# Phase 6.3 — Redundancy Runtime

> **Status**: **SHIPPED (core)** 2026-04-19 — Streams B (ServiceLevelCalculator + RecoveryStateManager) and D core (ApplyLeaseRegistry) merged to `v2` in PR #89. Exit gate in PR #90.
>
> Deferred follow-ups (tracked separately):
> - Stream A — RedundancyCoordinator cluster-topology loader (task #145).
> - Stream C — OPC UA node wiring: ServiceLevel + ServerUriArray + RedundancySupport (task #147).
> - Stream E — Admin UI RedundancyTab + OpenTelemetry metrics + SignalR (task #149).
> - Stream F — client interop matrix + Galaxy MXAccess failover test (task #150).
> - sp_PublishGeneration pre-publish validator rejecting unsupported RedundancyMode values (task #148 part 2 — SQL-side).
>
> Baseline pre-Phase-6.3: 1097 solution tests → post-Phase-6.3 core: 1137 passing (+40 net).
>
> **Branch**: `v2/phase-6-3-redundancy-runtime`
> **Estimated duration**: 2 weeks
> **Predecessor**: Phase 6.2 (Authorization) — reuses the Phase 6.1 health endpoints for cluster-peer probing
> **Successor**: Phase 6.4 (Admin UI completion)

## Phase Objective

Land the non-transparent redundancy protocol end-to-end: two `OtOpcUa.Server` instances in a `ServerCluster` each expose a live `ServiceLevel` node whose value reflects that instance's suitability to serve traffic, advertise each other via `ServerUriArray`, and transition role (Primary ↔ Backup) based on health + operator intent.

Closes these gaps:

1. **Dynamic `ServiceLevel`** — OPC UA Part 5 §6.3.34 specifies a Byte (0..255) that clients poll to pick the healthiest server. Our server publishes it as a static value today.
2. **`ServerUriArray` broadcast** — Part 4 specifies that every node in a redundant pair should advertise its peers' ApplicationUris. Currently advertises only its own.
3. **Primary / Backup role coordination** — entities carry `RedundancyRole` but the runtime doesn't read it; no peer health probing; no role-transfer on primary failure.
4. **Mid-apply dip** — decision-level expectation that a server mid-generation-apply should report a *lower* ServiceLevel so clients cut over to the peer during the apply window. Not implemented.

## Scope — What Changes

| Concern | Change |
|---------|--------|
| `OtOpcUa.Server` → new `Server.Redundancy` sub-namespace | `RedundancyCoordinator` singleton. Resolves the current node's `ClusterNode` row at startup, loads peers, runs **two-layer peer health probe**: (a) `/healthz` every 2 s as the fast-fail (inherits Phase 6.1 semantics — HTTP + DB/cache healthy); (b) `UaHealthProbe` every 10 s — opens a lightweight OPC UA client session to the peer + reads its `ServiceLevel` node + verifies endpoint serves data. Authority decisions use UaHealthProbe; `/healthz` is used only to avoid wasting UA probes when peer is obviously down. |
| Publish-generation fencing | Topology + role decisions are stamped with a monotonic `ConfigGenerationId` from the shared config DB. Coordinator re-reads topology via CAS on `(ClusterId, ExpectedGeneration)` → new row; peers reject state propagated from a lower generation. Prevents split-publish races. |
| `InvalidTopology` runtime state | If both nodes detect >1 Primary AFTER startup (config-DB drift during a publish), both self-demote to ServiceLevel 2 until convergence. Neither node serves authoritatively; clients pick the healthier alternative or reconnect later. |
| OPC UA server root | `ServiceLevel` variable node becomes a `BaseDataVariable` whose value updates on `RedundancyCoordinator` state change. `ServerUriArray` array variable includes **self + peers** in stable deterministic ordering (decision per OPC UA Part 4 §6.6.2.2). `RedundancySupport` stays static (set from `RedundancyMode` at startup); `Transparent` mode validated pre-publish, not rejected at startup. |
| `RedundancyCoordinator` computation | **8-state ServiceLevel matrix** — avoids OPC UA Part 5 §6.3.34 collision (`0=Maintenance`, `1=NoData`). Operator-declared maintenance only = **0**. Unreachable / Faulted = **1**. In-range operational states occupy **2..255**: Authoritative-Primary = **255**; Isolated-Primary (peer unreachable, self serving) = **230**; Primary-Mid-Apply = **200**; Recovering-Primary (post-fault, dwell not met) = **180**; Authoritative-Backup = **100**; Isolated-Backup (primary unreachable, "take over if asked") = **80**; Backup-Mid-Apply = **50**; Recovering-Backup = **30**; `InvalidTopology` (runtime detects >1 Primary) = **2** (detected-inconsistency band — below normal operation). Full matrix documented in `docs/Redundancy.md` update. |
| Role transition | Split-brain avoidance: role is *declared* in the shared config DB (`ClusterNode.RedundancyRole`), not elected at runtime. An operator flips the row (or a failover script does). Coordinator only reads; never writes. |
| `sp_PublishGeneration` hook | Uses named **apply leases** keyed to `(ConfigGenerationId, PublishRequestId)`. `await using var lease = coordinator.BeginApplyLease(...)`. Disposal on any exit path (success, exception, cancellation) decrements. Watchdog auto-closes any lease older than `ApplyMaxDuration` (default 10 min) → ServiceLevel can't stick at mid-apply. Pre-publish validator rejects unsupported `RedundancyMode` (e.g. `Transparent`) with a clear error so runtime never sees an invalid state. |
| Admin UI `/cluster/{id}` page | New `RedundancyTab.razor` — shows current node's role + ServiceLevel + peer reachability. FleetAdmin can trigger a role-swap by editing `ClusterNode.RedundancyRole` + publishing a draft. |
| Metrics | New OpenTelemetry metrics: `ot_opcua_service_level{cluster,node}`, `ot_opcua_peer_reachable{cluster,node,peer}`, `ot_opcua_apply_in_progress{cluster,node}`. Sink via Phase 6.1 observability layer. |

## Scope — What Does NOT Change

| Item | Reason |
|------|--------|
| OPC UA authn / authz | Phases 6.2 + prior. Redundancy is orthogonal. |
| Driver layer | Drivers aren't redundancy-aware; they run on each node independently against the same equipment. The server layer handles the ServiceLevel story. |
| Automatic failover / election | Explicitly out of scope. Non-transparent = client picks which server to use via ServiceLevel + ServerUriArray. We do NOT ship consensus, leader election, or automatic promotion. Operator-driven failover is the v2.0 model per decision #79–85. |
| Transparent redundancy (`RedundancySupport=Transparent`) | Not supported. If the operator asks for it the server fails startup with a clear error. |
| Historian redundancy | Galaxy Historian's own redundancy (two historians on two CPUs) is out of scope. The Galaxy driver talks to whichever historian is reachable from its node. |

## Entry Gate Checklist

- [ ] Phase 6.1 merged (uses `/healthz` for peer probing)
- [ ] `CLAUDE.md` §Redundancy + `docs/Redundancy.md` re-read
- [ ] Decisions #79–85 re-skimmed
- [ ] `ServerCluster`/`ClusterNode`/`RedundancyRole`/`RedundancyMode` entities + existing migration reviewed
- [ ] OPC UA Part 4 §Redundancy + Part 5 §6.3.34 (ServiceLevel) re-skimmed
- [ ] Dev box has two OtOpcUa.Server instances configured against the same cluster — one designated Primary, one Backup — for integration testing

## Task Breakdown

### Stream A — Cluster topology loader (3 days)

1. **A.1** `RedundancyCoordinator` startup path: reads `ClusterNode` row for the current node (identified by `appsettings.json` `Cluster:NodeId`), reads the cluster's peer list, validates invariants (no duplicate `ApplicationUri`, at most one `Primary` per cluster if `RedundancyMode.WarmActive`, at most two nodes total in v2.0 per decision #83).
2. **A.2** Topology subscription — coordinator re-reads on `sp_PublishGeneration` confirmation so an operator role-swap takes effect after publish (no process restart needed).
3. **A.3** Tests: two-node cluster seed, one-node cluster seed (degenerate), duplicate-uri rejection.

### Stream B — Peer health probing + ServiceLevel computation (6 days, widened)

1. **B.1** `PeerHttpProbeLoop` per peer at 2 s — calls peer's `/healthz`, 1 s timeout, exponential backoff on sustained failure. Used as fast-fail.
2. **B.2** `PeerUaProbeLoop` per peer at 10 s — opens an OPC UA client session to the peer (reuses Phase 5 `Driver.OpcUaClient` stack), reads peer's `ServiceLevel` node + verifies endpoint serves data. Short-circuit: if HTTP probe is failing, skip UA probe (no wasted sessions).
3. **B.3** `ServiceLevelCalculator.Compute(role, selfHealth, peerHttpHealthy, peerUaHealthy, applyInProgress, recoveryDwellMet, topologyValid) → byte`. 8-state matrix per §Scope. `topologyValid=false` forces InvalidTopology = 2 regardless of other inputs.
4. **B.4** `RecoveryStateManager`: after a `Faulted → Healthy` transition, hold driver in `Recovering` band (180 Primary / 30 Backup) for `RecoveryDwellTime` (default 60 s) AND require one positive publish witness (successful `Read` on a reference node) before entering Authoritative band.
5. **B.5** Calculator reacts to inputs via `IObserver` so changes immediately push to the OPC UA `ServiceLevel` node.
6. **B.6** Tests: **64-case matrix** covering role × self-health × peer-http × peer-ua × apply × recovery × topology. Specific cases flagged: Primary-with-unreachable-peer-serves-at-230 (authority retained); Backup-with-unreachable-primary-escalates-to-80 (not auto-promote); InvalidTopology demotes both nodes; Recovering dwell + publish-witness blocks premature return to 255.

### Stream C — OPC UA node wiring (3 days)

1. **C.1** `ServiceLevel` variable node created under `ServerStatus` at server startup. Type `Byte`, AccessLevel = CurrentRead only. Subscribe to `ServiceLevelCalculator` observable; push updates via `DataChangeNotification`.
2. **C.2** `ServerUriArray` variable node under `ServerCapabilities`. Array of `String`, **includes self + peers** with deterministic ordering (self first). Updates on topology change. Compliance test asserts local-plus-peer membership.
3. **C.3** `RedundancySupport` variable — static at startup from `RedundancyMode`. Values: `None`, `Cold`, `Warm`, `WarmActive`, `Hot`. Unsupported values (`Transparent`, `HotAndMirrored`) are rejected **pre-publish** by validator — runtime never sees them.
4. **C.4** Client.CLI cutover test: connect to primary, read `ServiceLevel` → 255; pause primary apply → 200; unreachable peer while apply in progress → 200 (apply dominates peer-unreachable per matrix); client sees peer via `ServerUriArray`; fail primary → client reconnects to peer at 80 (isolated-backup band).

### Stream D — Apply-window integration (3 days)

1. **D.1** `sp_PublishGeneration` caller wraps the apply in `await using var lease = coordinator.BeginApplyLease(generationId, publishRequestId)`. Lease keyed to `(ConfigGenerationId, PublishRequestId)` so concurrent publishes stay isolated. Disposal decrements on every exit path.
2. **D.2** `ApplyLeaseWatchdog` auto-closes leases older than `ApplyMaxDuration` (default 10 min) so a crashed publisher can't pin the node at mid-apply.
3. **D.3** Pre-publish validator in `sp_PublishGeneration` rejects unsupported `RedundancyMode` values (`Transparent`, `HotAndMirrored`) with a clear error message — runtime never sees an invalid mode.
4. **D.4** Tests: (a) mid-apply client subscribes → sees ServiceLevel drop → sees restore; (b) lease leak via `ThreadAbort` / cancellation → watchdog closes; (c) publish rejected for `Transparent` → operator-actionable error.

### Stream E — Admin UI + metrics (3 days)

1. **E.1** `RedundancyTab.razor` under `/cluster/{id}/redundancy`. Shows each node's role, current ServiceLevel (with band label per 8-state matrix), peer reachability (HTTP + UA probe separately), last apply timestamp. Role-swap button posts a draft edit on `ClusterNode.RedundancyRole`; publish applies.
2. **E.2** OpenTelemetry meter export: `ot_opcua_service_level{cluster,node}` gauge + `ot_opcua_peer_reachable{cluster,node,peer,kind=http|ua}` + `ot_opcua_apply_in_progress{cluster,node}` + `ot_opcua_topology_valid{cluster}`. Sink via Phase 6.1 observability.
3. **E.3** SignalR push: `FleetStatusHub` broadcasts ServiceLevel changes so the Admin UI updates within ~1 s of the coordinator observing a peer flip.

### Stream F — Client-interoperability matrix (3 days, new)

1. **F.1** Validate ServiceLevel-driven cutover against **Ignition 8.1 + 8.3**, **Kepware KEPServerEX 6.x**, **Aveva OI Gateway 2020R2 + 2023R1**. For each: configure the client with both endpoints, verify it honors `ServiceLevel` + `ServerUriArray` during primary failover.
2. **F.2** Clients that don't honour the standards (doc field — may include Kepware and OI Gateway per Codex review) get an explicit compatibility-matrix entry: "requires manual backup-endpoint config / vendor-specific redundancy primitives". Documented in `docs/Redundancy.md`.
3. **F.3** Galaxy MXAccess failover test — boot Galaxy.Proxy on both nodes, kill Primary, assert Galaxy consumer reconnects to Backup within `(SessionTimeout + KeepAliveInterval × 3)`. Document required session-timeout config in `docs/Redundancy.md`.

## Compliance Checks (run at exit gate)

- [ ] **OPC UA band compliance**: `0=Maintenance` reserved, `1=NoData` reserved. Operational states in 2..255 per 8-state matrix.
- [ ] **Authoritative-Primary** ServiceLevel = 255.
- [ ] **Isolated-Primary** (peer unreachable, self serving) = 230 — Primary retains authority.
- [ ] **Primary-Mid-Apply** = 200.
- [ ] **Recovering-Primary** = 180 with dwell + publish witness enforced.
- [ ] **Authoritative-Backup** = 100.
- [ ] **Isolated-Backup** (primary unreachable) = 80 — does NOT auto-promote.
- [ ] **InvalidTopology** = 2 — both nodes self-demote when >1 Primary detected runtime.
- [ ] **ServerUriArray** returns self + peer URIs, self first.
- [ ] **UaHealthProbe authority**: integration test — peer returns HTTP 200 but OPC UA endpoint unreachable → coordinator treats peer as UA-unhealthy; peer is not a valid authority source.
- [ ] **Apply-lease disposal**: leases close on exception, cancellation, and watchdog timeout; ServiceLevel never sticks at mid-apply band.
- [ ] **Transparent-mode rejection**: attempting to publish `RedundancyMode=Transparent` is blocked at `sp_PublishGeneration`; runtime never sees an invalid mode.
- [ ] **Role transition via operator publish**: FleetAdmin swaps `RedundancyRole` in a draft, publishes; both nodes re-read topology on publish confirmation + flip ServiceLevel — no restart.
- [ ] **Client.CLI cutover**: with primary halted, Client.CLI that was connected to primary sees primary drop + reconnects to backup via `ServerUriArray`.
- [ ] **Client interoperability matrix** (Stream F): Ignition 8.1 + 8.3 honour ServiceLevel; Kepware + Aveva OI Gateway findings documented.
- [ ] **Galaxy MXAccess failover**: end-to-end test — primary kill → Galaxy consumer reconnects to backup within session-timeout budget.
- [ ] No regression in existing driver test suites; no regression in `/healthz` reachability under redundancy load.

## Risks and Mitigations

| Risk | Likelihood | Impact | Mitigation |
|------|:----------:|:------:|------------|
| Split-brain from operator race (both nodes marked Primary) | Low | High | Coordinator rejects startup if its cluster has >1 Primary row; logs + fails fast. Document as a publish-time validation in `sp_PublishGeneration`. |
| ServiceLevel thrashing on flaky peer | Medium | Medium | 2 s probe interval + 3-sample smoothing window; only declares a peer unreachable after 3 consecutive failed probes |
| Client ignores ServiceLevel and stays on broken primary | Medium | Medium | Documented in `docs/Redundancy.md` — non-transparent redundancy requires client cooperation; most SCADA clients (Ignition, Kepware, Aveva OI Gateway) honor it. Unit-test the advertised values; field behavior is client-responsibility |
| Apply-window counter leaks on exception | Low | High | `BeginApplyWindow` returns `IDisposable`; `using` syntax enforces paired decrement; unit test for exception-in-apply path |
| `HttpClient` probe leaks sockets | Low | Medium | Single shared `HttpClient` per coordinator (not per-probe); timeouts tight to avoid keeping connections open during peer downtime |

## Completion Checklist

- [ ] Stream A: topology loader + tests
- [ ] Stream B: peer probe + ServiceLevel calculator + 32-case matrix tests
- [ ] Stream C: ServiceLevel / ServerUriArray / RedundancySupport node wiring + Client.CLI smoke test
- [ ] Stream D: apply-window integration + nested-apply counter
- [ ] Stream E: Admin `RedundancyTab` + OpenTelemetry metrics + SignalR push
- [ ] `phase-6-3-compliance.ps1` exits 0; exit-gate doc; `docs/Redundancy.md` updated with the ServiceLevel matrix

## Adversarial Review — 2026-04-19 (Codex, thread `019da490-3fa0-7340-98b8-cceeca802550`)

1. **Crit · ACCEPT** — No publish-generation fencing enables split-publish advertising both as authoritative. **Change**: coordinator CAS on a monotonic `ConfigGenerationId`; every topology decision is generation-stamped; peers reject state propagated from a lower generation.
2. **Crit · ACCEPT** — `>1 Primary` at startup covered but runtime containment missing when invalid topology appears later (mid-apply race). **Change**: add runtime `InvalidTopology` state — both nodes self-demote to ServiceLevel 2 (the "detected inconsistency" band, below normal operation) until convergence.
3. **High · ACCEPT** — `0 = Faulted` collides with OPC UA Part 5 §6.3.34 semantics where 0 means **Maintenance** and 1 means NoData. **Change**: reserve **0** for operator-declared maintenance-mode only; Faulted/unreachable uses **1** (NoData); in-range degraded states occupy 2..199.
4. **High · ACCEPT** — Matrix collapses distinct operational states onto the same value. **Change**: matrix expanded to Authoritative-Primary=255, Isolated-Primary=230 (peer unreachable — still serving), Primary-Mid-Apply=200, Recovering-Primary=180, Authoritative-Backup=100, Isolated-Backup=80 (primary unreachable — "take over if asked"), Backup-Mid-Apply=50, Recovering-Backup=30.
5. **High · ACCEPT** — `/healthz` from 6.1 is HTTP-healthy but doesn't guarantee OPC UA data plane. **Change**: add a redundancy-specific probe `UaHealthProbe` — issues a `ReadAsync(ServiceLevel)` against the peer's OPC UA endpoint via a lightweight client session. `/healthz` remains the fast-fail; the UA probe is the authority signal.
6. **High · ACCEPT** — `ServerUriArray` must include self + peers, not peers only. **Change**: array contains `[self.ApplicationUri, peer.ApplicationUri]` in stable deterministic ordering; compliance test asserts local-plus-peer membership.
7. **Med · ACCEPT** — No `Faulted → Recovering → Healthy` path. **Change**: add `Recovering` state with min dwell time (60 s default) + positive publish witness (one successful Read on a reference node) before returning to Healthy. Thrash-prevention.
8. **Med · ACCEPT** — Topology change during in-flight probe undefined. **Change**: every probe task tagged with `ConfigGenerationId` at dispatch; obsolete results discarded; in-flight probes cancelled on topology reload.
9. **Med · ACCEPT** — Apply-window counter race on exception/cancellation/async ownership. **Change**: apply-window is a named lease keyed to `(ConfigGenerationId, PublishRequestId)` with disposal enforced via `await using`; watchdog detects leased-but-abandoned and force-closes after `ApplyMaxDuration` (default 10 min).
10. **High · ACCEPT** — Ignition + Kepware + Aveva OI Gateway `ServiceLevel` compliance is unverified. **Change**: risk elevated to High; add Stream F (new) — build an interop matrix: validate against Ignition 8.1/8.3, Kepware KEPServerEX 6.x, Aveva OI Gateway 2020R2 + 2023R1. Document per-client cutover behaviour. Field deployments get a documented compatibility table; clients that ignore ServiceLevel documented as requiring explicit backup-endpoint config.
11. **Med · ACCEPT** — Galaxy MXAccess re-session on Primary death not in acceptance. **Change**: Stream F adds an end-to-end failover smoke test that boots Galaxy.Proxy on both nodes, kills Primary, asserts Galaxy consumer reconnects to Backup within `(SessionTimeout + KeepAliveInterval × 3)` budget. `docs/Redundancy.md` updated with required session timeouts.
12. **Med · ACCEPT** — Transparent-mode startup rejection is outage-prone. **Change**: `sp_PublishGeneration` validates `RedundancyMode` pre-publish — unsupported values reject the publish attempt with a clear validation error; runtime never sees an unsupported mode. Last-good config stays active.