dohertj2/lmxopcua
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
731092595fe78d77880c98ec7f89d4fa3370df74
lmxopcua/service_info.md

684 lines
40 KiB
Markdown
Raw Blame History

# Service Update Summary
Updated service instance: `C:\publish\lmxopcua\instance1`
Update time: `2026-03-25 12:54-12:55 America/New_York`
Backup created before deploy: `C:\publish\lmxopcua\backups\20260325-125444`
Configuration preserved:
- `C:\publish\lmxopcua\instance1\appsettings.json` was not overwritten.
Deployed binary:
- `C:\publish\lmxopcua\instance1\ZB.MOM.WW.LmxOpcUa.Host.exe`
- Last write time: `2026-03-25 12:53:58`
- Size: `143360`
Windows service:
- Name: `LmxOpcUa`
- Display name: `LMX OPC UA Server`
- Account: `LocalSystem`
- Status after update: `Running`
- Process ID after restart: `29236`
Restart evidence:
- Service log file: `C:\publish\lmxopcua\instance1\logs\lmxopcua-20260325_004.log`
- Last startup line: `2026-03-25 12:55:08.619 -04:00 [INF] The LmxOpcUa service was started.`
## CLI Verification
Endpoint from deployed config:
- `opc.tcp://localhost:4840/LmxOpcUa`
CLI used:
- `C:\Users\dohertj2\Desktop\lmxopcua\tools\opcuacli-dotnet\bin\Debug\net10.0\opcuacli-dotnet.exe`
Commands run:
```powershell
opcuacli-dotnet.exe connect -u opc.tcp://localhost:4840/LmxOpcUa
opcuacli-dotnet.exe read -u opc.tcp://localhost:4840/LmxOpcUa -n 'ns=1;s=MESReceiver_001.MoveInPartNumbers'
opcuacli-dotnet.exe read -u opc.tcp://localhost:4840/LmxOpcUa -n 'ns=1;s=MESReceiver_001.MoveInPartNumbers[]'
```
Observed results:
- `connect`: succeeded, server reported as `LmxOpcUa`.
- `read ns=1;s=MESReceiver_001.MoveInPartNumbers`: succeeded with good status `0x00000000`.
- `read ns=1;s=MESReceiver_001.MoveInPartNumbers[]`: failed with `BadNodeIdUnknown` (`0x80340000`).
---
## Instance 2 (Redundant Secondary)
Deployed: `2026-03-28`
Deployment path: `C:\publish\lmxopcua\instance2`
Configuration:
- `OpcUa.Port`: `4841`
- `OpcUa.ServerName`: `LmxOpcUa2`
- `OpcUa.ApplicationUri`: `urn:localhost:LmxOpcUa:instance2`
- `Dashboard.Port`: `8082`
- `MxAccess.ClientName`: `LmxOpcUa2`
- `Redundancy.Enabled`: `true`
- `Redundancy.Mode`: `Warm`
- `Redundancy.Role`: `Secondary`
- `Redundancy.ServerUris`: `["urn:localhost:LmxOpcUa:instance1", "urn:localhost:LmxOpcUa:instance2"]`
Windows service:
- Name: `LmxOpcUa2`
- Display name: `LMX OPC UA Server (Instance 2)`
- Account: `LocalSystem`
- Endpoint: `opc.tcp://localhost:4841/LmxOpcUa`
Instance 1 redundancy update (same date):
- `OpcUa.ApplicationUri`: `urn:localhost:LmxOpcUa:instance1`
- `Redundancy.Enabled`: `true`
- `Redundancy.Mode`: `Warm`
- `Redundancy.Role`: `Primary`
- `Redundancy.ServerUris`: `["urn:localhost:LmxOpcUa:instance1", "urn:localhost:LmxOpcUa:instance2"]`
CLI verification:
```
opcuacli-dotnet.exe redundancy -u opc.tcp://localhost:4840/LmxOpcUa
→ Redundancy Mode: Warm, Service Level: 200, Application URI: urn:localhost:LmxOpcUa:instance1
opcuacli-dotnet.exe redundancy -u opc.tcp://localhost:4841/LmxOpcUa
→ Redundancy Mode: Warm, Service Level: 150, Application URI: urn:localhost:LmxOpcUa:instance2
```
Both instances report the same `ServerUriArray` and expose the same Galaxy namespace (`urn:ZB:LmxOpcUa`).
## LDAP Authentication Update
Updated: `2026-03-28`
Both instances updated to use LDAP authentication via GLAuth.
Configuration changes (both instances):
- `Authentication.AllowAnonymous`: `true` (anonymous can browse/read)
- `Authentication.AnonymousCanWrite`: `false` (anonymous writes blocked)
- `Authentication.Ldap.Enabled`: `true`
- `Authentication.Ldap.Host`: `localhost`
- `Authentication.Ldap.Port`: `3893`
- `Authentication.Ldap.BaseDN`: `dc=lmxopcua,dc=local`
LDAP server: GLAuth v2.4.0 at `C:\publish\glauth\` (Windows service: `GLAuth`)
Permission verification (instance1, port 4840):
```
anonymous read → allowed
anonymous write → denied (BadUserAccessDenied)
readonly read → allowed
readonly write → denied (BadUserAccessDenied)
readwrite write → allowed
admin write → allowed
alarmack write → denied (BadUserAccessDenied)
bad password → denied (connection rejected)
```
## Alarm Notifier Chain Update
Updated: `2026-03-28`
Both instances updated with alarm event propagation up the notifier chain.
Code changes:
- Alarm events now walk up the parent chain (`ReportEventUpNotifierChain`), reporting to every ancestor node
- `EventNotifier = SubscribeToEvents` is set on all ancestors of alarm-containing nodes (`EnableEventNotifierUpChain`)
- Removed separate `Server.ReportEvent` call (no longer needed — the walk reaches the root)
No configuration changes required — alarm tracking was already enabled (`AlarmTrackingEnabled: true`).
Verification (instance1, port 4840):
```
alarms --node TestArea --refresh:
TestMachine_001.TestAlarm001 → visible (Severity=500, Retain=True)
TestMachine_001.TestAlarm002 → visible (Severity=500, Retain=True)
TestMachine_001.TestAlarm003 → visible (Severity=500, Retain=True)
TestMachine_002.TestAlarm001 → visible (Severity=500, Retain=True)
TestMachine_002.TestAlarm003 → visible (Severity=500, Retain=True)
alarms --node DEV --refresh:
Same 5 alarms visible at DEV (grandparent) level
```
## Auth Consolidation Update
Updated: `2026-03-28`
Both instances updated to consolidate LDAP roles into OPC UA session roles (`RoleBasedIdentity.GrantedRoleIds`).
Code changes:
- LDAP groups now map to custom OPC UA role NodeIds in `urn:zbmom:lmxopcua:roles` namespace
- Roles stored on session identity via `GrantedRoleIds` — no username-to-role side cache
- Permission checks use `GrantedRoleIds.Contains()` instead of username extraction
- `AnonymousCanWrite` behavior is consistent regardless of LDAP state
- Galaxy namespace moved from `ns=2` to `ns=3` (roles namespace is `ns=2`)
No configuration changes required.
Verification (instance1, port 4840):
```
anonymous read → allowed
anonymous write → denied (BadUserAccessDenied, AnonymousCanWrite=false)
readonly write → denied (BadUserAccessDenied)
readwrite write → allowed
admin write → allowed
alarmack write → denied (BadUserAccessDenied)
bad password → rejected (connection failed)
```
## Granular Write Roles Update
Updated: `2026-03-28`
Both instances updated with granular write roles replacing the single ReadWrite role.
Code changes:
- `ReadWrite` role replaced by `WriteOperate`, `WriteTune`, `WriteConfigure`
- Write permission checks now consider the Galaxy security classification of the target attribute
- `SecurityClassification` stored in `TagMetadata` for per-node lookup at write time
GLAuth changes:
- New groups: `WriteOperate` (5502), `WriteTune` (5504), `WriteConfigure` (5505)
- New users: `writeop`, `writetune`, `writeconfig`
- `admin` user added to all groups (5502, 5503, 5504, 5505)
Config changes (both instances):
- `Authentication.Ldap.ReadWriteGroup` replaced by `WriteOperateGroup`, `WriteTuneGroup`, `WriteConfigureGroup`
Verification (instance1, port 4840, Operate-classified attributes):
```
anonymous read → allowed
anonymous write → denied (AnonymousCanWrite=false)
readonly write → denied (no write role)
writeop write → allowed (WriteOperate matches Operate classification)
writetune write → denied (WriteTune doesn't match Operate)
writeconfig write → denied (WriteConfigure doesn't match Operate)
admin write → allowed (has all write roles)
```
## Historian SDK Migration
Updated: `2026-04-06`
Both instances updated to use the Wonderware Historian SDK (`aahClientManaged.dll`) instead of direct SQL queries for historical data access.
Code changes:
- `HistorianDataSource` rewritten from `SqlConnection`/`SqlDataReader` to `ArchestrA.HistorianAccess` SDK
- Persistent connection with lazy connect and auto-reconnect on failure
- `HistorianConfiguration.ConnectionString` replaced with `ServerName`, `IntegratedSecurity`, `UserName`, `Password`, `Port`
- `HistorianDataSource` now implements `IDisposable`, disposed on service shutdown
- `ConfigurationValidator` validates Historian SDK settings at startup
SDK DLLs deployed to both instances:
- `aahClientManaged.dll` (primary SDK, v2.0.0.0)
- `aahClient.dll`, `aahClientCommon.dll` (dependencies)
- `Historian.CBE.dll`, `Historian.DPAPI.dll`, `ArchestrA.CloudHistorian.Contract.dll`
Configuration changes (both instances):
- `Historian.ConnectionString` removed
- `Historian.ServerName`: `"localhost"`
- `Historian.IntegratedSecurity`: `true`
- `Historian.Port`: `32568`
- `Historian.Enabled`: `true` (unchanged)
Verification (instance1 startup log):
```
Historian.Enabled=true, ServerName=localhost, IntegratedSecurity=true, Port=32568
Historian.CommandTimeoutSeconds=30, MaxValuesPerRead=10000
=== Configuration Valid ===
LmxOpcUa service started successfully
```
## HistoryServerCapabilities and Continuation Points
Updated: `2026-04-06`
Both instances updated with OPC UA Part 11 spec compliance improvements.
Code changes:
- `HistoryServerCapabilities` node populated under `ServerCapabilities` with all boolean capability properties
- `AggregateFunctions` folder populated with references to 7 supported aggregate functions
- `HistoryContinuationPointManager` added — stores remaining data when results exceed `NumValuesPerNode`
- `HistoryReadRawModified` and `HistoryReadProcessed` now return `ContinuationPoint` in `HistoryReadResult` for partial reads
- Follow-up requests with `ContinuationPoint` resume from stored state; invalid/expired points return `BadContinuationPointInvalid`
No configuration changes required.
Verification (instance1 startup log):
```
HistoryServerCapabilities configured with 7 aggregate functions
LmxOpcUa service started successfully
```
## Remaining Historian Gaps Fix
Updated: `2026-04-06`
Both instances updated with remaining OPC UA Part 11 spec compliance fixes.
Code changes:
- **Gap 4**: `HistoryReadRawModified` returns `BadHistoryOperationUnsupported` when `IsReadModified=true`
- **Gap 5**: `HistoryReadAtTime` override added with `ReadAtTimeAsync` using SDK `HistorianRetrievalMode.Interpolated`
- **Gap 8**: `HistoricalDataConfigurationState` child nodes added to historized variables (`Stepped=false`, `Definition="Wonderware Historian"`)
- **Gap 10**: `ReturnBounds` parameter handled — boundary `DataValue` entries with `BadBoundNotFound` inserted at StartTime/EndTime
- **Gap 11**: `StandardDeviation` aggregate added to client enum, mapper, CLI (aliases: `stddev`/`stdev`), and UI dropdown
No configuration changes required.
## Historical Event Access
Updated: `2026-04-06`
Both instances updated with OPC UA historical event access (Gap 7).
Code changes:
- `HistorianDataSource.ReadEventsAsync` queries Historian event store via separate `HistorianConnectionType.Event` connection
- `LmxNodeManager.HistoryReadEvents` override maps `HistorianEvent` records to OPC UA `HistoryEventFieldList` entries
- `AccessHistoryEventsCapability` set to `true` when `AlarmTrackingEnabled` is true
- Event fields: EventId, EventType, SourceNode, SourceName, Time, ReceiveTime, Message, Severity
No configuration changes required. All historian gaps (1-11) are now resolved.
## Data Access Gaps Fix
Updated: `2026-04-06`
Both instances updated with OPC UA DA spec compliance fixes.
Code changes:
- `ConfigureServerCapabilities()` populates `ServerCapabilities` node: `ServerProfileArray`, `LocaleIdArray`, `MinSupportedSampleRate`, continuation point limits, array/string limits, and 12 `OperationLimits` values
- `Server_ServerDiagnostics_EnabledFlag` set to `true` — SDK auto-tracks session/subscription counts
- `OnModifyMonitoredItemsComplete` override logs monitored item modifications
No configuration changes required. All DA gaps (1-8) resolved.
## Alarms & Conditions Gaps Fix
Updated: `2026-04-06`
Both instances updated with OPC UA Part 9 alarm spec compliance fixes.
Code changes:
- Wired `OnConfirm`, `OnAddComment`, `OnEnableDisable`, `OnShelve`, `OnTimedUnshelve` handlers on each `AlarmConditionState`
- Shelving: `SetShelvingState()` manages `TimedShelve`, `OneShotShelve`, `Unshelve` state machine
- `ReportAlarmEvent` now populates `LocalTime` (timezone offset + DST) and `Quality` event fields
- Flaky `Monitor_ProbeDataChange_PreventsStaleReconnect` test fixed (increased stale threshold from 2s to 5s)
No configuration changes required. All A&C gaps (1-10) resolved.
## Security Gaps Fix
Updated: `2026-04-06`
Both instances updated with OPC UA Part 2/4/7 security spec compliance fixes.
Code changes:
- `SecurityProfileResolver`: Added 4 modern AES profiles (`Aes128_Sha256_RsaOaep-Sign/SignAndEncrypt`, `Aes256_Sha256_RsaPss-Sign/SignAndEncrypt`)
- `OnImpersonateUser`: Added `X509IdentityToken` handling with CN extraction and role assignment
- `BuildUserTokenPolicies`: Advertises `UserTokenType.Certificate` when non-None security profiles are configured
- `OnCertificateValidation`: Enhanced logging with certificate thumbprint, subject, and expiry
- Authentication audit logging: `AUDIT:` prefixed log entries for success/failure with session ID and roles
No configuration changes required. All security gaps (1-10) resolved.
## Historian Plugin Runtime Load + Dashboard Health
Updated: `2026-04-12 18:47-18:49 America/New_York`
Both instances updated to the latest build. Brings in the runtime-loaded Historian plugin (`Historian/` subfolder next to the Host) and the status dashboard health surface for historian plugin + alarm-tracking misconfiguration.
Backups created before deploy:
- `C:\publish\lmxopcua\backups\20260412-184713-instance1`
- `C:\publish\lmxopcua\backups\20260412-184713-instance2`
Configuration preserved:
- `C:\publish\lmxopcua\instance1\appsettings.json` was not overwritten.
- `C:\publish\lmxopcua\instance2\appsettings.json` was not overwritten.
Layout change:
- Flat historian interop DLLs removed from each instance root (`aahClient*.dll`, `ArchestrA.CloudHistorian.Contract.dll`, `Historian.CBE.dll`, `Historian.DPAPI.dll`).
- Historian plugin + interop DLLs now live under `<instance>\Historian\` (including `ZB.MOM.WW.LmxOpcUa.Historian.Aveva.dll`), loaded by `HistorianPluginLoader`.
Deployed binary (both instances):
- `ZB.MOM.WW.LmxOpcUa.Host.exe`
- Last write time: `2026-04-12 18:46:22 -04:00`
- Size: `7938048`
Windows services:
- `LmxOpcUa` — Running, PID `40176`
- `LmxOpcUa2` — Running, PID `34400`
Restart evidence (instance1 `logs/lmxopcua-20260412.log`):
```
2026-04-12 18:48:02.968 -04:00 [INF] Historian.Enabled=true, ServerName=localhost, IntegratedSecurity=true, Port=32568
2026-04-12 18:48:02.971 -04:00 [INF] === Configuration Valid ===
2026-04-12 18:48:09.658 -04:00 [INF] Historian plugin loaded from C:\publish\lmxopcua\instance1\Historian\ZB.MOM.WW.LmxOpcUa.Historian.Aveva.dll
2026-04-12 18:48:13.691 -04:00 [INF] LmxOpcUa service started successfully
```
Restart evidence (instance2 `logs/lmxopcua-20260412.log`):
```
2026-04-12 18:49:08.152 -04:00 [INF] Historian.Enabled=true, ServerName=localhost, IntegratedSecurity=true, Port=32568
2026-04-12 18:49:08.155 -04:00 [INF] === Configuration Valid ===
2026-04-12 18:49:14.744 -04:00 [INF] Historian plugin loaded from C:\publish\lmxopcua\instance2\Historian\ZB.MOM.WW.LmxOpcUa.Historian.Aveva.dll
2026-04-12 18:49:18.777 -04:00 [INF] LmxOpcUa service started successfully
```
CLI verification (via `dotnet run --project src/ZB.MOM.WW.LmxOpcUa.Client.CLI`):
```
connect opc.tcp://localhost:4840/LmxOpcUa → Server: LmxOpcUa
connect opc.tcp://localhost:4841/LmxOpcUa → Server: LmxOpcUa2
redundancy opc.tcp://localhost:4840/LmxOpcUa → Warm, ServiceLevel=200, urn:localhost:LmxOpcUa:instance1
redundancy opc.tcp://localhost:4841/LmxOpcUa → Warm, ServiceLevel=150, urn:localhost:LmxOpcUa:instance2
```
Both instances report the same `ServerUriArray` and the primary advertises the higher ServiceLevel, matching the prior redundancy baseline.
## Endpoints Panel on Dashboard
Updated: `2026-04-13 08:46-08:50 America/New_York`
Both instances updated with a new `Endpoints` panel on the status dashboard surfacing the opc.tcp base addresses, active OPC UA security profiles (mode + policy name + full URI), and user token policies.
Code changes:
- `StatusData.cs` — added `EndpointsInfo` / `SecurityProfileInfo` DTOs on `StatusData`.
- `OpcUaServerHost.cs` — added `BaseAddresses`, `SecurityPolicies`, `UserTokenPolicies` runtime accessors reading `ApplicationConfiguration.ServerConfiguration` live state.
- `StatusReportService.cs` — builds `EndpointsInfo` from the host and renders a new panel with a graceful empty state when the server is not started.
No configuration changes required.
Verification (instance1 @ `http://localhost:8085/`):
```
Base Addresses: opc.tcp://localhost:4840/LmxOpcUa
Security Profiles: None / None / http://opcfoundation.org/UA/SecurityPolicy#None
User Token Policies: Anonymous, UserName
```
Verification (instance2 @ `http://localhost:8086/`):
```
Base Addresses: opc.tcp://localhost:4841/LmxOpcUa
Security Profiles: None / None / http://opcfoundation.org/UA/SecurityPolicy#None
User Token Policies: Anonymous, UserName
```
## Template-Based Alarm Object Filter
Updated: `2026-04-13 09:39-09:43 America/New_York`
Both instances updated with a new configurable alarm object filter. When `OpcUa.AlarmFilter.ObjectFilters` is non-empty, only Galaxy objects whose template derivation chain matches a pattern (and their containment-tree descendants) contribute `AlarmConditionState` nodes. When the list is empty, the current unfiltered behavior is preserved (backward-compatible default).
Backups created before deploy:
- `C:\publish\lmxopcua\backups\20260413-093900-instance1`
- `C:\publish\lmxopcua\backups\20260413-093900-instance2`
Deployed binary (both instances):
- `ZB.MOM.WW.LmxOpcUa.Host.exe`
- Last write time: `2026-04-13 09:38:46 -04:00`
- Size: `7951360`
Windows services:
- `LmxOpcUa` — Running, PID `40900`
- `LmxOpcUa2` — Running, PID `29936`
Code changes:
- `gr/queries/hierarchy.sql` — added recursive CTE on `gobject.derived_from_gobject_id` and a new `template_chain` column (pipe-delimited, innermost template first).
- `Domain/GalaxyObjectInfo.cs` — added `TemplateChain: List<string>` populated from the new SQL column.
- `GalaxyRepositoryService.cs` — reads the new column and splits into `TemplateChain`.
- `Configuration/AlarmFilterConfiguration.cs` (new) — `List<string> ObjectFilters`; entries may themselves be comma-separated. Attached to `OpcUaConfiguration.AlarmFilter`.
- `Configuration/ConfigurationValidator.cs` — logs the effective filter and warns if patterns are configured while `AlarmTrackingEnabled == false`.
- `Domain/AlarmObjectFilter.cs` (new) — compiles wildcard patterns (`*` only) to case-insensitive regexes with Galaxy `$` prefix normalized on both sides; walks the hierarchy top-down with cycle defense; returns a `HashSet<int>` of included gobject IDs plus `UnmatchedPatterns` for startup warnings.
- `OpcUa/LmxNodeManager.cs` — constructor accepts the filter; the two alarm-creation loops (`BuildAddressSpace` full build and the subtree rebuild path) both call `ResolveAlarmFilterIncludedIds(sorted)` and skip any object not in the resolved set. New public properties expose filter state to the dashboard: `AlarmFilterEnabled`, `AlarmFilterPatternCount`, `AlarmFilterIncludedObjectCount`.
- `OpcUa/OpcUaServerHost.cs`, `OpcUa/LmxOpcUaServer.cs`, `OpcUaService.cs`, `OpcUaServiceBuilder.cs` — plumbing to construct and thread the filter from `appsettings.json` down to the node manager.
- `Status/StatusData.cs` + `Status/StatusReportService.cs` — `AlarmStatusInfo` gains `FilterEnabled`, `FilterPatternCount`, `FilterIncludedObjectCount`; a filter summary line renders in the Alarms panel when the filter is active.
Tests:
- 36 new unit tests in `tests/.../Domain/AlarmObjectFilterTests.cs` covering pattern parsing, wildcard semantics, regex escaping, Galaxy `$` normalization, template-chain matching, subtree propagation, set semantics, orphan/cycle defense, and `UnmatchedPatterns` tracking.
- 5 new integration tests in `tests/.../Integration/AlarmObjectFilterIntegrationTests.cs` spinning up a real `LmxNodeManager` via `OpcUaServerFixture` and asserting `AlarmConditionCount`/`AlarmFilterIncludedObjectCount` under various filters.
- 1 new Status test verifying JSON exposes the filter counters.
- Full suite: **446/446 tests passing** (no regressions).
Configuration change: both instances have `OpcUa.AlarmFilter.ObjectFilters: []` (filter disabled, unfiltered alarm tracking preserved).
Live verification against instance1 Galaxy (filter temporarily set to `"TestMachine"`):
```
2026-04-13 09:41:31 [INF] OpcUa.AlarmTrackingEnabled=true, AlarmFilter.ObjectFilters=[TestMachine]
2026-04-13 09:41:42 [INF] Alarm filter: 42 of 49 objects included (1 pattern(s))
Dashboard Alarms panel: Tracking: True | Conditions: 60 | Active: 4
Filter: 1 pattern(s), 42 object(s) included
```
Final configuration restored to empty filter. Dashboard confirms unfiltered behavior on both endpoints:
```
instance1 @ http://localhost:8085/ → Conditions: 60 | Active: 4 (no filter line)
instance2 @ http://localhost:8086/ → Conditions: 60 | Active: 4 (no filter line)
```
Filter syntax quick reference (documented in `AlarmFilterConfiguration.cs` XML-doc):
- `*` is the only wildcard (glob-style; zero or more characters).
- Matching is case-insensitive and ignores the Galaxy leading `$` template prefix on both the pattern and the stored chain entry, so operators write `TestMachine*` not `$TestMachine*`.
- Each entry may contain comma-separated patterns for convenience (e.g., `"TestMachine*, Pump_*"`).
- Empty list → filter disabled → current unfiltered behavior.
- Match semantics: an object is included when any template in its derivation chain matches any pattern, and the inclusion propagates to all descendants in the containment hierarchy. Each object is evaluated once regardless of how many patterns or ancestors match.
## Historian Runtime Health Surface
Updated: `2026-04-13 10:44-10:52 America/New_York`
Both instances updated with runtime historian query instrumentation so the status dashboard can detect silent query degradation that the load-time `PluginStatus` cannot catch.
Backups:
- `C:\publish\lmxopcua\backups\20260413-104406-instance1`
- `C:\publish\lmxopcua\backups\20260413-104406-instance2`
Code changes:
- `Host/Historian/HistorianHealthSnapshot.cs` (new) — DTO with `TotalQueries`, `TotalSuccesses`, `TotalFailures`, `ConsecutiveFailures`, `LastSuccessTime`, `LastFailureTime`, `LastError`, `ProcessConnectionOpen`, `EventConnectionOpen`.
- `Host/Historian/IHistorianDataSource.cs` — added `GetHealthSnapshot()` interface method.
- `Historian.Aveva/HistorianDataSource.cs` — added `_healthLock`-guarded counters, `RecordSuccess()` / `RecordFailure(path)` helpers called at every terminal site in all four read methods (raw, aggregate, at-time, events). Error messages carry a `raw:` / `aggregate:` / `at-time:` / `events:` prefix so operators can tell which SDK call is broken.
- `Host/OpcUa/LmxNodeManager.cs` — exposes `HistorianHealth` property that proxies to `IHistorianDataSource.GetHealthSnapshot()`.
- `Host/Status/StatusData.cs` — added 9 new fields on `HistorianStatusInfo`.
- `Host/Status/StatusReportService.cs` — `BuildHistorianStatusInfo()` populates the new fields from the node manager; panel color gradient: green → yellow (1-4 consecutive failures) → red (≥5 consecutive or plugin unloaded). Renders `Queries: N (Success: X, Failure: Y) | Consecutive Failures: Z`, `Process Conn: open/closed | Event Conn: open/closed`, plus `Last Success:` / `Last Failure:` / `Last Error:` lines when applicable.
- `Host/Status/HealthCheckService.cs` — new Rule 2b2: `Degraded` when `ConsecutiveFailures >= 3`. Threshold chosen to avoid flagging single transient blips.
Tests:
- 5 new unit tests in `HistorianDataSourceLifecycleTests` covering fresh zero-state, single failure, multi-failure consecutive increment, cross-read-path counting, and error-message-carries-path.
- Full suite: 16/16 plugin tests, 447/447 host tests passing.
Live verification on instance1:
```
Before any query:
Queries: 0 (Success: 0, Failure: 0) | Process Conn: closed | Event Conn: closed
After TestMachine_001.TestHistoryValue raw read:
Queries: 1 (Success: 1, Failure: 0) | Process Conn: open
Last Success: 2026-04-13T14:45:18Z
After aggregate hourly-average over 24h:
Queries: 2 (Success: 2, Failure: 0)
After historyread against an unknown node id (bad tag):
Queries: 2 (counter unchanged — rejected at node-lookup before reaching the plugin; correct)
```
JSON endpoint `/api/status` carries all 9 new fields with correct types. Both instances deployed; instance1 `LmxOpcUa` PID 33824, instance2 `LmxOpcUa2` PID 30200.
## Historian Read-Only Cluster Support
Updated: `2026-04-13 11:25-12:00 America/New_York`
Both instances updated with Wonderware Historian read-only cluster failover. Operators can supply an ordered list of historian cluster nodes; the plugin iterates them on each fresh connect and benches failed nodes for a configurable cooldown window. Single-node deployments are preserved via the existing `ServerName` field.
Backups:
- `C:\publish\lmxopcua\backups\20260413-112519-instance1`
- `C:\publish\lmxopcua\backups\20260413-112519-instance2`
Code changes:
- `Host/Configuration/HistorianConfiguration.cs` — added `ServerNames: List<string>` (defaults to `[]`) and `FailureCooldownSeconds: int` (defaults to 60). `ServerName` preserved as fallback when `ServerNames` is empty.
- `Host/Historian/HistorianClusterNodeState.cs` (new) — per-node DTO: `Name`, `IsHealthy`, `CooldownUntil`, `FailureCount`, `LastError`, `LastFailureTime`.
- `Host/Historian/HistorianHealthSnapshot.cs` — extended with `ActiveProcessNode`, `ActiveEventNode`, `NodeCount`, `HealthyNodeCount`, `Nodes: List<HistorianClusterNodeState>`.
- `Historian.Aveva/HistorianClusterEndpointPicker.cs` (new, internal) — pure picker with injected clock, thread-safe via lock, BFS-style `GetHealthyNodes()` / `MarkFailed()` / `MarkHealthy()` / `SnapshotNodeStates()`. Nodes iterate in configuration order; failed nodes skip until cooldown elapses; the cumulative `FailureCount` and `LastError` are retained across recovery for operator diagnostics.
- `Historian.Aveva/HistorianDataSource.cs` — new `ConnectToAnyHealthyNode(type)` method iterates picker candidates, clones `HistorianConfiguration` per attempt with the candidate as `ServerName`, and returns the first successful `(Connection, Node)` tuple. `EnsureConnected` and `EnsureEventConnected` both call it. `HandleConnectionError` and `HandleEventConnectionError` now mark the active node failed in the picker before nulling. `_activeProcessNode` / `_activeEventNode` track the live node for the dashboard. Both silos (process + event) share a single picker instance so a node failure on one immediately benches it for the other.
- `Host/Status/StatusData.cs` — added `NodeCount`, `HealthyNodeCount`, `ActiveProcessNode`, `ActiveEventNode`, `Nodes` to `HistorianStatusInfo`.
- `Host/Status/StatusReportService.cs` — Historian panel renders `Process Conn: open (<node>)` badges and a cluster table (when `NodeCount > 1`) showing each node's state, cooldown expiry, failure count, and last error. Single-node deployments render a compact `Node: <hostname>` line.
- `Host/Status/HealthCheckService.cs` — new Rule 2b3: `Degraded` when `NodeCount > 1 && HealthyNodeCount < NodeCount`. Lets operators alert on a partially-failed cluster even while queries are still succeeding via the remaining nodes.
- `Host/Configuration/ConfigurationValidator.cs` — logs the effective node list and `FailureCooldownSeconds` at startup, validates that `FailureCooldownSeconds >= 0`, warns when `ServerName` is set alongside a non-empty `ServerNames`.
Tests:
- `HistorianClusterEndpointPickerTests.cs` — 19 unit tests covering config parsing, ordered iteration, cooldown expiry, zero-cooldown mode, mark-healthy clears, cumulative failure counting, unknown-node safety, concurrent writers (thread-safety smoke test).
- `HistorianClusterFailoverTests.cs` — 6 integration tests driving `HistorianDataSource` via a scripted `FakeHistorianConnectionFactory`: first-node-fails-picks-second, all-nodes-fail, second-call-skips-cooled-down-node, single-node-legacy-behavior, picker-order-respected, shared-picker-across-silos.
- Full plugin suite: 41/41 tests passing. Host suite: 446/447 (1 pre-existing flaky MxAccess monitor test passes on retry).
Live verification on instance1 (cluster = `["does-not-exist-historian.invalid", "localhost"]`, `FailureCooldownSeconds=30`):
**Failover cycle 1** (fresh picker state, both nodes healthy):
```
2026-04-13 11:27:25.381 [WRN] Historian node does-not-exist-historian.invalid failed during connect attempt; trying next candidate
2026-04-13 11:27:25.910 [INF] Historian SDK connection opened to localhost:32568
```
- historyread returned 1 value successfully (`Queries: 1 (Success: 1, Failure: 0)`).
- Dashboard: panel yellow, `Cluster: 1 of 2 nodes healthy`, bad node `cooldown` until `11:27:55Z`, `Process Conn: open (localhost)`.
**Cooldown expiry**:
- At 11:29 UTC, the cooldown window had elapsed. Panel back to green, both nodes healthy, but `does-not-exist-historian.invalid` retains `FailureCount=1` and `LastError` as history.
**Failover cycle 2** (service restart to drop persistent connection):
```
2026-04-13 14:00:39.352 [WRN] Historian node does-not-exist-historian.invalid failed during connect attempt; trying next candidate
2026-04-13 14:00:39.885 [INF] Historian SDK connection opened to localhost:32568
```
- historyread returned 1 value successfully on the second restart cycle — proves the picker re-admits a cooled-down node and the whole failover cycle repeats cleanly.
**Single-node restoration**:
- Changed instance1 back to `"ServerNames": []`, restarted. Dashboard renders `Node: localhost` (no cluster table), panel green, backward compat verified.
Final configuration: both instances running with empty `ServerNames` (single-node mode). `LmxOpcUa` PID 31064, `LmxOpcUa2` PID 15012.
Operator configuration shape:
```json
"Historian": {
"Enabled": true,
"ServerName": "localhost", // ignored when ServerNames is non-empty
"ServerNames": ["historian-a", "historian-b"],
"FailureCooldownSeconds": 60,
...
}
```
## Galaxy Runtime Status Probes + Subtree Quality Invalidation
Updated: `2026-04-13 15:28-16:19 America/New_York`
Both instances updated with per-host Galaxy runtime status tracking ($WinPlatform + $AppEngine), proactive subtree quality invalidation when a host transitions to Stopped, and an OPC UA Read short-circuit so operators can no longer read stale-Good cached values from a dead runtime host.
This ships the feature described in the `runtimestatus.md` plan file. Addresses the production issue reported earlier: "when an AppEngine is set to scan off, LMX updates are received for every tag, causing OPC UA client freeze and sometimes not all OPC UA tags are set to bad quality."
Backups:
- `C:\publish\lmxopcua\backups\20260413-152824-instance1`
- `C:\publish\lmxopcua\backups\20260413-152824-instance2`
Deployed binary (both instances):
- `ZB.MOM.WW.LmxOpcUa.Host.exe` — commit `98ed6bd`
- Two incremental deploys during verification: 15:28 (initial), 15:52 (Read-handler patch), 16:06 (dispatch-thread deadlock fix)
Windows services:
- `LmxOpcUa` — Running, PID `29528`
- `LmxOpcUa2` — Running, PID `30684`
### Code changes — what shipped
**New config** — `MxAccessConfiguration`:
- `RuntimeStatusProbesEnabled: bool` (default `true`) — enables `<Host>.ScanState` probing for every deployed `$WinPlatform` and `$AppEngine`.
- `RuntimeStatusUnknownTimeoutSeconds: int` (default `15`) — only applies to the Unknown → Stopped transition; running hosts never time out because `ScanState` is delivered on-change only.
**New hierarchy columns** — `hierarchy.sql` and `GalaxyObjectInfo`:
- `CategoryId: int` — populated from `template_definition.category_id` (1 = $WinPlatform, 3 = $AppEngine).
- `HostedByGobjectId: int` — populated from `gobject.hosted_by_gobject_id` (the actual column name on this Galaxy schema; the plan document's guess of `host_gobject_id` was wrong). Walked up to find each variable's nearest Platform/Engine ancestor.
**New domain types** — `Host/Domain/`:
- `GalaxyRuntimeState` enum (`Unknown` / `Running` / `Stopped`).
- `GalaxyRuntimeStatus` DTO with callback/state-change timestamps, `LastScanState`, `LastError`, cumulative counters.
**New probe manager** — `Host/MxAccess/GalaxyRuntimeProbeManager.cs`:
- Pure manager, no SDK leakage. `AdviseSupervisory`s `<Host>.ScanState` for every runtime host on `SyncAsync`.
- State predicate: `isRunning = vtq.Quality.IsGood() && vtq.Value is bool b && b`. Everything else is Stopped.
- `GetSnapshot()` forces every entry to `Unknown` when the MxAccess transport is disconnected — prevents misleading "every host stopped" display when the actual problem is the transport.
- `Tick()` only advances Unknown → Stopped on the configured timeout; Running hosts never time out (on-change delivery semantic).
- `IsHostStopped(gobjectId)` — used by the Read-path short-circuit; uses underlying state directly (not the snapshot force-unknown rewrite) so a transport outage doesn't double-flag reads.
- `Dispose()` unadvises every active probe before MxAccess teardown.
**New hosted-variables map** — `LmxNodeManager`:
- `_hostedVariables: Dictionary<int, List<BaseDataVariableState>>` — host gobject_id → list of every descendant variable, populated during `BuildAddressSpace` by walking each variable's `HostedByGobjectId` chain up to the nearest Platform/Engine. A variable hosted by an Engine inside a Platform appears in BOTH lists.
- `_hostIdsByTagRef: Dictionary<string, List<int>>` — reverse index used by the Read short-circuit, populated alongside `_hostedVariables`.
- Public `MarkHostVariablesBadQuality(int gobjectId)` — walks `_hostedVariables[gobjectId]`, sets `StatusCode = BadOutOfService` on each, calls `ClearChangeMasks(ctx, false)` to push through the OPC UA publisher.
- Public `ClearHostVariablesBadQuality(int gobjectId)` — inverse, resets to `Good` on recovery.
**OPC UA Read short-circuit** — `LmxNodeManager.Read`:
- Before the normal `_mxAccessClient.ReadAsync(tagRef)` round-trip, check `IsTagUnderStoppedHost(tagRef)`. If true, return a `DataValue { StatusCode = BadOutOfService, Value = cachedVar?.Value }` directly. Covers both direct Read requests AND OPC UA monitored-item sampling, which both flow through this override.
**Deadlock fix — `_pendingHostStateChanges` queue**:
- First draft invoked `MarkHostVariablesBadQuality` synchronously from the probe callback. MxAccess delivers `OnDataChange` on the STA thread; the callback took the node manager `Lock`. Meanwhile any worker thread inside `Read` could hold `Lock` and wait on a pending `ReadAsync` that needed the STA thread — **classic STA deadlock** (first real deploy hung in ~30s).
- Fix: probe transitions are enqueued on `ConcurrentQueue<(int GobjectId, bool Stopped)>` and the dispatch thread drains the queue inside its existing 100ms `WaitOne` loop. The dispatch thread takes `Lock` naturally without STA involvement, so no cycle. Live verified with the IDE OffScan/OnScan cycle after the fix.
**Dashboard** — `Host/Status/`:
- New `RuntimeStatusInfo` DTO + "Galaxy Runtime" panel between Galaxy Info and Historian. Shows total/running/stopped/unknown counts plus a per-host table with Name / Kind / State / Since / Last Error columns. Panel color: green (all Running), yellow (some Unknown, none Stopped), red (any Stopped), gray (MxAccess disconnected forces every row to Unknown).
- Subscriptions panel gets a new `Probes: N (bridge-owned runtime status)` line when non-zero.
- `HealthCheckService` Rule 2e: `Degraded` when any host is Stopped, ordered after Rule 1 (MxAccess transport) to avoid double-messaging when the transport is the root cause.
### Tests
- **24** new `GalaxyRuntimeProbeManagerTests`: state transitions (Unknown/Running/Stopped/recovery), unknown-resolution timeout, transport gating, sync diff, dispose, callback exception safety, `IsHostStopped` for Read-path short-circuit (Unknown/Running/Stopped/recovery/unknown-id/transport-disconnected-contract).
- Full Host suite: **471/471** tests passing. No regressions.
### Live end-to-end verification (today, against real IDE OffScan action)
**Baseline** (before OffScan, dashboard at 15:44:00):
```
Galaxy Runtime: green, 2 of 2 hosts running
DevAppEngine $AppEngine Running 2026-04-13T19:29:12.9475357Z
DevPlatform $WinPlatform Running 2026-04-13T19:29:12.9345208Z
TestMachine_001.MachineID → Status 0x00000000 (Good), value "admin_test"
```
**After operator Set OffScan on DevAppEngine in IDE** (log at 15:44:25):
```
15:44:25.554 Galaxy runtime DevAppEngine.ScanState transitioned Running → Stopped (ScanState = false (OffScan))
15:44:25.557 Marked 3971 variable(s) BadOutOfService for stopped host gobject_id=1043
```
Dashboard: red panel, `1 of 2 hosts running (1 stopped, 0 unknown)`. Health: `Degraded — Galaxy runtime has 1 of 2 host(s) stopped: DevAppEngine`. Critical: 3ms from probe callback to subtree walk complete.
**Read during stop — found bug #1** (Read handler bypassed cached state):
- Initial deploy: `TestMachine_001.MachineID` still read `0x00000000` Good with a post-stop source time from MxAccess. Revealed that `LmxNodeManager.Read` calls `_mxAccessClient.ReadAsync()` directly and never consults the in-memory `BaseDataVariableState.StatusCode` we set during the walk.
- Fix: `IsTagUnderStoppedHost` short-circuit in Read override. After patch: `[808D0000] BadOutOfService` on all three test tags.
**Read during stop — found bug #2** (deadlock):
- After shipping the Read patch, the service hung on the next OffScan. HTTP listener accepted connections but never responded, and service shutdown stuck at STOP_PENDING for 15+ seconds until manually killed.
- Diagnosis: the probe callback fires `HandleProbeUpdate` → `MarkHostVariablesBadQuality` → acquires `Lock` on the STA thread. Meanwhile the dispatch thread can sit inside `Read` holding `Lock` and waiting for an STA-routed `ReadAsync`. Circular wait.
- Fix: enqueue probe transitions onto `ConcurrentQueue` and drain on the dispatch thread where `Lock` acquisition is safe. Second deploy resolved the hang.
**A/B verification** (instance1 patched, instance2 not yet):
| Instance | `TestMachine_001.MachineID` |
|---|---|
| `LmxOpcUa` (patched) | `0x808D0000` BadOutOfService ✅ |
| `LmxOpcUa2` (old) | `0x00000000` Good, stale ❌ |
Clean A/B confirmed the Read patch is required; instance2 subsequently updated to match.
**Recovery** (operator Set OnScan on DevAppEngine, log at 16:10:05):
```
16:10:05.129 Galaxy runtime DevAppEngine.ScanState transitioned → Running
16:10:05.130 Cleared bad-quality override on 3971 variable(s) for recovered host gobject_id=1043
```
Dashboard: back to green, `DevAppEngine` Running with new `Since = 20:10:05.129Z`. All three test tags back to `0x00000000` Good with fresh source timestamps. 1ms from probe callback to subtree clear.
### Client freeze observation — phase 2 decision gate
The original production issue had two symptoms: (1) incomplete quality flip and (2) OPC UA client freeze. The subtree walk + Read short-circuit fixes (1) definitively. For (2), there's still a pending dispatch-queue flood of per-tag MxAccess callbacks that MxAccess fans out when a host stops — the bridge doesn't currently drop them. We **deliberately did not** ship dispatch suppression in this pass, on the grounds that the subtree walk may coalesce notifications sufficiently at the SDK publisher level to resolve the freeze on its own. The verification against the live Galaxy with no OPC UA clients subscribed doesn't tell us one way or the other — the next subscribed-client test against a real stop will be the deciding measurement. If the client still freezes after the walk, phase 2 adds pre-dispatch filtering for tags under Stopped hosts.
### What's deferred
- **Synthetic OPC UA child nodes** (`$RuntimeState`, `$LastCallbackTime`, etc.) under each host object. Dashboard + health surface give operators visibility today; the OPC UA synthetic nodes are a follow-up.
- **Dispatch suppression** — gated on observing whether the subtree walk alone resolves the client freeze in production.
- **Documentation updates** — the `docs/` guides (`MxAccessBridge.md`, `StatusDashboard.md`, `Configuration.md`, `HistoricalDataAccess.md`) still describe the pre-runtime-status behavior. Need a consolidated doc pass covering this feature plus the historian cluster + health surface updates from earlier today.
## Notes
The service deployment and restart succeeded. The live CLI checks confirm the endpoint is reachable and that the array node identifier has changed to the bracketless form. The array value on the live service still prints as blank even though the status is good, so if this environment should have populated `MoveInPartNumbers`, the runtime data path still needs follow-up investigation.
Reference in New Issue View Git Blame Copy Permalink