docs(717): record the owner's 2026-11-13 answers in the Warm redundancy spec and plan

Standby writes and Part 9 methods refused BadInvalidState under Warm, promote while the Primary is
unreachable (OpcUa:Redundancy:PromoteWhilePrimaryUnreachable, default on), standby status code
BadOutOfService, no ConfigDb migration, bands as drafted. ScadaBridge#138 is the filed prerequisite.

Claude-Session: https://claude.ai/code/session_01GT67ACmpMR7b9LGRPLijMi
This commit is contained in:
Joseph Doherty
2026-11-13 02:43:00 -05:00
parent c0ba450d46
commit a42c9d7f7c
2 changed files with 60 additions and 40 deletions
@@ -1,8 +1,9 @@
# Warm non-transparent redundancy with a degraded standby (Gitea #717): design spec # Warm non-transparent redundancy with a degraded standby (Gitea #717): design spec
Status: **DRAFT for owner approval.** No product code changes until this spec and the companion plan Status: **APPROVED by the owner 2026-11-13** (#717), together with the companion plan
(`2026-11-11-warm-redundancy-717-plan.md`) are approved. Parent gate: #615 (G8 CTT, G9 production-client (`2026-11-11-warm-redundancy-717-plan.md`). The §9 questions are answered; the answers are recorded in §9 and
cutover). These stay open and separate from everything here. applied throughout. Parent gate: #615 (G8 CTT, G9 production-client cutover). These stay open and separate from
everything here.
## 1. Owner decisions this spec builds on (not re-opened) ## 1. Owner decisions this spec builds on (not re-opened)
@@ -69,6 +70,15 @@ this node — the node is **Standby**, whatever the member count. This is strict
lone-node allow (`PrimaryGatePolicy.cs:21`): a lone node goes Active within one debounce (~250 ms) of its own lone-node allow (`PrimaryGatePolicy.cs:21`): a lone node goes Active within one debounce (~250 ms) of its own
singleton's first publish. That gate is **not** changed for writes/acks (D7). The Warm verdict is a separate output. singleton's first publish. That gate is **not** changed for writes/acks (D7). The Warm verdict is a separate output.
**Promote while the Primary is unreachable (owner answer Q2, `OpcUa:Redundancy:PromoteWhilePrimaryUnreachable`,
default `true`).** A SIGKILLed Primary stays an Up-but-unreachable member until `auto-down` removes it (~25–30 s),
and the singleton usually died with it, so (2) and (3) would hold the survivor Standby for that whole window. With
the option on, (2) and (3) are evaluated with **unreachable** members excluded: (2′) self is the oldest Up `driver`
member among the members this node can reach, and (3′) the latest accepted snapshot names either self or an
**unreachable** member as `Primary`. This is the #672 pattern (`ContinuousHistorization:CaptureWhilePrimaryUnreachable`).
A node that has never accepted a snapshot is still Standby. The cost is dual-Active while a live Primary is merely
unreachable, and for the whole of a real partition.
Accepted trade: an `auto-down` partition can give two Active nodes, as it already does for the data-plane gate Accepted trade: an `auto-down` partition can give two Active nodes, as it already does for the data-plane gate
(`docs/Redundancy.md` §Split-brain). For reads this is benign, because both nodes serve the same Good data. (`docs/Redundancy.md` §Split-brain). For reads this is benign, because both nodes serve the same Good data.
@@ -118,15 +128,15 @@ creation from the node instance and the `AttributeId`, never from namespace or C
| Service | Standby behaviour | Status the client sees | | Service | Standby behaviour | Status the client sees |
|---|---|---| |---|---|---|
| Read (ordinary tag `Value`) | Projected. The stored value is kept internally and continues to update from acquisition. | `BadNoCommunication`, `Value = null`, `ServerTimestamp = now`. Operation result Good. | | Read (ordinary tag `Value`) | Projected. The stored value is kept internally and continues to update from acquisition. | `BadOutOfService`, `Value = null`, `ServerTimestamp = now`. Operation result Good. |
| Read (other attributes, condition children, ns=0 incl. ServiceLevel/ServerStatus/ServerRedundancy) | Unchanged | Good / as today | | Read (other attributes, condition children, ns=0 incl. ServiceLevel/ServerStatus/ServerRedundancy) | Unchanged | Good / as today |
| CreateMonitoredItems (ordinary tag) | Created normally (ACL unchanged). Initial value is projected. | Item result Good; first notification `BadNoCommunication`/null | | CreateMonitoredItems (ordinary tag) | Created normally (ACL unchanged). Initial value is projected. | Item result Good; first notification `BadOutOfService`/null |
| Existing data items at demotion | One `BadNoCommunication`/null notification is queued, ignoring filters. Later samples project to the same value, so the SDK's change filter emits nothing further. **D4:** any Good ordinary-tag notification that would be first-delivered after the demotion step is rewritten to `BadNoCommunication`/null at Publish/Republish (§4.6). | `BadNoCommunication` | | Existing data items at demotion | One `BadOutOfService`/null notification is queued, ignoring filters. Later samples project to the same value, so the SDK's change filter emits nothing further. **D4:** any Good ordinary-tag notification that would be first-delivered after the demotion step is rewritten to `BadOutOfService`/null at Publish/Republish (§4.6). | `BadOutOfService` |
| Write | **Unchanged (D7):** the existing gate rejects off-Primary through the #5 self-correction (synchronous Good, then revert + audit). Under Warm the revert is invisible because the value is projected Bad anyway. See open question Q1. | as today | | Write (ordinary tag `Value`) | **Refused synchronously under Warm only (owner answer Q1).** The node manager refuses the write before it reaches the driver; nothing is reverted and no device is touched. `Unset`/`None` keep today's #5 self-correction (synchronous Good, then revert + audit) off-Primary (D7). | `BadInvalidState` |
| HistoryRead | Unchanged. Served from the shared backend, which is not live acquisition. | as today | | HistoryRead | Unchanged. Served from the shared backend, which is not live acquisition. | as today |
| HistoryUpdate | Unchanged | `BadInvalidState` off-Primary (#674) | | HistoryUpdate | Unchanged | `BadInvalidState` off-Primary (#674) |
| Browse / TranslateBrowsePaths / RegisterNodes | Unchanged | as today | | Browse / TranslateBrowsePaths / RegisterNodes | Unchanged | as today |
| Call: Part 9 methods | Unchanged (D7): native ack `BadInvalidState` off-Primary; scripted ack/shelve accepted. The resulting condition event counts as a new alarm event and is suppressed (§4.5). See Q4. | as today | | Call: Part 9 methods | **Refused under Warm (owner answer Q4).** Every Part 9 condition method (Acknowledge, Confirm, AddComment, Enable/Disable, the shelving methods) on a native or scripted condition is refused before it changes any state. `Unset`/`None` keep today's behaviour: native ack `BadInvalidState` off-Primary, scripted ack/shelve accepted (D7). | `BadInvalidState` |
| Event subscriptions (Server object, folders, direct) | Stay established, in Reporting mode, with the ACL unchanged | — | | Event subscriptions (Server object, folders, direct) | Stay established, in Reporting mode, with the ACL unchanged | — |
| Model-change / audit events | **Not suppressed.** They are not alarm events, and browse consistency needs them. | — | | Model-change / audit events | **Not suppressed.** They are not alarm events, and browse consistency needs them. | — |
| Session / secure channel / ServerStatus.State | Unchanged (`Running`) | — | | Session / secure channel / ServerStatus.State | Unchanged (`Running`) | — |
@@ -168,7 +178,7 @@ mechanism that passed the provenance spike (c7851, 66 assertions) and the delive
cached SDK objects are never mutated. cached SDK objects are never mutated.
- If the node is Active at first delivery, the original message is the choice. - If the node is Active at first delivery, the original message is the choice.
- If the node is Standby, a detached clone is the choice, with every `OrdinaryTag` notification replaced by - If the node is Standby, a detached clone is the choice, with every `OrdinaryTag` notification replaced by
`BadNoCommunication`/null. `BadOutOfService`/null.
- A Republish returns the recorded choice. First delivery *through* Republish is decided the same way. - A Republish returns the recorded choice. First delivery *through* Republish is decided the same way.
- A short lock orders the choice against the role flip. No lock spans an await. - A short lock orders the choice against the role flip. No lock spans an await.
3. **Out of scope of the rewrite**, by D4: messages already in flight before the demotion step, and Republish of 3. **Out of scope of the rewrite**, by D4: messages already in flight before the demotion step, and Republish of
@@ -197,7 +207,7 @@ this order.
2. Switch the delivery-history choice to Standby. This is the D4 boundary. 2. Switch the delivery-history choice to Standby. This is the D4 boundary.
3. Suppress alarm delivery. 3. Suppress alarm delivery.
4. Close the read projection. 4. Close the read projection.
5. Queue `BadNoCommunication` to each live ordinary-tag item. 5. Queue `BadOutOfService` to each live ordinary-tag item.
Invariant: **an Active-band ServiceLevel is never visible while the projection is closed.** If the ServiceLevel Invariant: **an Active-band ServiceLevel is never visible while the projection is closed.** If the ServiceLevel
publish fails (#719 retry path), a promotion stops before step 6 and retries. A demotion continues anyway: failing publish fails (#719 retry path), a promotion stops before step 6 and retries. A demotion continues anyway: failing
@@ -217,25 +227,27 @@ ScadaBridge consumes this server as an OPC UA client, through `RealOpcUaClient`
`FailoverRetryCount` consecutive *unstable disconnects* `FailoverRetryCount` consecutive *unstable disconnects*
(`../ScadaBridge/src/ZB.MOM.WW.ScadaBridge.DataConnectionLayer/Actors/DataConnectionActor.cs:655-690`; (`../ScadaBridge/src/ZB.MOM.WW.ScadaBridge.DataConnectionLayer/Actors/DataConnectionActor.cs:655-690`;
`RealOpcUaClient.cs` sits in that project's `Adapters/`). `RealOpcUaClient.cs` sits in that project's `Adapters/`).
`BadNoCommunication` is not a connection-level status (`:1974-1983`), so it maps to plain Bad quality, and since `BadOutOfService` is not a connection-level status (`:1974-1983`), so it maps to plain Bad quality, and since
ScadaBridge#112 a Bad-with-null update keeps the last-known value. Nothing in ScadaBridge reads `ServiceLevel`, ScadaBridge#112 a Bad-with-null update keeps the last-known value. Nothing in ScadaBridge reads `ServiceLevel`,
`RedundancySupport` or `ServerUriArray`. `RedundancySupport` or `ServerUriArray`.
- **Consequence.** A ScadaBridge connection that lands on the Standby will sit on Bad quality indefinitely on a - **Consequence.** A ScadaBridge connection that lands on the Standby will sit on Bad quality indefinitely on a
healthy session, and never fail over. Today the same connection gets Good data at 240. Warm is therefore an healthy session, and never fail over. Today the same connection gets Good data at 240. Warm is therefore an
**outage for ScadaBridge** unless ScadaBridge becomes ServiceLevel-aware: monitor `i=2267` and treat a Degraded **outage for ScadaBridge** unless ScadaBridge becomes ServiceLevel-aware: monitor `i=2267` and treat a Degraded
band (< 200) sustained for N s, while the other endpoint reports a higher level, as a failover trigger. That band (< 200) sustained for N s, while the other endpoint reports a higher level, as a failover trigger. That
change must ship and be deployed first. The plan files it as a ScadaBridge issue (`cross-repo`, blocking). change must ship and be deployed first. It is filed as **ScadaBridge#138** (`cross-repo`, blocking); no tier sets
`Mode=Warm` until it is deployed there.
- **Alarms are compatible.** ScadaBridge already handles a RefreshStart/RefreshEnd replay as a snapshot - **Alarms are compatible.** ScadaBridge already handles a RefreshStart/RefreshEnd replay as a snapshot
(`RealOpcUaClient.cs:72`) and triggers its own ConditionRefresh on (re)subscribe (`:1179-1193`). Under D5 that (`RealOpcUaClient.cs:72`) and triggers its own ConditionRefresh on (re)subscribe (`:1179-1193`). Under D5 that
refresh returns retained conditions on the Standby too. The promotion refresh arrives as an unsolicited refresh returns retained conditions on the Standby too. The promotion refresh arrives as an unsolicited
Start/End pair on existing items; this needs a ScadaBridge-side test. Start/End pair on existing items; this needs a ScadaBridge-side test.
- **Status mapping is compatible.** ScadaBridge#117 maps by severity bits, so `BadNoCommunication` is Bad. - **Status mapping is compatible.** ScadaBridge#117 maps by severity bits, so `BadOutOfService` is Bad.
## 6. Configuration surface ## 6. Configuration surface
| Key | Default | Notes | | Key | Default | Notes |
|---|---|---| |---|---|---|
| `OpcUa:Redundancy:Mode` | `Unset` | `Unset` / `None` / `Warm`. `Hot` and unknown values fail host start. Startup-static. Flip **both** pair nodes in one window, after both run a revision-capable build (§4.2). | | `OpcUa:Redundancy:Mode` | `Unset` | `Unset` / `None` / `Warm`. `Hot` and unknown values fail host start. Startup-static. Flip **both** pair nodes in one window, after both run a revision-capable build (§4.2). |
| `OpcUa:Redundancy:PromoteWhilePrimaryUnreachable` | `true` | Warm only. Lets the survivor go Active while the Primary is unreachable but not yet downed (§4.1, owner answer Q2). |
| `OpcUa:PeerApplicationUris` | `[]` | Existing key, now also the `ServerUriArray` peer source. Required (≥ 1) for `Warm`, must be empty for `None`. | | `OpcUa:PeerApplicationUris` | `[]` | Existing key, now also the `ServerUriArray` peer source. Required (≥ 1) for `Warm`, must be empty for `None`. |
| `OpcUa:ApplicationUri` | `urn:OtOpcUa` | Must be unique within the pair under `Warm` (validated against the effective config). | | `OpcUa:ApplicationUri` | `urn:OtOpcUa` | Must be unique within the pair under `Warm` (validated against the effective config). |
@@ -269,22 +281,20 @@ There are no ConfigDb schema changes. `ServerCluster.RedundancyMode` is only cro
- Sequence-number wrap. - Sequence-number wrap.
- Encoded-byte fidelity of cloned payloads beyond the plan's bounded matrix (c7890). - Encoded-byte fidelity of cloned payloads beyond the plan's bounded matrix (c7890).
## 9. Open questions for the owner (short) ## 9. Owner answers (2026-11-13, #717)
1. **Standby writes.** Keep today's optimistic Good + revert (D7), or refuse synchronously with `BadInvalidState` The draft carried six open questions. The owner answered them on 2026-11-13 and approved the spec and plan
under Warm, as HistoryUpdate and native ack already do? *Recommend synchronous refusal under Warm only.* (`2be43aa9`). The answers are applied in the sections above.
2. **Crash-failover gap.** A SIGKILLed Primary stays an unreachable member until auto-down, ~25–30 s. Under (2)/(3)
the survivor stays Standby for that whole window, with no Good data anywhere, where today the Secondary keeps 1. **Standby writes.** Refuse synchronously with `BadInvalidState` under `Warm` only (§4.4). `Unset`/`None` keep the
serving. Should the survivor be allowed to promote when the Primary is *unreachable* (the #672 pattern, default #5 optimistic Good + revert.
on), accepting dual-Active during a real partition? *Recommend yes, behind 2. **Crash-failover gap.** Yes: the survivor may promote while the Primary is unreachable, behind
`OpcUa:Redundancy:PromoteWhilePrimaryUnreachable`.* `OpcUa:Redundancy:PromoteWhilePrimaryUnreachable` (default on) (§4.1). Dual-Active during a real partition is accepted.
3. **Standby status code.** `BadNoCommunication` is spike-proven and is plain Bad in ScadaBridge, but it is the 3. **Standby status code.** `BadOutOfService`, not the spike's `BadNoCommunication`, so a standby is distinguishable
same code as real device comms loss; only ServiceLevel tells them apart. Accept it, or use `BadOutOfService`? from real device comms loss without reading ServiceLevel. Re-proven on 2.0.1 in T0.
4. **Scripted Part 9 methods on the Standby.** They are accepted (D7) but their event is suppressed (D3), so the 4. **Scripted Part 9 methods on the Standby.** Refuse with `BadInvalidState` under `Warm` (§4.4).
caller sees Good and no event until promotion or refresh. Accept, or refuse with `BadInvalidState` under Warm? 5. **ConfigDb `Hot`.** Leave the CHECK; rely on the §3 mismatch Error. No migration.
5. **ConfigDb `Hot`.** Narrow the CHECK to `None/Warm` with a migration, or leave it and rely on the §3 mismatch 6. **Bands.** Accepted as drafted: Standby 90 (healthy) / 50 (degraded), boot 1 (§4.3).
Error?
6. **Bands.** Are Standby 90/50 and boot 1 acceptable (§4.3)?
## 10. What must be re-proven on SDK 2.0.1 ## 10. What must be re-proven on SDK 2.0.1
@@ -1,7 +1,9 @@
# Warm non-transparent redundancy (Gitea #717): implementation plan # Warm non-transparent redundancy (Gitea #717): implementation plan
Status: **DRAFT for owner approval**, together with `2026-11-11-warm-redundancy-717-design.md` (referred to below Status: **APPROVED by the owner 2026-11-13** (#717), together with `2026-11-11-warm-redundancy-717-design.md`
as "the spec", with § references). No task starts until both documents and the §9 open questions are settled. (referred to below as "the spec", with § references). The spec's §9 records the owner's answers: writes and Part 9
methods are refused `BadInvalidState` under Warm, promote-while-unreachable is on by default, the standby status is
`BadOutOfService`, there is no ConfigDb migration, and the bands are as drafted.
Every task ships **dark**: `OpcUa:Redundancy:Mode` defaults to `Unset`, so merging any task changes no deployed Every task ships **dark**: `OpcUa:Redundancy:Mode` defaults to `Unset`, so merging any task changes no deployed
behaviour. behaviour.
@@ -44,7 +46,7 @@ D4 itself stays an owner decision.
**Files:** **Files:**
- `src/Server/ZB.MOM.WW.OtOpcUa.OpcUaServer/OpcUaApplicationHost.cs`: options at `:29-121`. Add - `src/Server/ZB.MOM.WW.OtOpcUa.OpcUaServer/OpcUaApplicationHost.cs`: options at `:29-121`. Add
`OpcUaRedundancyOptions Redundancy` (`Mode`; `PromoteWhilePrimaryUnreachable` if Q2 = yes). `OpcUaRedundancyOptions Redundancy` (`Mode`; `PromoteWhilePrimaryUnreachable`, default `true`, owner answer Q2).
- `src/Server/ZB.MOM.WW.OtOpcUa.Host/Configuration/OpcUaApplicationHostOptionsValidator.cs`. - `src/Server/ZB.MOM.WW.OtOpcUa.Host/Configuration/OpcUaApplicationHostOptionsValidator.cs`.
- A post-load check in `StartAsync` (`:218-250`) against the effective `ApplicationConfiguration.ApplicationUri`. - A post-load check in `StartAsync` (`:218-250`) against the effective `ApplicationConfiguration.ApplicationUri`.
@@ -103,7 +105,7 @@ D4 itself stays an owner decision.
- `src/Core/ZB.MOM.WW.OtOpcUa.Cluster/Redundancy/ServiceLevelCalculator.cs`: add `ComputeWarm(inputs, isActive)`. - `src/Core/ZB.MOM.WW.OtOpcUa.Cluster/Redundancy/ServiceLevelCalculator.cs`: add `ComputeWarm(inputs, isActive)`.
`Compute` itself is unchanged. `Compute` itself is unchanged.
- `OpcUaPublishActor.RecomputeServiceLevel` (`:1576-1636`) branches on mode. It subscribes to `MemberUp/Removed` - `OpcUaPublishActor.RecomputeServiceLevel` (`:1576-1636`) branches on mode. It subscribes to `MemberUp/Removed`
(and to reachability if Q2 = yes) so the verdict re-derives on membership changes, not only on snapshots. and to reachability (owner answer Q2) so the verdict re-derives on membership changes, not only on snapshots.
**Tests:** **Tests:**
- A truth table covering every row of §4.3 and the invariant *max(Standby) < min(Active serving)*. - A truth table covering every row of §4.3 and the invariant *max(Standby) < min(Active serving)*.
@@ -132,9 +134,13 @@ D4 itself stays an owner decision.
- `src/Server/ZB.MOM.WW.OtOpcUa.OpcUaServer/EquipmentVariableState.cs`: the hook decided in T0b, reading a - `src/Server/ZB.MOM.WW.OtOpcUa.OpcUaServer/EquipmentVariableState.cs`: the hook decided in T0b, reading a
node-manager-wide volatile projection flag. Stored `Value` is never modified. node-manager-wide volatile projection flag. Stored `Value` is never modified.
- `OtOpcUaNodeManager` implements `IStandbyProjection.SetDataProjection`. - `OtOpcUaNodeManager` implements `IStandbyProjection.SetDataProjection`.
- Owner answer Q1: under Warm, the ordinary-tag write handler refuses a write on the Standby with `BadInvalidState`
before it reaches the node-write gateway. `Unset`/`None` keep the #5 optimistic path.
**Tests (wire):** **Tests (wire):**
- Standby Read of a raw tag, a UNS reference, a virtual tag and a Calculation tag returns `BadNoCommunication`/null. - Warm Standby: a Write returns `BadInvalidState` and the gateway is never called; on the Active node the same Write
reaches the gateway; under `Unset` a node that is not Primary still takes today's optimistic path.
- Standby Read of a raw tag, a UNS reference, a virtual tag and a Calculation tag returns `BadOutOfService`/null.
- A Standby Read of a condition child, `i=2267` or `i=3709` stays Good. - A Standby Read of a condition child, `i=2267` or `i=3709` stays Good.
- A Read with an IndexRange on an array tag is projected; a DataEncoding request is projected. - A Read with an IndexRange on an array tag is projected; a DataEncoding request is projected.
- Acquisition continues: flip to Active and the latest value is read without any new driver publish. - Acquisition continues: flip to Active and the latest value is read without any new driver publish.
@@ -150,7 +156,7 @@ D4 itself stays an owner decision.
disturbing the ACL filtering or `_liveItems`. disturbing the ACL filtering or `_liveItems`.
**Tests (wire):** port the spikes' assertion matrix to product code. **Tests (wire):** port the spikes' assertion matrix to product code.
- Queued Good samples, then demotion, then the first Publish carries `BadNoCommunication`. - Queued Good samples, then demotion, then the first Publish carries `BadOutOfService`.
- Prebuilt sequences, then demotion: the first delivery is projected and a later Republish returns the same - Prebuilt sequences, then demotion: the first delivery is projected and a later Republish returns the same
projected payload. projected payload.
- Good delivered while Active, then demotion, then Republish returns the original Good. - Good delivered while Active, then demotion, then Republish returns the original Good.
@@ -169,7 +175,7 @@ D4 itself stays an owner decision.
disabled-mode-drops-queue lesson applies. ACL-suppressed items are skipped. disabled-mode-drops-queue lesson applies. ACL-suppressed items are skipped.
**Tests (wire):** **Tests (wire):**
- Demotion delivers exactly one `BadNoCommunication` per item, then silence while values keep changing internally. - Demotion delivers exactly one `BadOutOfService` per item, then silence while values keep changing internally.
- Promotion delivers the current Good value without a driver publish. - Promotion delivers the current Good value without a driver publish.
- An ACL-stood-down item receives nothing either way. - An ACL-stood-down item receives nothing either way.
- A DeadbandFilter item still receives the promotion value. - A DeadbandFilter item still receives the promotion value.
@@ -182,8 +188,12 @@ D4 itself stays an owner decision.
- `IStandbyProjection.RefreshAllEventItems()` runs the existing `ConditionRefresh` (`:6586`) Server-object path and - `IStandbyProjection.RefreshAllEventItems()` runs the existing `ConditionRefresh` (`:6586`) Server-object path and
`RefreshDirectItems` (`:6974`) over every live event item, from `_allEventsItems` and `_eventItems`. `RefreshDirectItems` (`:6974`) over every live event item, from `_allEventsItems` and `_eventItems`.
- `ReportToServerObject` (`:6060`) is untouched. - `ReportToServerObject` (`:6060`) is untouched.
- Owner answer Q4: under Warm, every Part 9 condition method (native or scripted) is refused `BadInvalidState` on
the Standby before it changes condition state.
**Tests (wire, ACL off and on):** **Tests (wire, ACL off and on):**
- Warm Standby: Acknowledge and a shelving method on a scripted condition return `BadInvalidState` and route no
command; the Active node accepts them; `Unset` keeps today's scripted-method acceptance.
- A new alarm on the Standby reaches no Server-object, folder or direct item. - A new alarm on the Standby reaches no Server-object, folder or direct item.
- Model-change and audit events are still delivered on the Standby. - Model-change and audit events are still delivered on the Standby.
- A client ConditionRefresh **and** ConditionRefresh2 on the Standby return Start, retained, End (D5). - A client ConditionRefresh **and** ConditionRefresh2 on the Standby return Start, retained, End (D5).
@@ -210,7 +220,7 @@ D4 itself stays an owner decision.
## T11: ScadaBridge prerequisite (cross-repo) ## T11: ScadaBridge prerequisite (cross-repo)
- File a ScadaBridge issue (`type:feature`, `cross-repo`, `priority:p1`, linked to #717). The DCL becomes - **Filed as ScadaBridge#138** (2026-11-13). A ScadaBridge issue (`type:feature`, `cross-repo`, `priority:p1`, linked to #717). The DCL becomes
ServiceLevel-aware: monitor `i=2267`, and fail over when it is below 200 for a configurable window while the ServiceLevel-aware: monitor `i=2267`, and fail over when it is below 200 for a configurable window while the
other endpoint reports higher. Also add a test for an unsolicited RefreshStart/End on existing alarm items other endpoint reports higher. Also add a test for an unsolicited RefreshStart/End on existing alarm items
(spec §5). (spec §5).
@@ -245,14 +255,14 @@ everything yourself. Use `otopcua-cli` from `src/Client/ZB.MOM.WW.OtOpcUa.Client
|---|---|---| |---|---|---|
| L1 metadata | `redundancy` on :4840 and :4841 | `RedundancySupport=Warm`. ServerUriArray is self first and differs per node. Same URIs in both. | | L1 metadata | `redundancy` on :4840 and :4841 | `RedundancySupport=Warm`. ServerUriArray is self first and differs per node. Same URIs in both. |
| L2 bands | `read -n i=2267` on both | Primary 250, Standby 90 | | L2 bands | `read -n i=2267` on both | Primary 250, Standby 90 |
| L3 standby data | `read` and `subscribe` of a raw tag and its UNS reference on the Standby | `BadNoCommunication`/null; the Primary shows Good at the same moment | | L3 standby data | `read` and `subscribe` of a raw tag and its UNS reference on the Standby | `BadOutOfService`/null; the Primary shows Good at the same moment |
| L4 standby alarms | `alarms` on both. Trip a scripted alarm via its source tag. | Event on the Primary only. A ConditionRefresh on the Standby lists it as retained (D5). | | L4 standby alarms | `alarms` on both. Trip a scripted alarm via its source tag. | Event on the Primary only. A ConditionRefresh on the Standby lists it as retained (D5). |
| L5 manual failover | AdminUI manual failover (`docs/Redundancy.md` §Manual failover). Keep the CLI subscriptions open on both. | Old Primary: SL drops to 90, then `BadNoCommunication`, with no Good delivered after the drop. New Primary: Good values, then RefreshStart / retained / RefreshEnd, then SL 250, in that order (timestamps). | | L5 manual failover | AdminUI manual failover (`docs/Redundancy.md` §Manual failover). Keep the CLI subscriptions open on both. | Old Primary: SL drops to 90, then `BadOutOfService`, with no Good delivered after the drop. New Primary: Good values, then RefreshStart / retained / RefreshEnd, then SL 250, in that order (timestamps). |
| L6 crash | `docker kill central-<primary>` | The survivor goes Active within the gap Q2 selects. Record the measured gap. | | L6 crash | `docker kill central-<primary>` | The survivor goes Active once the Primary is unreachable (`PromoteWhilePrimaryUnreachable`, owner answer Q2). Record the measured gap. |
| L7 rejoin | Restart the killed node | It comes back Standby (90) and never shows 250 or 255. The boot read is 1 or 90. | | L7 rejoin | Restart the killed node | It comes back Standby (90) and never shows 250 or 255. The boot read is 1 or 90. |
| L8 standalone | site-b-2 (or a one-node overlay) with `Mode=None` | `RedundancySupport=None`, no `i=11314`, data Good | | L8 standalone | site-b-2 (or a one-node overlay) with `Mode=None` | `RedundancySupport=None`, no `i=11314`, data Good |
| L9 Unset control | Another node left on `Unset` | Byte-identical to pre-change: SL 240/250, SDK-default `i=3709` | | L9 Unset control | Another node left on `Unset` | Byte-identical to pre-change: SL 240/250, SDK-default `i=3709` |
| L10 writes/history | Write and HistoryRead on the Standby | Behaviour per Q1. HistoryRead is Good. | | L10 writes/history | Write and HistoryRead on the Standby | The write is refused `BadInvalidState` (owner answer Q1); HistoryRead is Good. |
Record evidence in `docs/plans/2026-11-xx-warm-redundancy-717-live-gate.md`. Do **not** mark G8/G9 (#615); they Record evidence in `docs/plans/2026-11-xx-warm-redundancy-717-live-gate.md`. Do **not** mark G8/G9 (#615); they
need CTT and a production client (Ignition 8.3 / ScadaBridge after T11). need CTT and a production client (Ignition 8.3 / ScadaBridge after T11).