Merge #809 live PercentGood/DurationGood count Good only (Part 13), matching HistoryRead

Claude-Session: https://claude.ai/code/session_01GT67AD99zQQqpLCVm3p7xk
This commit is contained in:
Joseph Doherty
2026-11-11 12:05:51 -05:00
11 changed files with 421 additions and 126 deletions
+1 -1
View File
@@ -32,7 +32,7 @@ historian tagname, which is the tag's driver `FullName`.
### Pluggable backends, SQL Server and HistoryUpdate (2026-10)
- **Selection:** `ServerHistorian:Backend` (`Gateway` default | `SqlServer` | `QuestDb`; unknown fails host start). `HistorianBackendRegistry` → `IHistorianBackendProvider` builds all five seams; the gateway keys are only required for `Gateway`. `SqlServer` uses `ServerHistorian:SqlServer:{ConnectionString (required), Schema=hist, CommandTimeout=30s, AutoCreateSchema=true}`.
- **SQL Server** (`Driver.Historian.SqlServer`): `hist.Sample` PK `(TagId, SourceTimeUtc)` — writes are insert-if-absent, so outbox redelivery is idempotent. Stores every value kind with its **real status**; per-sample aggregates (10, simple definitions) are pushed down to T-SQL with bucket math in **100 ns ticks**; the Part 13 time-weighted seven (TimeAverage/TimeAverage2/Total/Total2/Interpolative, #671 — ⚠️ `Total` is now the time-weighted integral, no longer `SUM` — plus the **duration-based** PercentGood/PercentBad, #680, and DurationGood/DurationBad, #820 — ms, Uncertain time in neither; the live AggregateFilter still counts Uncertain as Good, #809) stream window + 4 bound rows into the shared `TimeWeightedAggregator` (both SQL backends, identical results; reverse `Interpolative` anchored at the later time; window capped by `ServerHistorian:MaxTimeWeightedRawRows` ⇒ `BadTooManyOperations`, percentages included). ⚠️ **#680 status rules (client-visible):** TimeAverage/Total are Uncertain if any part of the interval uses an Uncertain value, is bridged across a Bad sample, is extrapolated past the last sample (⇒ a live tag's latest interval is Uncertain) or is not covered; TimeAverage2/Total2 keep the SDK time-based rule (Uncertain time counts as Good) but never extrapolate past the last sample — that interval is Partial, later ones `BadNoData` (a report-by-exception tag with nothing after its last sample reads `BadNoData` until its next sample arrives). `docs/Historian.md` §"Time-weighted aggregates". ⚠️ **#680 per-sample (client-visible):** `Count` counts **Good** samples only (value-based status; an all-Bad interval is generic Bad, value null) and an empty interval **inside the tag's data range** is `0`/Good (outside it still `BadNoData`); `Start`/`End` return the **raw** first/last sample of any quality — its own timestamp and status, value null when Bad; `Delta` is Uncertain only when the interval's first or last sample is unusable; ⚠️ #684: `Average`/`Minimum`/`Maximum`/`Range`/`*ActualTime` use **Good samples only** (was Good + Uncertain; an Uncertain-only interval is `BadNoData`), Min/Max-family status Uncertain on a Bad sample, an Uncertain value beyond the Good extreme (Part 13 — the SDK ignores that) or an Uncertain non-numeric sample, and only a **Good** String/DateTime sample refuses them (`BadAggregateInvalidInputs`; Delta still refuses any usable one); and every per-sample aggregate but `Average` sets **end-of-data Partial** on the interval holding the last sample (forward incl. its earlier edge, reverse strictly inside), never on a Bad result. The data-range lookup is a key seek on SQL Server and a `tag` symbol-index seek on QuestDB (#683 — `{p}sample.tag` is `SYMBOL INDEX`; the bootstrapper ADDs it to a pre-#683 table, idempotent), read BEFORE the bucket statement so a racing append can never yield a Good `Count` of 0 (a concurrent HistoryUpdate delete is the mirror case and still can, transiently). `docs/Historian.md` §"Aggregates". Docker-dev recipe, schema and aggregate semantics: `docs/Historian.md` §"SQL Server backend". Integration tests need `HISTORIAN_SQL_TEST_CONNECTION` (shared host `10.100.0.35,14330`, `sa`/`OtOpcUaDev_2026!` — not docker-dev's `OtOpcUa!Dev123`; QuestDB: `HISTORIAN_QUESTDB_TEST_CONNECTION` against the shared `/opt/otopcua-questdb` fixture `10.100.0.35:8812`, #822, `infra/README.md` §2a). **Retention (#667):** opt-in `SqlServer:RetentionDays` (+ `PurgeInterval` 1h, `PurgeBatchSize` 4000) -> `HistorianRetentionService` purges old `hist.Sample` (per tag) / `hist.Event` in `DELETE TOP(n)` batches, Primary-gated on `IHistorizationRoleView`, one purger per DB via `sp_getapplock`; counter `otopcua.historian.retention.purged{table}`. Optional `IHistorianBackendProvider.CreateRetention` (default null). `docs/Historian.md` §Retention.
- **SQL Server** (`Driver.Historian.SqlServer`): `hist.Sample` PK `(TagId, SourceTimeUtc)` — writes are insert-if-absent, so outbox redelivery is idempotent. Stores every value kind with its **real status**; per-sample aggregates (10, simple definitions) are pushed down to T-SQL with bucket math in **100 ns ticks**; the Part 13 time-weighted seven (TimeAverage/TimeAverage2/Total/Total2/Interpolative, #671 — ⚠️ `Total` is now the time-weighted integral, no longer `SUM` — plus the **duration-based** PercentGood/PercentBad, #680, and DurationGood/DurationBad, #820 — ms, Uncertain time in neither; since #809 the live AggregateFilter agrees — `GoodOnlyStatusAggregateCalculator` overrides only those two of the frozen calculators) stream window + 4 bound rows into the shared `TimeWeightedAggregator` (both SQL backends, identical results; reverse `Interpolative` anchored at the later time; window capped by `ServerHistorian:MaxTimeWeightedRawRows` ⇒ `BadTooManyOperations`, percentages included). ⚠️ **#680 status rules (client-visible):** TimeAverage/Total are Uncertain if any part of the interval uses an Uncertain value, is bridged across a Bad sample, is extrapolated past the last sample (⇒ a live tag's latest interval is Uncertain) or is not covered; TimeAverage2/Total2 keep the SDK time-based rule (Uncertain time counts as Good) but never extrapolate past the last sample — that interval is Partial, later ones `BadNoData` (a report-by-exception tag with nothing after its last sample reads `BadNoData` until its next sample arrives). `docs/Historian.md` §"Time-weighted aggregates". ⚠️ **#680 per-sample (client-visible):** `Count` counts **Good** samples only (value-based status; an all-Bad interval is generic Bad, value null) and an empty interval **inside the tag's data range** is `0`/Good (outside it still `BadNoData`); `Start`/`End` return the **raw** first/last sample of any quality — its own timestamp and status, value null when Bad; `Delta` is Uncertain only when the interval's first or last sample is unusable; ⚠️ #684: `Average`/`Minimum`/`Maximum`/`Range`/`*ActualTime` use **Good samples only** (was Good + Uncertain; an Uncertain-only interval is `BadNoData`), Min/Max-family status Uncertain on a Bad sample, an Uncertain value beyond the Good extreme (Part 13 — the SDK ignores that) or an Uncertain non-numeric sample, and only a **Good** String/DateTime sample refuses them (`BadAggregateInvalidInputs`; Delta still refuses any usable one); and every per-sample aggregate but `Average` sets **end-of-data Partial** on the interval holding the last sample (forward incl. its earlier edge, reverse strictly inside), never on a Bad result. The data-range lookup is a key seek on SQL Server and a `tag` symbol-index seek on QuestDB (#683 — `{p}sample.tag` is `SYMBOL INDEX`; the bootstrapper ADDs it to a pre-#683 table, idempotent), read BEFORE the bucket statement so a racing append can never yield a Good `Count` of 0 (a concurrent HistoryUpdate delete is the mirror case and still can, transiently). `docs/Historian.md` §"Aggregates". Docker-dev recipe, schema and aggregate semantics: `docs/Historian.md` §"SQL Server backend". Integration tests need `HISTORIAN_SQL_TEST_CONNECTION` (shared host `10.100.0.35,14330`, `sa`/`OtOpcUaDev_2026!` — not docker-dev's `OtOpcUa!Dev123`; QuestDB: `HISTORIAN_QUESTDB_TEST_CONNECTION` against the shared `/opt/otopcua-questdb` fixture `10.100.0.35:8812`, #822, `infra/README.md` §2a). **Retention (#667):** opt-in `SqlServer:RetentionDays` (+ `PurgeInterval` 1h, `PurgeBatchSize` 4000) -> `HistorianRetentionService` purges old `hist.Sample` (per tag) / `hist.Event` in `DELETE TOP(n)` batches, Primary-gated on `IHistorizationRoleView`, one purger per DB via `sp_getapplock`; counter `otopcua.historian.retention.purged{table}`. Optional `IHistorianBackendProvider.CreateRetention` (default null). `docs/Historian.md` §Retention.
- **QuestDB** (`Driver.Historian.QuestDb`, `ServerHistorian:QuestDb:*`, Npgsql over PG wire; docker-dev overlay `docker-compose.questdb.yml`): same kinds/status/aggregates/HistoryUpdate as SqlServer, soft delete by tombstone (INSERT only — never `UPDATE`/`DELETE`; every read filters `NOT deleted`). ⚠️ **Timestamps are µs** (`QuestDbConnection.Round`/`BindTimestamp`; `IHistorianDataSource.TimestampResolutionTicks` = 10 steps HistoryRead paging), **pre-1970 cannot be stored** (writer drops, HistoryUpdate `BadOutOfRange`), every write waits for WAL visibility, TTL is enforced on write commits only, set Npgsql `Maximum Pool Size` < QuestDB's 64. `docs/Historian.md` §"QuestDB backend".
- ⚠️ **Reverse processed reads are Part 13/SDK-anchored (#677, client-visible):** every aggregate on SqlServer/QuestDB now anchors a reverse read at its later time — window `(end, start]`, partial interval earliest, each stamped with its later edge (was: earlier-anchored, stamped earlier edge); on forward reads Partial also marks the interval the data starts inside (not Average — the SDK's rule, forward-only); a non-default client `AggregateConfiguration` ⇒ `BadAggregateConfigurationRejected`; Gateway provisioning sends `IntegralDivisor = 1`. SDK-parity tests: `SdkAggregateParityTests`. `docs/Historian.md` §Aggregates.
- ⚠️ **Reverse / end-only RAW reads work on Gateway (#686, client-visible):** `IHistorianDataSource.SupportsReverseRaw` (SqlServer/QuestDB/Null `true`, Gateway `false`, interface default `false`, guard-tested) — the node manager sends `start > end` only to a backend that declares it; otherwise `ReverseRawReader` serves it from FORWARD reads: uncapped ⇒ ONE forward read reversed in memory; capped ⇒ a bounded walk back in segments (ms-aligned request bounds, width aimed from observed density at ≈ N + 8 rows and carried page to page in the continuation point — #689: it used to aim at half `MaxTieClusterOverfetch`, ~32k rows per page on a dense tag; a read asks for `2N + 18` rows (a discovery segment ≤ 1,026), `MaxTieClusterOverfetch` is only the tie-cluster ceiling — a full read is re-read, never answered truncated; only more than max(1, `MaxTieClusterOverfetch`) samples within ONE MILLISECOND ⇒ `BadTooManyOperations` (per ms, not per timestamp — requests are ms-rounded, and letting a segment narrow below 1 ms stepped a page through one ms a tick per read, ~10k reads); walks no later than now + 1 day and never before the Unix epoch; #691: after two empty segments it asks once for the whole remainder, and the stretch a page proved empty rides the continuation point — the last pages of a deep page-through no longer each walk ~14 reads to 1970). The raw/events tie-path oversized-cluster refusal is now `BadTooManyOperations` too (was `BadHistoryOperationUnsupported`). Uncapped Gateway raw reads (forward too) used to send `max_values = 0`, which the client/gateway reject — now unbounded (`int.MaxValue`; a ceiling refusal ⇒ `BadTooManyOperations`), and a cancelled read is no longer a backend failure. Since #669 such reads failed the node on Gateway and counted as a backend fault (the gateway rejects start > end — `GatewayHistorianDataSource.ReadRawAsync` now refuses one unsent); before #669 an end-only Gateway read returned the OLDEST N ascending — now the latest N at/before EndTime, newest first. Events follow since #691.4: a reverse/end-only capped events read on Gateway (`SupportsReverseEvents` false; Null now reports `true` and answers empty) is walked back the same way (ties sorted by EventId — the gateway orders by time only; one extra event found ⇒ continuation point) — it used to be refused `BadHistoryOperationUnsupported` (explicit reverse) or answered with the OLDEST N ascending (end-only); needs the gateway's `RuntimeDb:EventReadMaxRows` 0/unlimited (it truncates silently). Uncapped Gateway events reads (forward too) used to throw "Array dimensions exceeded" (result list pre-sized to `int.MaxValue`) — fixed. Above a truncated read the walk gallops up from the newest row seen, then bisects (never below 1 ms): a burst that ended 100 days ago ≈ 18 reads, a dense tag that stopped days ago ≈ 26. `docs/Historian.md` §"Open-ended and reverse raw reads".
+9 -5
View File
@@ -1827,10 +1827,13 @@ is `Good` (plus Calculated / Partial). Value = `100 × region time ÷ interval d
**`DurationGood` / `DurationBad`** (#820) are the same Good / Bad region times as `PercentGood` / `PercentBad` — every rule above
(regions, simple bounds, time before the first sample = Bad, end of data, `BadNoData`, status, Partial) applies unchanged — reported as
an OPC UA `Duration`, i.e. a `double` in **milliseconds** (the SDK's unit): `DurationGood = PercentGood × interval ms ÷ 100`. Uncertain
time counts toward neither (Mantis 0006166). Before #820 both were `BadAggregateNotSupported` on the SQL backends. ⚠️ A **live**
`AggregateFilter` subscription computes DurationGood/PercentGood with the frozen 1.5.378.182 calculators, which count Uncertain time as
Good — they disagree with HistoryRead whenever Uncertain data is present (#809, owner decision pending). The Gateway backend still
refuses both (`BadAggregateNotSupported`, not advertised in i=11201).
time counts toward neither (Mantis 0006166). Before #820 both were `BadAggregateNotSupported` on the SQL backends. ⚠️ **#809
(client-visible, owner-approved 2026-11-11):** a **live** `AggregateFilter` subscription now computes DurationGood/PercentGood the same
way — Good time only. They used to come from the frozen 1.5.378.182 calculators, which count Uncertain time as Good; the server now
serves those two through the application's `GoodOnlyStatusAggregateCalculator` (every other live aggregate is still the frozen
baseline). A live PercentGood/DurationGood over Uncertain data therefore drops by the Uncertain share. Pinned live-vs-history by
`SdkAggregateParityTests.Live_AggregateFilter_PercentGood_and_DurationGood_equal_HistoryRead_over_Uncertain_samples`. The Gateway
backend still refuses both (`BadAggregateNotSupported`, not advertised in i=11201).
All carry the Calculated bit. A non-Bad result also carries the **Partial** bit (`0x4`, `AggregateBits.Partial`) — never on a Bad
result — for: the **truncated interval** (shorter than the processing interval — the latest one forward, the earliest one reverse;
@@ -1911,7 +1914,8 @@ registers for **live** `AggregateFilter` items and from SDK 2.0.1's own stock ca
`SdkAggregateParityTests.An_Uncertain_first_sample_in_a_late_start_interval_makes_the_SDK_integrate_the_time_before_the_data`
(ours / frozen / stock 2.0.1, with a Good-first control) and `OpcUaServer.IntegrationTests/LiveAggregateLateStartFactsTests` (through
the booted server's runtime `AggregateManager`), and on both SQL backends by `QualityMixConformance.Time_weighted_values_over_partial_data_cover_only_the_data`
(forward) and `…_in_a_reverse_read` (#821). Whether live subscriptions should follow is the open #809/#821 owner decision. The same
(forward) and `…_in_a_reverse_read` (#821). Whether live subscriptions should follow is the open #821 owner decision (#809, the
PercentGood/DurationGood half, is decided: live follows Part 13). The same
suite feeds PercentGood/PercentBad, DurationGood/DurationBad (and Count, Average, TimeAverage(2), Total(2), Start/End/Delta, Interpolative) Good/Uncertain/Bad
mixes, all-Uncertain intervals (0 % Good **and** 0 % Bad, Mantis 0006166) and partial-data intervals, hand-computed and against the
#796 oracle — the baseline calculator's Uncertain-as-Good PercentGood fails it.
@@ -6,10 +6,16 @@ namespace ZB.MOM.WW.OtOpcUa.OpcUaServer.Compatibility.Aggregates;
/// <summary>Adapts the versioned baseline algorithms to the SDK 2 calculator protocol without a second SDK runtime.</summary>
internal static class CompatibilityAggregateFactory
{
/// <remarks>
/// PercentGood / DurationGood are served by the application-owned <see cref="GoodOnlyStatusAggregateCalculator"/>
/// (#809: Uncertain time is not Good, as in Part 13 and HistoryRead); every other id is the unchanged baseline.
/// </remarks>
internal static IAggregateCalculator Create(NodeId id, DateTimeUtc start, DateTimeUtc end, double interval,
bool stepped, AggregateConfiguration configuration, ITelemetryContext telemetry) =>
new Adapter(Aggregators.CreateStandardCalculator(id, (DateTime)start, (DateTime)end, interval, stepped, configuration, telemetry)
?? throw new ArgumentOutOfRangeException(nameof(id), "Unsupported baseline aggregate."));
new Adapter(GoodOnlyStatusAggregateCalculator.Overrides(id)
? new GoodOnlyStatusAggregateCalculator(id, (DateTime)start, (DateTime)end, interval, stepped, configuration, telemetry)
: Aggregators.CreateStandardCalculator(id, (DateTime)start, (DateTime)end, interval, stepped, configuration, telemetry)
?? throw new ArgumentOutOfRangeException(nameof(id), "Unsupported baseline aggregate."));
/// <summary>Replaces exactly the existing standard factories; no aggregate IDs are added or removed.</summary>
internal static async ValueTask RegisterAsync(AggregateManager manager, CancellationToken ct)
@@ -0,0 +1,83 @@
using Opc.Ua;
using DataValue = ZB.MOM.WW.OtOpcUa.OpcUaServer.Compatibility.Aggregates.LegacyDataValue;
namespace ZB.MOM.WW.OtOpcUa.OpcUaServer.Compatibility.Aggregates;
/// <summary>
/// Application-owned override (#809) of the baseline <see cref="StatusAggregateCalculator"/> for
/// <c>PercentGood</c> and <c>DurationGood</c> only: a region counts toward the Good time only when its status is
/// Good. Uncertain time is in neither Good nor (unless <c>TreatUncertainAsBad</c>) Bad, per Part 13
/// §5.4.3.31/§5.4.3.33 (Mantis 0006166) — the rule the HistoryRead evaluators (<c>TimeWeightedAggregator</c>,
/// SqlServer + QuestDB) already follow.
/// </summary>
/// <remarks>
/// The baseline 1.5.378.182 <c>ComputeDurationGoodBad</c> counts Uncertain regions as Good when
/// <c>TreatUncertainAsBad</c> = false. This class leaves the copied file untouched and replaces only the Good
/// branch: the slice values, simple bounds, stepped regions, empty-slice <c>BadNoData</c>, timestamp, value type
/// and Calculated status bit are the baseline's. <c>DurationBad</c> / <c>PercentBad</c> / <c>WorstQuality*</c>
/// fall through to the baseline unchanged.
/// </remarks>
internal sealed class GoodOnlyStatusAggregateCalculator : StatusAggregateCalculator
{
/// <summary>True for the two aggregate ids this class overrides.</summary>
internal static bool Overrides(NodeId aggregateId) =>
aggregateId == ObjectIds.AggregateFunction_PercentGood || aggregateId == ObjectIds.AggregateFunction_DurationGood;
/// <summary>Initializes the calculator; same contract as the baseline constructor.</summary>
public GoodOnlyStatusAggregateCalculator(
NodeId aggregateId,
DateTime startTime,
DateTime endTime,
double processingInterval,
bool stepped,
AggregateConfiguration configuration,
ITelemetryContext? telemetry)
: base(aggregateId, startTime, endTime, processingInterval, stepped, configuration, telemetry)
{
}
/// <inheritdoc />
protected override DataValue ComputeValue(TimeSlice slice)
{
if (AggregateId == ObjectIds.AggregateFunction_DurationGood) return ComputeGoodOnly(slice, usePercent: false);
if (AggregateId == ObjectIds.AggregateFunction_PercentGood) return ComputeGoodOnly(slice, usePercent: true);
return base.ComputeValue(slice);
}
/// <summary>The baseline <c>ComputeDurationGoodBad(slice, isBad: false, usePercent)</c> with Good-only counting.</summary>
private DataValue ComputeGoodOnly(TimeSlice slice, bool usePercent)
{
List<DataValue>? values = GetValuesWithSimpleBounds(slice);
if (values == null || values.Count == 0)
{
return GetNoDataValue(slice);
}
List<SubRegion> regions = GetRegionsInValueSet(values, false, true);
double duration = 0;
double total = 0;
foreach (var region in regions)
{
total += region.Duration;
if (StatusCode.IsGood(region.StatusCode))
{
duration += region.Duration;
}
}
if (usePercent)
{
duration = duration / total * 100;
}
var value = new DataValue
{
WrappedValue = SdkValueBoundary.FromClr(duration),
SourceTimestamp = GetTimestamp(slice),
ServerTimestamp = GetTimestamp(slice)
};
value.StatusCode = value.StatusCode.WithAggregateBits(AggregateBits.Calculated);
return value;
}
}
@@ -17,3 +17,14 @@ The isolated `scripts/validation/opcua-aggregate-baseline` process pins only SDK
Run the generator with a **new** output path, then compare it with `tests/Server/ZB.MOM.WW.OtOpcUa.OpcUaServer.IntegrationTests/Fixtures/opcua-1.5.378.182-aggregates.json`. It refuses to overwrite evidence. Never regenerate expectations from the compatibility implementation. Existing historian parity tests retain their application-specific documented divergences from the baseline SDK, while new compatibility tests check the complete old-calculator output.
This preserves older behavior deliberately; it does not claim that all older SDK algorithms implement the newest OPC UA interpretation. A future semantic migration needs its own consumer-impact decision and versioned tests.
## Recorded semantic decisions
### #809 — PercentGood / DurationGood count Good time only (owner-approved 2026-11-11)
The baseline `StatusAggregateCalculator.ComputeDurationGoodBad` counts Uncertain regions as **Good** when `TreatUncertainAsBad = false` (1.5.378.106 and SDK 2.0.1 count Good only). Part 13 §5.4.3.31/§5.4.3.33 (Mantis 0006166) and the HistoryRead evaluators (`TimeWeightedAggregator`, SqlServer + QuestDB) count Good regions only, so live AggregateFilter results disagreed with HistoryRead whenever Uncertain data was present. ScadaBridge uses neither AggregateFilter nor these aggregates.
- The copied upstream files above are **unchanged** (their SHA-256s still match `baseline-sources.json`).
- `GoodOnlyStatusAggregateCalculator.cs` (application code, not a baseline copy) subclasses the baseline `StatusAggregateCalculator` and replaces only the PercentGood / DurationGood computation: the same slice values, simple bounds, stepped regions, empty-slice `BadNoData`, timestamps, value type and Calculated bit, with a region counted as Good only when its status is Good — whatever `TreatUncertainAsBad` says (when true, Uncertain is Bad, so it was never Good). `CompatibilityAggregateFactory.Create` routes exactly those two IDs to it; the other 35 IDs, including DurationBad / PercentBad / WorstQuality, are the baseline unchanged.
- Golden test: the 64 vectors this changes (38 PercentGood + 26 DurationGood, all with an effective `TreatUncertainAsBad = false` over data with Uncertain regions) are exempted **exactly** — the PercentGood / DurationGood vectors whose baseline output differs from the same case's baseline vector with `TreatUncertainAsBad = true`, where 1.5.378.182 already counted Good regions only — and each is pinned to that twin's output. The pin is therefore baseline evidence, never regenerated from the application's calculator; any further drift in any of the 6,144 cases fails. `Sdk2AggregateCompatibilityTests.Issue809_*` also pins a hand-computed case.
- Live-vs-history parity: `SdkAggregateParityTests.Live_AggregateFilter_PercentGood_and_DurationGood_equal_HistoryRead_over_Uncertain_samples` compares the registered calculator with `TimeWeightedAggregator` over Good / Uncertain / Bad mixes with no masks, and checks that the data discriminates against the pre-#809 baseline.
@@ -10,7 +10,9 @@ namespace ZB.MOM.WW.OtOpcUa.OpcUaServer;
/// <para>
/// The aggregate evaluators (<c>AggregateEvaluator</c> / <c>TimeWeightedAggregator</c> in
/// <c>Driver.Historian.Sql.Common</c>) are written against fixed settings, not parameterised: Uncertain time counts as
/// Good (<see cref="TreatUncertainAsBad"/> = false) in the time-weighted family; the per-sample Average / Minimum /
/// Good (<see cref="TreatUncertainAsBad"/> = false) in TimeAverage2 / Total2, while PercentGood / DurationGood count
/// Good time only and PercentBad / DurationBad Bad time only — Uncertain in neither (Part 13 §5.4.3.31-34, Mantis
/// 0006166); the per-sample Average / Minimum /
/// Maximum / Range / *ActualTime / Count are defined by Part 13 over Good values whatever this setting, as the
/// SDK computes them under it; an interval is Good only when all of
/// it is (<see cref="PercentDataGood"/> = 100) and Bad only when none of it is non-Bad
@@ -23,6 +25,16 @@ namespace ZB.MOM.WW.OtOpcUa.OpcUaServer;
/// the SDK manager so live AggregateFilter subscriptions agree with history reads; a test pins both sides.
/// </para>
/// <para>
/// The live calculators are the application's frozen 1.5.378.182 copies (<c>Compatibility/Aggregates</c>), whose
/// <c>ComputeDurationGoodBad</c> counts Uncertain time as Good under <see cref="TreatUncertainAsBad"/> = false. That
/// made live PercentGood / DurationGood disagree with HistoryRead whenever Uncertain data was present. #809
/// (owner-approved 2026-11-11) overrides exactly those two (<c>GoodOnlyStatusAggregateCalculator</c>), so live and
/// HistoryRead now agree on them — pinned by
/// <c>SdkAggregateParityTests.Live_AggregateFilter_PercentGood_and_DurationGood_equal_HistoryRead_over_Uncertain_samples</c>.
/// Other known live-vs-HistoryRead differences (e.g. #821, an Uncertain first sample in a late-start interval) are
/// documented in <c>docs/Historian.md</c>.
/// </para>
/// <para>
/// A request asking for <c>UseServerCapabilitiesDefaults</c>, or spelling out exactly these values, is served;
/// any other configuration is refused per node with <c>BadAggregateConfigurationRejected</c> rather than silently
/// answered under settings the client did not ask for (it used to be ignored).
@@ -71,7 +71,7 @@ public abstract class QualityMixConformance : ConformanceBase
/// <summary>Hand-computed Part 13 DurationBad of Q's eight intervals, seconds.</summary>
private static readonly double[] QDurationBad = [3, 0, 0, 0, 5, 0, 0, 3];
/// <summary>What the 1.5.378.182 baseline answers for Q's PercentGood (Uncertain counted as Good, #796).</summary>
/// <summary>What the untouched 1.5.378.182 baseline answers for Q's PercentGood (Uncertain counted as Good, #796; live before #809).</summary>
private static readonly double[] QBaselinePercentGood = [70, 100, 100, 100, 50, 100, 100, 70];
/// <summary>
@@ -182,7 +182,8 @@ public abstract class QualityMixConformance : ConformanceBase
var pGood = (await src.ReadProcessedAsync("Q", At(0), At(80), Ten, HistoryAggregateType.PercentGood, Ct)).Samples;
ShouldBeClose(Values(good), Values(pGood).Select(p => p * Ten.TotalMilliseconds / 100).ToArray(), "DurationGood vs PercentGood");
// The baseline (the live AggregateFilter calculator, #809) answers Q's DurationGood with the Uncertain time added.
// The untouched baseline (the live AggregateFilter calculator before #809) answers Q's DurationGood with the
// Uncertain time added; the live calculator now answers the Part 13 values above (SdkAggregateParityTests).
var baseline = SdkAggregateOracle.RunSdkAsIs(HistoryAggregateType.DurationGood, At(0), At(80), Ten,
Q.Select(r => (r.T, r.V, r.Status)));
ShouldBeClose(baseline.Select(r => Convert.ToDouble(r.Value)).ToArray(),
@@ -2,7 +2,7 @@ using Opc.Ua;
using Opc.Ua.Server;
using ZB.MOM.WW.OtOpcUa.OpcUaServer.Compatibility.Aggregates;
using BaselineAggregateCalculator = ZB.MOM.WW.OtOpcUa.OpcUaServer.Compatibility.Aggregates.AggregateCalculator;
using BaselineStatusAggregateCalculator = ZB.MOM.WW.OtOpcUa.OpcUaServer.Compatibility.Aggregates.StatusAggregateCalculator;
using BaselineAggregators = ZB.MOM.WW.OtOpcUa.OpcUaServer.Compatibility.Aggregates.Aggregators;
using Shouldly;
using ZB.MOM.WW.OtOpcUa.Core.Abstractions;
@@ -18,17 +18,18 @@ namespace ZB.MOM.WW.OtOpcUa.Driver.Historian.Conformance;
/// <remarks>
/// <para>Shared (source-linked) by <c>Sql.Common.Tests</c> and both SQL backends' integration suites.</para>
/// <para>
/// <b>PercentGood (and, since #820, DurationGood) are the aggregates not taken as-is (#796).</b> The 1.5.378.182 baseline changed
/// <b>PercentGood / DurationGood (#796, #820, #809).</b> The 1.5.378.182 baseline changed
/// <c>StatusAggregateCalculator.ComputeDurationGoodBad</c> (1.5.378.106 and earlier counted Good regions only):
/// with TreatUncertainAsBad = false it counts Uncertain regions as <b>Good</b> time (as Bad when true). Part 13
/// keeps Uncertain out of both (§5.4.3.31 DurationGood counts Good regions; §5.4.3.33 PercentGood =
/// DurationGood ÷ interval; Mantis 0006166, closed "no change required": an all-Uncertain interval is
/// PercentGood 0 <i>and</i> PercentBad 0), and so does ours. <see cref="Run"/> therefore answers PercentGood
/// (and DurationGood) through <see cref="Part13GoodCalculator"/>: the SAME baseline calculator, region machinery (simple
/// bounds, end-of-data 1 ms region, status and Partial bits) and server-default configuration, counting only
/// Good regions — exactly what the pre-.182 oracle computed. PercentBad / DurationBad still count Bad regions only
/// (unchanged by .182) and are compared as-is. <see cref="RunSdkAsIs"/> runs the baseline untouched, for the test that pins
/// the divergence.
/// PercentGood 0 <i>and</i> PercentBad 0), and so does ours. Since #809 the LIVE calculators follow Part 13 too:
/// <c>CompatibilityAggregateFactory</c> serves those two through the application's
/// <c>GoodOnlyStatusAggregateCalculator</c> (the same baseline region machinery — simple bounds, end-of-data 1 ms
/// region, status and Partial bits — counting only Good regions), so <see cref="Run"/> / <see cref="RunLive"/> are
/// exactly what a live AggregateFilter computes, for every aggregate. PercentBad / DurationBad count Bad regions only
/// (unchanged by .182). <see cref="RunSdkAsIs"/> runs the untouched baseline (what live served before #809), for the
/// tests that pin that divergence.
/// </para>
/// <para>
/// Not <c>100 − PercentBad</c> under TreatUncertainAsBad = true: that flag also changes the region machinery
@@ -54,8 +55,9 @@ public static class SdkAggregateOracle
};
/// <summary>
/// Runs the Part 13 oracle for <paramref name="aggregate"/> over <paramref name="samples"/>: the baseline calculator,
/// except PercentGood / DurationGood, which count Good regions only (see the remarks, #796/#820).
/// Runs the Part 13 oracle for <paramref name="aggregate"/> over <paramref name="samples"/>: the live calculator
/// (<see cref="RunLive"/>) — the baseline, except PercentGood / DurationGood, which count Good regions only (see the
/// remarks, #796/#820/#809).
/// </summary>
/// <param name="aggregate">The aggregate.</param>
/// <param name="start">Request start (later than <paramref name="end"/> for a reverse read).</param>
@@ -65,16 +67,29 @@ public static class SdkAggregateOracle
/// <returns>The oracle's processed values, in the order it returns them.</returns>
public static IReadOnlyList<Result> Run(HistoryAggregateType aggregate, DateTime start, DateTime end, TimeSpan interval,
IEnumerable<(DateTime Time, HistorianValue Value, uint Status)> samples) =>
aggregate is HistoryAggregateType.PercentGood or HistoryAggregateType.DurationGood
? RunCore(new LegacyCalculatorAdapter(new Part13GoodCalculator(AggregateId(aggregate),
start, end, interval.TotalMilliseconds, false, DefaultConfiguration, Telemetry,
usePercent: aggregate == HistoryAggregateType.PercentGood)), start, end, samples)
: RunSdkAsIs(aggregate, start, end, interval, samples);
RunLive(aggregate, start, end, interval, samples);
/// <summary>
/// Runs the baseline (1.5.378.182) calculator for <paramref name="aggregate"/> exactly as it computes it, under
/// <see cref="DefaultConfiguration"/> — including its PercentGood rule (Uncertain counted as Good) that
/// <see cref="Run"/> deliberately does not use (#796).
/// Runs exactly the calculator the server registers for live AggregateFilter monitored items
/// (<c>CompatibilityAggregateFactory.Create</c>, the #809 Good-only PercentGood / DurationGood included).
/// </summary>
/// <param name="aggregate">The aggregate.</param>
/// <param name="start">Request start (later than <paramref name="end"/> for a reverse read).</param>
/// <param name="end">Request end.</param>
/// <param name="interval">Processing interval.</param>
/// <param name="samples">Every stored sample of the tag, any order (sorted here).</param>
/// <param name="configuration">The configuration (default <see cref="DefaultConfiguration"/>).</param>
/// <returns>The calculator's processed values, in the order it returns them.</returns>
public static IReadOnlyList<Result> RunLive(HistoryAggregateType aggregate, DateTime start, DateTime end,
TimeSpan interval, IEnumerable<(DateTime Time, HistorianValue Value, uint Status)> samples,
AggregateConfiguration? configuration = null) =>
RunCore(CompatibilityAggregateFactory.Create(AggregateId(aggregate), start, end, interval.TotalMilliseconds, false,
configuration ?? DefaultConfiguration, Telemetry), start, end, samples);
/// <summary>
/// Runs the untouched baseline (1.5.378.182) calculator for <paramref name="aggregate"/> exactly as it computes it,
/// under <see cref="DefaultConfiguration"/> — including its PercentGood / DurationGood rule (Uncertain counted as
/// Good) that neither HistoryRead nor, since #809, the live calculators use.
/// </summary>
/// <param name="aggregate">The aggregate.</param>
/// <param name="start">Request start (later than <paramref name="end"/> for a reverse read).</param>
@@ -86,8 +101,9 @@ public static class SdkAggregateOracle
public static IReadOnlyList<Result> RunSdkAsIs(HistoryAggregateType aggregate, DateTime start, DateTime end,
TimeSpan interval, IEnumerable<(DateTime Time, HistorianValue Value, uint Status)> samples,
AggregateConfiguration? configuration = null) =>
RunCore(CompatibilityAggregateFactory.Create(AggregateId(aggregate), start, end, interval.TotalMilliseconds, false,
configuration ?? DefaultConfiguration, Telemetry), start, end, samples);
RunCore(new LegacyCalculatorAdapter(BaselineAggregators.CreateStandardCalculator(AggregateId(aggregate), start, end,
interval.TotalMilliseconds, false, configuration ?? DefaultConfiguration, Telemetry)
?? throw new ArgumentOutOfRangeException(nameof(aggregate))), start, end, samples);
private static IReadOnlyList<Result> RunCore(IAggregateCalculator calc, DateTime start, DateTime end,
IEnumerable<(DateTime Time, HistorianValue Value, uint Status)> samples)
@@ -155,42 +171,7 @@ public static class SdkAggregateOracle
problems.ShouldBeEmpty($"{label} differs from the SDK:\n" + string.Join("\n", problems));
}
/// <summary>
/// The baseline <see cref="BaselineStatusAggregateCalculator"/> with PercentGood / DurationGood counting <b>Good</b>
/// regions only — Part 13 §5.4.3.31/§5.4.3.33, and the 1.5.378.106 rule. Everything else (bounds, regions, Partial,
/// status, the millisecond Duration unit) is the baseline's own code, run under the same configuration as
/// PercentBad / DurationBad.
/// </summary>
private sealed class Part13GoodCalculator(NodeId aggregateId, DateTime startTime, DateTime endTime,
double processingInterval, bool stepped, AggregateConfiguration configuration, ITelemetryContext telemetry,
bool usePercent)
: BaselineStatusAggregateCalculator(aggregateId, startTime, endTime, processingInterval, stepped, configuration, telemetry)
{
protected override LegacyDataValue ComputeValue(TimeSlice slice)
{
var values = GetValuesWithSimpleBounds(slice);
if (values == null || values.Count == 0) return GetNoDataValue(slice);
var regions = GetRegionsInValueSet(values, false, true);
double good = 0, total = 0;
foreach (var region in regions)
{
total += region.Duration;
if (StatusCode.IsGood(region.StatusCode)) good += region.Duration;
}
var value = new LegacyDataValue
{
WrappedValue = ZB.MOM.WW.OtOpcUa.OpcUaServer.SdkValueBoundary.FromClr(usePercent ? good / total * 100 : good),
SourceTimestamp = GetTimestamp(slice),
ServerTimestamp = GetTimestamp(slice),
};
value.StatusCode = value.StatusCode.WithAggregateBits(AggregateBits.Calculated);
return value;
}
}
/// <summary>The <see cref="CompatibilityAggregateFactory"/> adapter, for a calculator built here.</summary>
/// <summary>The <see cref="CompatibilityAggregateFactory"/> adapter, for a baseline calculator built here.</summary>
private sealed class LegacyCalculatorAdapter(BaselineAggregateCalculator calculator) : IAggregateCalculator
{
public NodeId AggregateId => calculator.AggregateId;
@@ -414,9 +414,10 @@ public sealed class SdkAggregateParityTests
public void PercentGood_keeps_Uncertain_out_of_Good_per_Part_13_where_the_baseline_SDK_counts_it_as_Good(
double from, double to)
{
// #796, documented divergence (SdkAggregateOracle remarks). Part 13 §5.4.3.31 / §5.4.3.33: PercentGood is the Good
// share of the interval. The 1.5.378.182 StatusAggregateCalculator, under the server's TreatUncertainAsBad = false,
// adds Uncertain regions to it (1.5.378.106 did not). Hand-computed over 4 × 5 s regions G / U / B / G:
// #796, documented divergence of the untouched baseline (SdkAggregateOracle remarks; live follows Part 13 since
// #809). Part 13 §5.4.3.31 / §5.4.3.33: PercentGood is the Good share of the interval. The 1.5.378.182
// StatusAggregateCalculator, under the server's TreatUncertainAsBad = false, adds Uncertain regions to it
// (1.5.378.106 did not). Hand-computed over 4 × 5 s regions G / U / B / G:
// Good 10 s ⇒ 50 %, Bad 5 s ⇒ 25 %, Uncertain 5 s in neither; the baseline says 75 % Good.
Pct(AsIs(HistoryAggregateType.PercentGood, from, to, GoodUncertainBadGood))
.ShouldBe(75d, 1e-9, "the baseline counts the Uncertain 5 s as Good — if this moves, revisit the oracle (#796)");
@@ -524,15 +525,16 @@ public sealed class SdkAggregateParityTests
public void An_Uncertain_first_sample_in_a_late_start_interval_makes_the_SDK_integrate_the_time_before_the_data(
double from, double to)
{
// #821, pinned divergence (pending the #809/#821 owner decision — this test records, it does not choose).
// #821, pinned divergence (pending the #821 owner decision — this test records, it does not choose).
// Part 13 §5.4.3.3/§5.4.3.8: the interpolated-bounding signal is defined from the first sample only — 10 → 11 over
// [5,12) (73.5) and 11 → 11 + 8/13 over [12,20) (4·(22 + 8/13)) ⇒ Total 163.96…, TimeAverage = Total ÷ 15 s.
const double part13Total = 73.5 + 88 + 32d / 13;
Pct(Ours(HistoryAggregateType.Total, from, to, 0, UncertainFirstLateStart)).ShouldBe(part13Total, 1e-9);
Pct(Ours(HistoryAggregateType.TimeAverage, from, to, 0, UncertainFirstLateStart)).ShouldBe(part13Total / 15, 1e-9);
// The frozen 1.5.378.182 calculators — exactly what the server registers for live AggregateFilter items
// (CompatibilityAggregateFactory; Sdk2AggregateCompatibilityTests proves the runtime manager serves the same code)
// The frozen 1.5.378.182 calculators — for Total / TimeAverage exactly what the server registers for live
// AggregateFilter items (CompatibilityAggregateFactory; Sdk2AggregateCompatibilityTests proves the runtime manager
// serves the same code)
// — integrate the uncovered [0,5) as a ramp 0 → 10 (+25) and divide TimeAverage by the WHOLE 20 s interval.
Pct(AsIs(HistoryAggregateType.Total, from, to, UncertainFirstLateStart)).ShouldBe(part13Total + 25, 1e-9);
Pct(AsIs(HistoryAggregateType.TimeAverage, from, to, UncertainFirstLateStart)).ShouldBe((part13Total + 25) / 20, 1e-9);
@@ -557,16 +559,108 @@ public sealed class SdkAggregateParityTests
Stock201(agg, from, to, GoodFirstLateStart).Single().Value!.Value.ShouldBe(expected, 1e-9);
}
// #809 on the same input: DurationGood / PercentGood — Part 13 (ours) and stock SDK 2.0.1 count Good regions only
// ([12,20) = 8 s); the frozen live calculators also count the Uncertain [5,12) as Good (15 s).
// #809 on the same input: DurationGood / PercentGood — Part 13 (ours), stock SDK 2.0.1 and, since #809, the live
// calculators count Good regions only ([12,20) = 8 s); the untouched baseline (live before #809) also counted the
// Uncertain [5,12) as Good (15 s).
Pct(Ours(HistoryAggregateType.DurationGood, from, to, 0, UncertainFirstLateStart)).ShouldBe(8_000d, 1e-9);
Stock201(HistoryAggregateType.DurationGood, from, to, UncertainFirstLateStart).Single().Value!.Value.ShouldBe(8_000d, 1e-9);
Pct(Sdk(HistoryAggregateType.DurationGood, from, to, 0, UncertainFirstLateStart)).ShouldBe(8_000d, 1e-9);
Pct(AsIs(HistoryAggregateType.DurationGood, from, to, UncertainFirstLateStart)).ShouldBe(15_000d, 1e-9);
Pct(Ours(HistoryAggregateType.PercentGood, from, to, 0, UncertainFirstLateStart)).ShouldBe(40d, 1e-9);
Stock201(HistoryAggregateType.PercentGood, from, to, UncertainFirstLateStart).Single().Value!.Value.ShouldBe(40d, 1e-9);
Pct(Sdk(HistoryAggregateType.PercentGood, from, to, 0, UncertainFirstLateStart)).ShouldBe(40d, 1e-9);
Pct(AsIs(HistoryAggregateType.PercentGood, from, to, UncertainFirstLateStart)).ShouldBe(75d, 1e-9);
}
/// <summary>
/// #809: the #810 QualityMix tag Q — Good / Uncertain / Bad samples and an outage marker, Uncertain regions starting
/// at a sample, through a sloped bound and held into an outage, with Good samples before (−10) and after (+90) every
/// window so no interval holds the start or the end of the data.
/// </summary>
private static readonly StoredSample[] UncertainMix =
[
S(-10, HistorianValue.FromDouble(10), Good),
S(0, HistorianValue.FromDouble(10), Good),
S(4, HistorianValue.FromDouble(12), Uncertain),
S(7, HistorianValue.FromDouble(13), Bad),
S(10, HistorianValue.FromDouble(14), Good),
S(13, HistorianValue.FromDouble(15), Uncertain),
S(20, HistorianValue.FromDouble(16), Uncertain),
S(25, HistorianValue.FromDouble(17), Uncertain),
S(33, HistorianValue.FromDouble(18), Uncertain),
S(40, HistorianValue.FromDouble(20), Good),
S(45, HistorianValue.Empty, BadNoCommunication),
S(50, HistorianValue.FromDouble(22), Uncertain),
S(56, HistorianValue.FromDouble(23), Good),
S(64, HistorianValue.FromDouble(25), Uncertain),
S(67, HistorianValue.FromDouble(26), Good),
S(75, HistorianValue.FromDouble(27), Bad),
S(78, HistorianValue.FromDouble(28), Good),
S(90, HistorianValue.FromDouble(30), Good),
];
/// <summary>Windows of whole intervals only (no truncated interval — ours flags it Partial, a documented divergence).</summary>
public static TheoryData<double, double, double> WholeIntervalWindows => new()
{
{ 0, 80, 20 }, { 80, 0, 20 },
{ 0, 80, 10 }, { 80, 0, 10 },
{ 5, 80, 15 }, { 80, 5, 15 },
{ 0, 60, 0 }, { 60, 0, 0 },
};
[Theory]
[MemberData(nameof(WholeIntervalWindows))]
public void Live_AggregateFilter_PercentGood_and_DurationGood_equal_HistoryRead_over_Uncertain_samples(
double from, double to, double intervalS)
{
// #809: a live AggregateFilter monitored item and a HistoryRead-Processed (SqlServer / QuestDB share
// TimeWeightedAggregator) answer PercentGood / DurationGood identically over Uncertain data — same values,
// timestamps, status and Partial / Calculated bits, with NO documented-divergence mask. RunLive is exactly
// CompatibilityAggregateFactory.Create, the calculator the server registers (Sdk2AggregateCompatibilityTests
// proves the booted runtime manager serves the same code).
var interval = intervalS > 0 ? TimeSpan.FromSeconds(intervalS) : (At(to) - At(from)).Duration();
var failures = new List<string>();
var discriminated = 0;
foreach (var agg in new[] { HistoryAggregateType.PercentGood, HistoryAggregateType.DurationGood })
foreach (var (name, data) in new[] { ("UncertainMix", UncertainMix), ("DenseUncertain", DenseUncertain) })
{
var rows = data.Select(x => (x.TimeUtc, x.Value, x.Status)).ToArray();
var live = SdkAggregateOracle.RunLive(agg, At(from), At(to), interval, rows);
try
{
SdkAggregateOracle.ShouldMatch(Ours(agg, from, to, intervalS, data), live, $"live {agg} {name} {from}→{to}/{intervalS}s");
}
catch (Exception ex)
{
failures.Add(ex.Message);
}
// The data must discriminate: the pre-#809 live calculator (the untouched baseline) counts the Uncertain
// time as Good, so it disagrees somewhere — or this parity would pass against the old behaviour too.
var before = SdkAggregateOracle.RunSdkAsIs(agg, At(from), At(to), interval, rows);
if (before.Zip(live).Any(p => Math.Abs(Convert.ToDouble(p.First.Value) - Convert.ToDouble(p.Second.Value)) > 1e-9))
discriminated++;
}
failures.ShouldBeEmpty(string.Join("\n", failures));
discriminated.ShouldBe(4, "every aggregate × data set must differ from the pre-#809 live answer somewhere");
}
[Fact]
public void Live_AggregateFilter_PercentGood_and_DurationGood_are_the_hand_computed_Part_13_values()
{
// #809, hand-computed over UncertainMix's eight 10 s intervals (QualityMixConformance derives them region by
// region): Good seconds 4, 3, 0, 0, 5, 4, 3, 2. The pre-#809 live answer added the Uncertain time
// (7, 10, 10, 10, 5, 10, 10, 7 s).
double[] goodSeconds = [4, 3, 0, 0, 5, 4, 3, 2];
var rows = UncertainMix.Select(x => (x.TimeUtc, x.Value, x.Status)).ToArray();
var ten = TimeSpan.FromSeconds(10);
SdkAggregateOracle.RunLive(HistoryAggregateType.DurationGood, At(0), At(80), ten, rows)
.Select(r => Convert.ToDouble(r.Value)).ShouldBe(goodSeconds.Select(g => g * 1000), tolerance: 1e-9);
SdkAggregateOracle.RunLive(HistoryAggregateType.PercentGood, At(0), At(80), ten, rows)
.Select(r => Convert.ToDouble(r.Value)).ShouldBe(goodSeconds.Select(g => g * 10), tolerance: 1e-9);
SdkAggregateOracle.RunSdkAsIs(HistoryAggregateType.DurationGood, At(0), At(80), ten, rows)
.Select(r => Convert.ToDouble(r.Value)).ShouldBe([7000d, 10000, 10000, 10000, 5000, 10000, 10000, 7000], tolerance: 1e-9);
}
/// <summary>#684: Good 10 / 30 with an Uncertain value BELOW the Good minimum (+4) and one ABOVE the maximum (+6).</summary>
private static readonly StoredSample[] UncertainBeyond =
[
@@ -6,9 +6,10 @@ namespace ZB.MOM.WW.OtOpcUa.OpcUaServer.IntegrationTests;
/// <summary>
/// #821 / #809 facts, through the BOOTED server's runtime <c>AggregateManager</c> — the calculators a live
/// AggregateFilter monitored item gets, under the server's own default configuration. This pins current behaviour for
/// the owner decision; it does not choose. The HistoryRead (Part 13) answers and SDK 2.0.1's stock answers for the
/// same input are pinned beside it in <c>SdkAggregateParityTests</c>.
/// AggregateFilter monitored item gets, under the server's own default configuration. The #821 Total / TimeAverage
/// facts pin current behaviour for the owner decision; they do not choose. #809 is decided (2026-11-11): live
/// PercentGood / DurationGood count Good time only, equal to HistoryRead. The HistoryRead (Part 13) answers and SDK
/// 2.0.1's stock answers for the same input are pinned beside it in <c>SdkAggregateParityTests</c>.
/// </summary>
public sealed class LiveAggregateLateStartFactsTests
{
@@ -45,9 +46,10 @@ public sealed class LiveAggregateLateStartFactsTests
const double part13Total = 73.5 + 88 + 32d / 13;
One(ObjectIds.AggregateFunction_Total).ShouldBe(part13Total + 25, 1e-9); // 188.96…
One(ObjectIds.AggregateFunction_TimeAverage).ShouldBe((part13Total + 25) / 20, 1e-9); // 9.448… (Part 13: 10.93…)
// #809: the Uncertain [5,12) counts as Good live (HistoryRead: 8 000 ms / 40 %).
One(ObjectIds.AggregateFunction_DurationGood).ShouldBe(15_000d, 1e-9);
One(ObjectIds.AggregateFunction_PercentGood).ShouldBe(75d, 1e-9);
// #809: the Uncertain [5,12) is not Good — only [12,20) is, live as in HistoryRead (8 000 ms / 40 %; before #809
// the live calculator counted the Uncertain time too: 15 000 ms / 75 %).
One(ObjectIds.AggregateFunction_DurationGood).ShouldBe(8_000d, 1e-9);
One(ObjectIds.AggregateFunction_PercentGood).ShouldBe(40d, 1e-9);
// DurationBad / PercentBad agree with HistoryRead: the 5 s before the data is Bad.
One(ObjectIds.AggregateFunction_DurationBad).ShouldBe(5_000d, 1e-9);
One(ObjectIds.AggregateFunction_PercentBad).ShouldBe(25d, 1e-9);
@@ -9,8 +9,20 @@ using ZB.MOM.WW.OtOpcUa.OpcUaServer.Compatibility.Aggregates;
namespace ZB.MOM.WW.OtOpcUa.OpcUaServer.IntegrationTests;
/// <summary>Compares all preserved aggregate factories with independently generated SDK 1.5.378.182 vectors.</summary>
/// <remarks>
/// #809 (owner-approved 2026-11-11): PercentGood / DurationGood no longer count Uncertain time as Good. The baseline
/// vectors that changes are exempted EXACTLY — a PercentGood / DurationGood vector whose output differs from the same
/// case computed by the baseline SDK with <c>TreatUncertainAsBad = true</c> (its "twin": there 1.5.378.182 already
/// counted Good regions only) — and each is pinned to that twin instead. The twin is baseline evidence, not output of
/// the application's calculator, so the pin never regenerates expectations from the implementation.
/// </remarks>
public sealed class Sdk2AggregateCompatibilityTests
{
private const string PercentGood = "AggregateFunction_PercentGood";
private const string DurationGood = "AggregateFunction_DurationGood";
private const int Issue809PercentGoodExemptions = 38;
private const int Issue809DurationGoodExemptions = 26;
[Theory]
[InlineData(false)]
[InlineData(true)]
@@ -19,57 +31,145 @@ public sealed class Sdk2AggregateCompatibilityTests
await using var rig = throughRuntimeManager
? await ServerEventRig.StartAsync("AggregateBaseline", encrypted: false, TestContext.Current.CancellationToken)
: null;
var baseline = JsonSerializer.Deserialize<Baseline>(File.ReadAllText(Path.Combine(AppContext.BaseDirectory, "Fixtures", "opcua-1.5.378.182-aggregates.json")), new JsonSerializerOptions { PropertyNameCaseInsensitive = true })!;
var telemetry = DefaultTelemetry.Create(_ => { });
var epoch = new DateTime(baseline.EpochTicks, DateTimeKind.Utc);
var baseline = LoadBaseline();
var exemptions = Issue809Exemptions(baseline);
var failures = new List<string>();
foreach (var vector in baseline.Vectors)
{
var id = (NodeId)typeof(ObjectIds).GetProperty(vector.Aggregate)!.GetValue(null)!;
var start = epoch.AddSeconds(vector.Reverse ? 70 : 0); var end = epoch.AddSeconds(vector.Reverse ? 0 : 70);
var profile = vector.Configuration;
var config = new AggregateConfiguration
{
TreatUncertainAsBad = profile?.TreatUncertainAsBad ?? vector.UncertainAsBad,
PercentDataBad = profile?.PercentDataBad ?? 100, PercentDataGood = profile?.PercentDataGood ?? 100,
UseSlopedExtrapolation = profile?.UseSlopedExtrapolation ?? false,
UseServerCapabilitiesDefaults = profile?.UseServerCapabilitiesDefaults ?? false,
};
var calculator = throughRuntimeManager
? rig!.Server.CurrentInstance.AggregateManager.CreateCalculator(id, start, end, baseline.IntervalMilliseconds, vector.Stepped, config)
: CompatibilityAggregateFactory.Create(id, start, end, baseline.IntervalMilliseconds, vector.Stepped, config, telemetry);
calculator.ShouldNotBeNull($"Runtime registration must serve {vector.Aggregate}.");
var output = new List<Value>(); var accepted = new List<bool>(); string? error = null;
try
{
var samples = baseline.Data[vector.DataSet].OrderBy(x => x.Seconds);
foreach (var sample in vector.Reverse ? samples.Reverse() : samples)
{
var time = epoch.AddSeconds(sample.Seconds);
accepted.Add(calculator.QueueRawValue(new DataValue(SdkValueBoundary.FromClr(Payload(sample)), sample.Status, time, time.AddTicks(sample.ServerOffsetTicks), sample.SourcePicoseconds, sample.ServerPicoseconds)));
if (vector.Incremental) Drain(false);
}
Drain(true);
}
catch (Exception ex) { error = ex.GetType().FullName; }
if (error != vector.Exception || !output.SequenceEqual(vector.Output) || (vector.QueueAccepted is { } expectedAccepted && !accepted.SequenceEqual(expectedAccepted)))
{
failures.Add($"{vector.Aggregate}/{vector.DataSet}/rev={vector.Reverse}/step={vector.Stepped}/uncertainBad={vector.UncertainAsBad}/incremental={vector.Incremental}: error {vector.Exception} vs {error}; expected {JsonSerializer.Serialize(vector.Output)} actual {JsonSerializer.Serialize(output)}");
}
void Drain(bool partial)
{
for (var guard = 0; guard < 100; guard++)
{
if (!calculator.TryGetProcessedValue(partial, out var value)) return;
var payload = SdkValueBoundary.ToClr(value.WrappedValue);
output.Add(new(partial, payload?.GetType().Name, Format(payload), value.StatusCode.Code, LegacyTimestampTicks(value.SourceTimestamp), LegacyTimestampTicks(value.ServerTimestamp), value.SourcePicoseconds, value.ServerPicoseconds));
}
throw new InvalidOperationException("Baseline calculator did not terminate.");
}
// #809: an exempted case is held to its baseline uncertain-as-bad twin; every other case to its own vector.
var expected = exemptions.TryGetValue(vector, out var twin) ? twin.Output : vector.Output;
if (Check(baseline, vector, expected, rig) is { } failure) failures.Add(failure);
}
baseline.Vectors.Length.ShouldBe(6144);
baseline.Vectors.Select(x => x.Aggregate).Distinct().Count().ShouldBe(37);
failures.ShouldBeEmpty("The application must retain baseline values, CLR types, complete status bits, timestamps and incremental processing.");
exemptions.Keys.Count(x => x.Aggregate == PercentGood).ShouldBe(Issue809PercentGoodExemptions);
exemptions.Keys.Count(x => x.Aggregate == DurationGood).ShouldBe(Issue809DurationGoodExemptions);
failures.ShouldBeEmpty("The application must retain baseline values, CLR types, complete status bits, timestamps and incremental processing (#809 exemptions: the uncertain-as-bad twin).");
}
/// <summary>
/// #809 pin: the exemptions are exactly the PercentGood / DurationGood cases that counted Uncertain as Good, and
/// every PercentGood / DurationGood case — exempt or not, whatever its configuration — now answers its baseline
/// uncertain-as-bad twin (Good regions only).
/// </summary>
[Fact]
public void Issue809_percent_and_duration_good_count_only_good_time()
{
var baseline = LoadBaseline();
var exemptions = Issue809Exemptions(baseline);
var failures = new List<string>();
var checkedCount = 0;
var twinned = 0;
foreach (var vector in baseline.Vectors.Where(v => v.Aggregate is PercentGood or DurationGood))
{
// A data set with no uncertain-as-bad twin (the numeric-width edge sets) has no Uncertain region to move.
var twin = UncertainAsBadTwin(baseline, vector);
if (exemptions.ContainsKey(vector))
EffectiveTreatUncertainAsBad(vector).ShouldBeFalse($"{Describe(vector)}: only an Uncertain-as-Good case may change.");
if (Check(baseline, vector, twin?.Output ?? vector.Output, rig: null) is { } failure) failures.Add(failure);
checkedCount++;
if (twin is not null) twinned++;
}
checkedCount.ShouldBe(352);
twinned.ShouldBeGreaterThan(exemptions.Count);
exemptions.Count.ShouldBe(Issue809PercentGoodExemptions + Issue809DurationGoodExemptions);
failures.ShouldBeEmpty("PercentGood / DurationGood must count Good regions only (Part 13 §5.4.3.31/§5.4.3.33, #809).");
}
/// <summary>
/// #809 hand-computed pin, independent of the expected outputs in any fixture: the "uncertain" data set (0 s Good,
/// 5 s Uncertain, 15 s Good, 25 s Bad, 35 s Good, 50 s Good) over 20 s intervals from 0 to 70 s, sloped, forward.
/// [0,20): Good 0–5 and 15–20 = 10 s (the baseline said 20 s); [20,40): the bound at 20 slopes into the Bad sample
/// (Uncertain), 25–35 Bad, 35–40 Good = 5 s (baseline 10 s); [40,60): 20 s; [60,70): all 10 s of the partial
/// interval = 100 %.
/// </summary>
[Theory]
[InlineData(DurationGood, new[] { 10000d, 5000d, 20000d, 10000d })]
[InlineData(PercentGood, new[] { 50d, 25d, 100d, 100d })]
public void Issue809_uncertain_time_is_not_good_hand_computed(string aggregate, double[] expected)
{
var baseline = LoadBaseline();
var vector = baseline.Vectors.Single(v => v.Aggregate == aggregate && v.DataSet == "uncertain" && !v.Reverse && !v.Stepped
&& !v.UncertainAsBad && !v.Incremental && v.Configuration is null);
var output = Run(baseline, vector, rig: null, out var error);
error.ShouldBeNull();
output.ShouldAllBe(x => x.Type == "Double");
output.Select(x => double.Parse(x.Payload!, CultureInfo.InvariantCulture)).ShouldBe(expected);
}
private static Baseline LoadBaseline() =>
JsonSerializer.Deserialize<Baseline>(File.ReadAllText(Path.Combine(AppContext.BaseDirectory, "Fixtures", "opcua-1.5.378.182-aggregates.json")), new JsonSerializerOptions { PropertyNameCaseInsensitive = true })!;
private static bool EffectiveTreatUncertainAsBad(Vector vector) => vector.Configuration?.TreatUncertainAsBad ?? vector.UncertainAsBad;
/// <summary>The same case computed by the baseline SDK with <c>TreatUncertainAsBad = true</c> and no explicit profile.</summary>
private static Vector? UncertainAsBadTwin(Baseline baseline, Vector vector) =>
baseline.Vectors.SingleOrDefault(v => v.Aggregate == vector.Aggregate && v.DataSet == vector.DataSet && v.Reverse == vector.Reverse
&& v.Stepped == vector.Stepped && v.Incremental == vector.Incremental && v.UncertainAsBad && v.Configuration is null);
/// <summary>#809: the PercentGood / DurationGood vectors whose baseline output differs from their uncertain-as-bad twin.</summary>
private static Dictionary<Vector, Vector> Issue809Exemptions(Baseline baseline) =>
baseline.Vectors
.Where(v => v.Aggregate is PercentGood or DurationGood)
.Select(v => (Vector: v, Twin: UncertainAsBadTwin(baseline, v)))
.Where(x => x.Twin is not null && !x.Vector.Output.SequenceEqual(x.Twin.Output))
.ToDictionary<(Vector Vector, Vector? Twin), Vector, Vector>(x => x.Vector, x => x.Twin!, System.Collections.Generic.ReferenceEqualityComparer.Instance);
private static string Describe(Vector vector) =>
$"{vector.Aggregate}/{vector.DataSet}/rev={vector.Reverse}/step={vector.Stepped}/uncertainBad={vector.UncertainAsBad}/incremental={vector.Incremental}/config={JsonSerializer.Serialize(vector.Configuration)}";
private static string? Check(Baseline baseline, Vector vector, Value[] expected, ServerEventRig? rig)
{
var output = Run(baseline, vector, rig, out var error);
return error != vector.Exception || !output.SequenceEqual(expected) || (vector.QueueAccepted is { } expectedAccepted && !output.Accepted.SequenceEqual(expectedAccepted))
? $"{Describe(vector)}: error {vector.Exception} vs {error}; expected {JsonSerializer.Serialize(expected)} actual {JsonSerializer.Serialize(output.ToArray())}"
: null;
}
private static RunOutput Run(Baseline baseline, Vector vector, ServerEventRig? rig, out string? error)
{
var telemetry = DefaultTelemetry.Create(_ => { });
var epoch = new DateTime(baseline.EpochTicks, DateTimeKind.Utc);
var id = (NodeId)typeof(ObjectIds).GetProperty(vector.Aggregate)!.GetValue(null)!;
var start = epoch.AddSeconds(vector.Reverse ? 70 : 0); var end = epoch.AddSeconds(vector.Reverse ? 0 : 70);
var profile = vector.Configuration;
var config = new AggregateConfiguration
{
TreatUncertainAsBad = profile?.TreatUncertainAsBad ?? vector.UncertainAsBad,
PercentDataBad = profile?.PercentDataBad ?? 100, PercentDataGood = profile?.PercentDataGood ?? 100,
UseSlopedExtrapolation = profile?.UseSlopedExtrapolation ?? false,
UseServerCapabilitiesDefaults = profile?.UseServerCapabilitiesDefaults ?? false,
};
var calculator = rig is not null
? rig.Server.CurrentInstance.AggregateManager.CreateCalculator(id, start, end, baseline.IntervalMilliseconds, vector.Stepped, config)
: CompatibilityAggregateFactory.Create(id, start, end, baseline.IntervalMilliseconds, vector.Stepped, config, telemetry);
calculator.ShouldNotBeNull($"Runtime registration must serve {vector.Aggregate}.");
var output = new RunOutput();
error = null;
try
{
var samples = baseline.Data[vector.DataSet].OrderBy(x => x.Seconds);
foreach (var sample in vector.Reverse ? samples.Reverse() : samples)
{
var time = epoch.AddSeconds(sample.Seconds);
output.Accepted.Add(calculator.QueueRawValue(new DataValue(SdkValueBoundary.FromClr(Payload(sample)), sample.Status, time, time.AddTicks(sample.ServerOffsetTicks), sample.SourcePicoseconds, sample.ServerPicoseconds)));
if (vector.Incremental) Drain(false);
}
Drain(true);
}
catch (Exception ex) { error = ex.GetType().FullName; }
return output;
void Drain(bool partial)
{
for (var guard = 0; guard < 100; guard++)
{
if (!calculator.TryGetProcessedValue(partial, out var value)) return;
var payload = SdkValueBoundary.ToClr(value.WrappedValue);
output.Add(new(partial, payload?.GetType().Name, Format(payload), value.StatusCode.Code, LegacyTimestampTicks(value.SourceTimestamp), LegacyTimestampTicks(value.ServerTimestamp), value.SourcePicoseconds, value.ServerPicoseconds));
}
throw new InvalidOperationException("Baseline calculator did not terminate.");
}
}
// UA FILETIME zero is the old CLR unset sentinel, not a different historical instant.
@@ -84,6 +184,7 @@ public sealed class Sdk2AggregateCompatibilityTests
"Single" => float.Parse(sample.Payload!, CultureInfo.InvariantCulture),
"Double" => double.Parse(sample.Payload!, CultureInfo.InvariantCulture), "Int64" => long.Parse(sample.Payload!, CultureInfo.InvariantCulture), "Boolean" => bool.Parse(sample.Payload!), _ => null };
static string? Format(object? value) => value switch { null => null, double number => number.ToString("R", CultureInfo.InvariantCulture), float number => number.ToString("R", CultureInfo.InvariantCulture), IFormattable f => f.ToString(null, CultureInfo.InvariantCulture), _ => value.ToString() };
private sealed class RunOutput : List<Value> { public List<bool> Accepted { get; } = []; }
record Baseline(long EpochTicks, int IntervalMilliseconds, Dictionary<string,Sample[]> Data, Vector[] Vectors);
record Sample(double Seconds,string Type,string? Payload,uint Status,ushort SourcePicoseconds = 0,ushort ServerPicoseconds = 0,long ServerOffsetTicks = 0);
record Value(bool PartialRequested,string? Type,string? Payload,uint Status,long SourceTicks,long ServerTicks,ushort SourcePicoseconds,ushort ServerPicoseconds);