circuit for the others).
Wrap StartAsync/SendAsync in try/catch on all four, matching the established
best-effort pattern already used in Hosts.razor and ScriptLog.razor: the live
banner / live refresh degrades but the page renders. Restoring functional hub
live-updates needs a token-based hub auth scheme (cookie forwarding is not
viable across the prerender/interactive boundary) and is left as follow-up.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Core-002 fixed TriePermissionEvaluator to evaluate each request against
the session's bound AuthGenerationId rather than whatever the cache
currently holds. AuthorizationGate.BuildSessionState was not updated at
the same time: it hardcoded AuthGenerationId = 0, so the evaluator's
GetTrie(cluster, 0) call returned null for any generation != 0, causing
every gated operation to silently fail with NotGranted regardless of
actual grants. The 42 gate/matrix/deferred-hardening tests all started
failing as a result.
Fix: add an optional PermissionTrieCache parameter to AuthorizationGate;
BuildSessionState now stamps AuthGenerationId from the cache's current
generation for the session's cluster. AuthorizationBootstrap.BuildGateAsync
passes the cache it creates. All 7 test MakeGate helpers updated to pass
the cache so tests produce a valid AuthGenerationId. 433/433 server tests
now pass.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Replace silent Enum.TryParse fallback to None with a ParseSecurityProfile
helper that emits a startup Log.Warning naming the unsupported value and
listing recognised profiles; operators now see the misconfiguration
before any client connects rather than getting an unexplained None posture.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Advertise UserName token policy on any non-None security profile when
Ldap.Enabled; emit a startup LogWarning when Ldap.Enabled=true but
SecurityProfile=None so the misconfiguration is surfaced before clients
connect rather than silently producing no credential path.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Default AutoAcceptUntrustedClientCertificates to false in both
OpcUaServerOptions and Program.cs config fallback, aligning with
docs/security.md; auto-accept is now explicitly opt-in for dev use only.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Add configDbHealthy parameter to OpcUaApplicationHost; wire a
DbHealthCache (CanConnectAsync cached 10 s) in Program.cs so /healthz
reflects real config-DB reachability instead of the previous always-true
default; /healthz now returns 503 on a DB outage unless stale-config
cache is warm.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Add _nodeManagerDisposed field; set it under Lock in Dispose before
detaching the alarm-service handler; check it in OnAlarmServiceTransition
under the same Lock so an in-flight transition cannot dispatch to a
ConditionSink whose DriverNodeManager is being concurrently disposed.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Fix ReadRawAsync: correct XML doc from newest-first to oldest-first
(ascending source timestamp per OPC UA Part 11); move maxValuesPerNode
cap inside the time-window filter loop so paging limits apply to
in-window results only, not the whole buffer snapshot.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Configuration-002: sp_PublishGeneration is transaction-nesting aware
(BEGIN TRANSACTION vs SAVE TRANSACTION on @@TRANCOUNT) so a caller's outer
transaction survives a publish failure; sp_ValidateDraft wrapped in TRY/CATCH.
Configuration-003: ValidatePathLength uses the cluster's actual Enterprise/Site
lengths when available, falling back to the conservative approximation.
Configuration-006: ResilientConfigReader treats a command-timeout
TaskCanceledException as a fault (not caller cancellation) and falls back.
Configuration-009: removed the checked-in plaintext sa connection string;
CreateDbContext now requires OTOPCUA_CONFIG_CONNECTION.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
