Files
32dff7f1d6f34b67af16b45a2973500524a47070
Joseph Doherty
1f3343e61f
OpenTelemetry redundancy metrics + RoleChanged SignalR push. Closes instrumentation + live-push slices of task #198; the exporter wiring (OTLP vs Prometheus package decision) is split to new task #201 because the collector/scrape-endpoint choice is a fleet-ops decision that deserves its own PR rather than hardcoded here. New RedundancyMetrics class (Singleton-registered in DI) owning a System.Diagnostics.Metrics.Meter("ZB.MOM.WW.OtOpcUa.Redundancy", "1.0.0"). Three ObservableGauge instruments — otopcua.redundancy.primary_count / secondary_count / stale_count — all tagged by cluster.id, populated by SetClusterCounts(clusterId, primary, secondary, stale) which the poller calls at the tail of every tick; ObservableGauge callbacks snapshot the last value set under a lock so the reader (OTel collector, dotnet-counters) sees consistent tuples. One Counter — otopcua.redundancy.role_transition — tagged cluster.id, node.id, from_role, to_role; ideal for tracking "how often does Cluster-X failover" + "which node transitions most" aggregate queries. In-box Metrics API means zero NuGet dep here — the exporter PR adds OpenTelemetry.Extensions.Hosting + OpenTelemetry.Exporter.OpenTelemetryProtocol or OpenTelemetry.Exporter.Prometheus.AspNetCore to actually ship the data somewhere. FleetStatusPoller extended with role-change detection. Its PollOnceAsync now pulls ClusterNode rows alongside the existing ClusterNodeGenerationState scan, and a new PollRolesAsync walks every node comparing RedundancyRole to the _lastRole cache. On change: records the transition to RedundancyMetrics + emits a RoleChanged SignalR message to both FleetStatusHub.GroupName(cluster) + FleetStatusHub.FleetGroup so cluster-scoped + fleet-wide subscribers both see it. First observation per node is a bootstrap (cache fill) + NOT a transition — avoids spurious churn on service startup or pod restart. UpdateClusterGauges groups nodes by cluster + sets the three gauge values, using ClusterNodeService.StaleThreshold (shared 30s convention) for staleness so the /hosts page + the gauge agree. RoleChangedMessage record lives alongside NodeStateChangedMessage in FleetStatusPoller.cs. RedundancyTab.razor subscribes to the fleet-status hub on first parameters-set, filters RoleChanged events to the current cluster, reloads the node list + paints a blue info banner ("Role changed on node-a: Primary → Secondary at HH:mm:ss UTC") so operators see the transition without needing to poll-refresh the page. IAsyncDisposable closes the connection on tab swap-away. Two new RedundancyMetricsTests covering RecordRoleTransition tag emission (cluster.id + node.id + from_role + to_role all flow through the MeterListener callback) + ObservableGauge snapshot for two clusters (assert primary_count=1 for c1, stale_count=1 for c2). Existing FleetStatusPollerTests ctor-line updated to pass a RedundancyMetrics instance; all tests still pass. Full Admin.Tests suite 87/87 passing (was 85, +2). Admin project builds 0 errors. Task #201 captures the exporter-wiring follow-up — OpenTelemetry.Extensions.Hosting + OTLP vs Prometheus + /metrics endpoint decision, driven by fleet-ops infra direction.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
OpenTelemetry redundancy metrics + RoleChanged SignalR push. Closes instrumentation + live-push slices of task #198; the exporter wiring (OTLP vs Prometheus package decision) is split to new task #201 because the collector/scrape-endpoint choice is a fleet-ops decision that deserves its own PR rather than hardcoded here. New RedundancyMetrics class (Singleton-registered in DI) owning a System.Diagnostics.Metrics.Meter("ZB.MOM.WW.OtOpcUa.Redundancy", "1.0.0"). Three ObservableGauge instruments — otopcua.redundancy.primary_count / secondary_count / stale_count — all tagged by cluster.id, populated by SetClusterCounts(clusterId, primary, secondary, stale) which the poller calls at the tail of every tick; ObservableGauge callbacks snapshot the last value set under a lock so the reader (OTel collector, dotnet-counters) sees consistent tuples. One Counter — otopcua.redundancy.role_transition — tagged cluster.id, node.id, from_role, to_role; ideal for tracking "how often does Cluster-X failover" + "which node transitions most" aggregate queries. In-box Metrics API means zero NuGet dep here — the exporter PR adds OpenTelemetry.Extensions.Hosting + OpenTelemetry.Exporter.OpenTelemetryProtocol or OpenTelemetry.Exporter.Prometheus.AspNetCore to actually ship the data somewhere. FleetStatusPoller extended with role-change detection. Its PollOnceAsync now pulls ClusterNode rows alongside the existing ClusterNodeGenerationState scan, and a new PollRolesAsync walks every node comparing RedundancyRole to the _lastRole cache. On change: records the transition to RedundancyMetrics + emits a RoleChanged SignalR message to both FleetStatusHub.GroupName(cluster) + FleetStatusHub.FleetGroup so cluster-scoped + fleet-wide subscribers both see it. First observation per node is a bootstrap (cache fill) + NOT a transition — avoids spurious churn on service startup or pod restart. UpdateClusterGauges groups nodes by cluster + sets the three gauge values, using ClusterNodeService.StaleThreshold (shared 30s convention) for staleness so the /hosts page + the gauge agree. RoleChangedMessage record lives alongside NodeStateChangedMessage in FleetStatusPoller.cs. RedundancyTab.razor subscribes to the fleet-status hub on first parameters-set, filters RoleChanged events to the current cluster, reloads the node list + paints a blue info banner ("Role changed on node-a: Primary → Secondary at HH:mm:ss UTC") so operators see the transition without needing to poll-refresh the page. IAsyncDisposable closes the connection on tab swap-away. Two new RedundancyMetricsTests covering RecordRoleTransition tag emission (cluster.id + node.id + from_role + to_role all flow through the MeterListener callback) + ObservableGauge snapshot for two clusters (assert primary_count=1 for c1, stale_count=1 for c2). Existing FleetStatusPollerTests ctor-line updated to pass a RedundancyMetrics instance; all tests still pass. Full Admin.Tests suite 87/87 passing (was 85, +2). Admin project builds 0 errors. Task #201 captures the exporter-wiring follow-up — OpenTelemetry.Extensions.Hosting + OTLP vs Prometheus + /metrics endpoint decision, driven by fleet-ops infra direction.
Phase 3 PR 27 — Fleet status dashboard page. New /fleet route shows per-node apply state (ClusterNodeGenerationState joined with ClusterNode for the ClusterId) in a sortable table with summary cards for Total / Applied / Stale / Failed node counts. Stale detection: LastSeenAt older than 30s triggers a table-warning row class + yellow count card. Failed rows get table-danger + red card. Badge classes per LastAppliedStatus: Applied=bg-success, Failed=bg-danger, Applying=bg-info, unknown=bg-secondary. Timestamps rendered as relative-age strings ('42s ago', '15m ago', '3h ago', then absolute date for >24h). Error column is truncated to 320px with the full message in a tooltip so the table stays readable on wide fleets. Initial data load on OnInitializedAsync; auto-refresh every 5s via a Timer that calls InvokeAsync(RefreshAsync) — matches the FleetStatusPoller's 5s cadence so the dashboard sees the most recent state without polling ahead of the broadcaster. A Refresh button also kicks a manual reload; _refreshing gate prevents double-runs when the timer fires during an in-flight query. IServiceScopeFactory (matches FleetStatusPoller's pattern) creates a fresh DI scope per refresh so the per-page DbContext can't race the timer with the render thread; no new DI registrations needed. Live SignalR hub push is deliberately deferred to a follow-up PR — the existing FleetStatusHub + NodeStateChangedMessage already works for external JavaScript clients; wiring an in-process Blazor Server consumer adds HubConnectionBuilder plumbing that's worth its own focused change. Sidebar link added to MainLayout between Overview and Clusters. Full Admin.Tests Unit suite 14 pass / 0 fail — unchanged, no tests regressed. Full Admin build clean (0 errors, 0 warnings). Closes the 'no per-driver dashboard' gap from lmx-followups item #7 at the fleet level; per-host (platform/engine/Modbus PLC) granularity still needs a dedicated page that consumes IHostConnectivityProbe.GetHostStatuses from the Server process — that's the live-SignalR follow-up.
Phase 1 Stream E Admin UI — finish Blazor pages so operators can run the draft → publish → rollback workflow end-to-end without hand-executing SQL. Adds eight new scoped services that wrap the Configuration stored procs + managed validators: EquipmentService (CRUD with auto-derived EquipmentId per decision #125), UnsService (areas + lines), NamespaceService, DriverInstanceService (generic JSON DriverConfig editor per decision #94 — per-driver schema validation lands in each driver's phase), NodeAclService (grant + revoke with bundled-preset permission sets; full per-flag editor + bulk-grant + permission simulator deferred to v2.1), ReservationService (fleet-wide active + released reservation inspector + FleetAdmin-only sp_ReleaseExternalIdReservation wrapper with required-reason invariant), DraftValidationService (hydrates a DraftSnapshot from the draft's rows plus prior-cluster Equipment + active reservations, runs the managed DraftValidator to surface every rule in one pass for inline validation panel), AuditLogService (recent ConfigAuditLog reader). Pages: /clusters list with create-new shortcut; /clusters/new wizard that creates the cluster row + initial empty draft in one go; /clusters/{id} detail with 8 tabs (Overview / Generations / Equipment / UNS Structure / Namespaces / Drivers / ACLs / Audit) — tabs that write always target the active draft, published generations stay read-only; /clusters/{id}/draft/{gen} editor with live validation panel (errors list with stable code + message + context; publish button disabled while any error exists) and tab-embedded sub-components; /clusters/{id}/draft/{gen}/diff three-column view backed by sp_ComputeGenerationDiff with Added/Removed/Modified badges; Generations tab with per-row rollback action wired to sp_RollbackToGeneration; /reservations FleetAdmin-only page (CanPublish policy) with active + released lists and a modal release dialog that enforces non-empty reason and round-trips through sp_ReleaseExternalIdReservation; /login scaffold with stub credential accept + FleetAdmin-role cookie issuance (real LDAP bind via the ScadaLink-parity LdapAuthService is deferred until live GLAuth integration — marked in the login view and in the Phase 1 partial-exit TODO). Layout: sidebar gets Overview / Clusters / Reservations + AuthorizeView with signed-in username + roles + sign-out POST to /auth/logout; cascading authentication state registered for <AuthorizeView> to work in RenderMode.InteractiveServer. Integration testing: AdminServicesIntegrationTests creates a throwaway per-run database (same pattern as the Configuration test fixture), applies all three migrations, and exercises (1) create-cluster → add-namespace+UNS+driver+equipment → validate (expects zero errors) → publish (expects Published status) → rollback (expects one new Published + at least one Superseded); (2) cross-cluster namespace binding draft → validates to BadCrossClusterNamespaceBinding per decision #122. Old flat Components/Pages/Clusters.razor moved to Components/Pages/Clusters/ClustersList.razor so the Clusters folder can host tab sub-components without the razor generator creating a type-and-namespace collision. Dev appsettings.json connection string switched from Integrated Security to sa auth to match the otopcua-mssql container on port 14330 (remapped from 1433 to coexist with the native MSSQL14 Galaxy ZB instance). Browser smoke test completed: home page, clusters list, new-cluster form, cluster detail with a seeded row, reservations (redirected to login for anon user) all return 200 / 302-to-login as expected; full solution 928 pass / 1 pre-existing Phase 0 baseline failure. Phase 1 Stream E items explicitly deferred with TODOs: CSV import for Equipment, SignalR FleetStatusHub + AlertHub real-time push, bulk-grant workflow, permission-simulator trie, merge-equipment draft, AppServer-via-OI-Gateway end-to-end smoke test (decision #142), and the real LDAP bind replacing the Login page stub.
Admin /hosts red-badge + resilience columns + Polly telemetry observer. Closes task #164 (the remaining slice of Phase 6.1 Stream E.3 after the earlier publisher + hub PR). Three cooperating pieces wired together so the operator-facing /hosts table actually reflects the live Polly counters that the pipeline builder is producing. DriverResiliencePipelineBuilder gains an optional DriverResilienceStatusTracker ctor param — when non-null, every built pipeline wires Polly's OnRetry/OnOpened/OnClosed strategy-options callbacks into the tracker. OnRetry → tracker.RecordFailure (so ConsecutiveFailures climbs per retry), OnOpened → tracker.RecordBreakerOpen (stamps LastCircuitBreakerOpenUtc), OnClosed → tracker.RecordSuccess (resets the failure counter once the target recovers). Absent tracker = silent, preserving the unit-test constructor path + any deployment that doesn't care about resilience observability. Cancellation stays excluded from the failure count via the existing ShouldHandle predicate. HostStatusService.HostStatusRow extends with four new fields — ConsecutiveFailures, LastCircuitBreakerOpenUtc, CurrentBulkheadDepth, LastRecycleUtc — populated via a second LEFT JOIN onto DriverInstanceResilienceStatuses keyed on (DriverInstanceId, HostName). LEFT JOIN because brand-new hosts haven't been sampled yet; a missing row means zero failures + never-opened breaker, which is the correct default. New FailureFlagThreshold constant (=3, matches plan decision #143's conservative half-of-breaker convention) + IsFlagged predicate so the UI can pre-warn before the breaker actually trips. Hosts.razor paints three new columns between State and Last-transition — Fail# (bold red when flagged), In-flight (bulkhead-depth proxy), Breaker-opened (relative age). Per-row "Flagged" red badge alongside State when IsFlagged is true. Above the first cluster table, a red alert banner summarises the flagged-host count when ≥1 host is flagged, so operators see the problem before scanning rows. Three new tests in DriverResiliencePipelineBuilderTests — Tracker_RecordsFailure_OnEveryRetry verifies ConsecutiveFailures reaches RetryCount after a transient-forever operation, Tracker_StampsBreakerOpen_WhenBreakerTrips verifies LastBreakerOpenUtc is set after threshold failures on a Write pipeline, Tracker_IsolatesCounters_PerHost verifies one dead host does not leak failure counts into a healthy sibling. Full suite — Core.Tests 14/14 resilience-builder tests passing (11 existing + 3 new), Admin.Tests 72/72 passing, Admin project builds 0 errors. SignalR live push of status changes + browser visual review are deliberately left to a follow-up — this PR keeps the structural change minimal (polling refresh already exists in the page's 10s timer; SignalR would be a structural add that touches hub registration + client subscription).
Phase 1 Stream E Admin UI — finish Blazor pages so operators can run the draft → publish → rollback workflow end-to-end without hand-executing SQL. Adds eight new scoped services that wrap the Configuration stored procs + managed validators: EquipmentService (CRUD with auto-derived EquipmentId per decision #125), UnsService (areas + lines), NamespaceService, DriverInstanceService (generic JSON DriverConfig editor per decision #94 — per-driver schema validation lands in each driver's phase), NodeAclService (grant + revoke with bundled-preset permission sets; full per-flag editor + bulk-grant + permission simulator deferred to v2.1), ReservationService (fleet-wide active + released reservation inspector + FleetAdmin-only sp_ReleaseExternalIdReservation wrapper with required-reason invariant), DraftValidationService (hydrates a DraftSnapshot from the draft's rows plus prior-cluster Equipment + active reservations, runs the managed DraftValidator to surface every rule in one pass for inline validation panel), AuditLogService (recent ConfigAuditLog reader). Pages: /clusters list with create-new shortcut; /clusters/new wizard that creates the cluster row + initial empty draft in one go; /clusters/{id} detail with 8 tabs (Overview / Generations / Equipment / UNS Structure / Namespaces / Drivers / ACLs / Audit) — tabs that write always target the active draft, published generations stay read-only; /clusters/{id}/draft/{gen} editor with live validation panel (errors list with stable code + message + context; publish button disabled while any error exists) and tab-embedded sub-components; /clusters/{id}/draft/{gen}/diff three-column view backed by sp_ComputeGenerationDiff with Added/Removed/Modified badges; Generations tab with per-row rollback action wired to sp_RollbackToGeneration; /reservations FleetAdmin-only page (CanPublish policy) with active + released lists and a modal release dialog that enforces non-empty reason and round-trips through sp_ReleaseExternalIdReservation; /login scaffold with stub credential accept + FleetAdmin-role cookie issuance (real LDAP bind via the ScadaLink-parity LdapAuthService is deferred until live GLAuth integration — marked in the login view and in the Phase 1 partial-exit TODO). Layout: sidebar gets Overview / Clusters / Reservations + AuthorizeView with signed-in username + roles + sign-out POST to /auth/logout; cascading authentication state registered for <AuthorizeView> to work in RenderMode.InteractiveServer. Integration testing: AdminServicesIntegrationTests creates a throwaway per-run database (same pattern as the Configuration test fixture), applies all three migrations, and exercises (1) create-cluster → add-namespace+UNS+driver+equipment → validate (expects zero errors) → publish (expects Published status) → rollback (expects one new Published + at least one Superseded); (2) cross-cluster namespace binding draft → validates to BadCrossClusterNamespaceBinding per decision #122. Old flat Components/Pages/Clusters.razor moved to Components/Pages/Clusters/ClustersList.razor so the Clusters folder can host tab sub-components without the razor generator creating a type-and-namespace collision. Dev appsettings.json connection string switched from Integrated Security to sa auth to match the otopcua-mssql container on port 14330 (remapped from 1433 to coexist with the native MSSQL14 Galaxy ZB instance). Browser smoke test completed: home page, clusters list, new-cluster form, cluster detail with a seeded row, reservations (redirected to login for anon user) all return 200 / 302-to-login as expected; full solution 928 pass / 1 pre-existing Phase 0 baseline failure. Phase 1 Stream E items explicitly deferred with TODOs: CSV import for Equipment, SignalR FleetStatusHub + AlertHub real-time push, bulk-grant workflow, permission-simulator trie, merge-equipment draft, AppServer-via-OI-Gateway end-to-end smoke test (decision #142), and the real LDAP bind replacing the Login page stub.
Admin RoleGrants page — LDAP-group → Admin-role mapping CRUD. Closes the RoleGrantsTab slice of task #144 (Phase 6.2 Stream D follow-up); the remaining three sub-items (Probe-this-permission on AclsTab, SignalR invalidation on role/ACL changes, draft-diff ACL section) are split into new follow-up task #196 so each can ship independently. The permission-trie evaluator + ILdapGroupRoleMappingService already exist from Phase 6.2 Streams A + B — this PR adds the consuming UI + the DI registration that was missing. New /role-grants page at Components/Pages/RoleGrants.razor registered in MainLayout's sidebar next to Certificates. Lists every LdapGroupRoleMapping row with columns LDAP group / Role / Scope (Fleet-wide or Cluster:X) / Created / Notes / Revoke. Add-grant form takes LDAP group DN + AdminRole dropdown (ConfigViewer, ConfigEditor, FleetAdmin) + Fleet-wide checkbox + Cluster dropdown (disabled when Fleet-wide checked) + optional Notes. Service-layer invariants — IsSystemWide=true + ClusterId=null, or IsSystemWide=false + ClusterId populated — enforced in ValidateInvariants; UI catches InvalidLdapGroupRoleMappingException and displays the message in a red alert. ILdapGroupRoleMappingService was present in the Configuration project from Stream A but never registered in the Admin DI container — this PR adds the AddScoped registration so the injection can resolve. Control-plane/data-plane separation note rendered in an info banner at the top of the page per decision #150 (these grants do NOT govern OPC UA data-path authorization; NodeAcl rows are read directly by the permission-trie evaluator without consulting role mappings). Admin project builds 0 errors; Admin.Tests 72/72 passing. Task #196 created to track: (1) AclsTab Probe-this-permission form that takes (ldap group, node path, permission flag) and runs it through the permission trie, showing which row granted it + the actual resolved grant; (2) SignalR invalidation — push a RoleGrantsChanged event when rows are created/deleted so connected Admin sessions reload without polling, ditto NodeAclChanged on ACL writes; (3) DiffViewer ACL section — show NodeAcl + LdapGroupRoleMapping deltas between draft + published alongside equipment/uns diffs.