TwinCATDriverProbe ran a first-phase TCP check against the AmsNetId's
first four octets on the AMS port (801/851). An AmsNetId is not an IP and
the AMS port is not a TCP port, so Test Connect timed out against every
real runtime (Z35474: dialled 5.118.56.93, an unrelated public host).
The probe now opens no socket of its own: it takes the driver's path,
AdsClient.Connect + ReadStateAsync through the local AMS router, bounded
by the probe budget (Task.Run + WaitAsync), and classifies failures as
no local AMS router / no route or target unreachable / timeout / other.
The degrade-to-green 'TCP reachability only' outcome is gone - with no
TCP check nothing was contacted, so a missing router is reported red.
Tests: the regression test listens on the AmsNetId-derived endpoint and
asserts no connection arrives (fails on the old probe); classification
and budget tests drive the handshake through an internal seam.
Claude-Session: https://claude.ai/code/session_01GT67AD99zQQqpLCVm3p7xk
RegisterNotificationAsync published a refused registration into
NativeRegistrations under ConnectGate with NextRetryUtc = MinValue; the
SubscribeAsync caller booked the refusal backoff (NoteDeadFailure refused:
true) only after the gate was released and after its WARN. A probe tick in
that window took the gate, saw the registration as due, and retried it at
once - the CI flake in A_rearm_during_an_in_flight_retry_is_not_lost
("AttemptsFor(\"MAIN.T\") should be 1 but was 2").
RegisterNotificationAsync now books the failure (refusal backoff, or the
comms-class every-tick retry) before it publishes the dead registration, still
under the gate. The caller no longer books the refusal a second time (that
would double the backoff); its comms-class booking is idempotent and stays.
Tests: TwinCATCapturingLogger gains an OnWarning hook. A logger that holds
the refusal WARN for ten probe intervals forces the window open, so
A_rearm_during_an_in_flight_retry_is_not_lost exercises it on every run, and
the new A_refused_registration_is_published_with_its_backoff_already_set
pins the backoff (booked once, not doubled).
Pass rate of A_rearm_during_an_in_flight_retry_is_not_lost:
- before, unmodified test: 40/40 runs, plus 200/200 in-process repeats under
CPU load (the natural window is a few instructions wide; the flake is rare);
- before, window held open: 0/32 (deterministic failure, "should be 1 but
was 2");
- after, window held open: 40/40.
Full TwinCAT suite 663/663 at AKKA_TEST_TIMEFACTOR 1 and 3.
Claude-Session: https://claude.ai/code/session_01GT67AEHd2Acv1YPwGfXEoV
AbLegacyDriver stamped LastSuccessfulRead (init, read, write), the read's sample
timestamps and host-state change times from raw DateTime.UtcNow; only the backoff
path read the injectable Clock. All of them now read Clock(), so
NativePcccBridgeStatusHealthTests drives a manual clock (init at T0, read at T1)
and asserts LastSuccessfulRead == T1 exactly instead of spin-waiting for the wall
clock to pass the init stamp (#767). Without the driver change the 6 tests fail
(init stamp is wall-clock, not T0).
Directory.Build.props sets TreatWarningsAsErrors for every project since #766, so
the 75 per-csproj copies (64 src + 11 tests) are removed and the props comment
says it is now the only place. tools/mem-probe keeps its explicit opt-out.
Release build: 0 warnings, 0 errors; a planted CS0168 still fails the build in a
src project and in a test project that used to carry its own copy.
Claude-Session: https://claude.ai/code/session_01GT67AEHd2Acv1YPwGfXEoV
- I1: the poll engine belongs to a teardown generation. ShutdownAsync/Dispose
retire the current engine and install a fresh one in the same swap as the
device/tag tables, then dispose the retired one (possibly in the background
after a cancelled or over-budget shutdown). A subscription made after the
next InitializeAsync can no longer land in an engine whose late dispose tail
would dispose its CTS without cancelling it (a loop no UnsubscribeAsync could
find). Poll handles are generation-qualified (TwinCATPollHandle) because each
engine numbers handles from 1; SubscribeAsync re-homes a subscription that
raced a retirement.
- M1: a write's status feeds per-device health - Good clears that device's
comms fault and advances LastSuccessfulRead, a comms status degrades, a
refusal keeps the state.
- M2: a symbol-version re-arm bumps a per-registration epoch; an in-flight
retry that fails afterwards no longer re-imposes the backoff.
- M3: a re-arm restarts the refusal backoff at 30 s.
- M4: documented that PLC-not-in-RUN detection needs the probe.
- M5: WSTRING and multi-dimensional array encode-plan tests; the cancelled-
shutdown test asserts Unknown health and awaits its background teardown.
- M6: an infinity spelled as a string ('Infinity', '-inf') passes like a
float/double infinity; a finite overflow stays BadOutOfRange.
Refs #739
Claude-Session: https://claude.ai/code/session_01Ub9hGmYLQvn9XETzw7CMsT
- IMPORTANT: a successful probe records the PLC's ADS state; a PLC not in Run
keeps the driver Degraded with 'PLC in {state}' while LastSuccessfulRead
still advances (the device answered). RecordDeviceAnswered never restores
Healthy while any device is out of Run.
- Health is per device, rolled up worst-of: a comms fault is remembered on the
device it was seen on and only that device answering clears it (device A
answering no longer clears device B's fault).
- Refused (non-comms) dead registrations back off 30 s doubling to 10 min and
are re-armed by a symbol-version change or retried by the reconnect replay;
comms-class failures keep the every-tick retry. A retry whose failure class
changes republishes the tag's status, once.
- Bit-RMW gates are kept across generations and teardowns (never cleared).
- ShutdownAsync honours its CancellationToken: it stops waiting, finishes the
bookkeeping and throws OperationCanceledException.
- Remarks on why a superseded InitializeAsync throws OperationCanceledException.
- REAL / LREAL values (scalar or array element) outside the target range are
OverflowException -> BadOutOfRange instead of being written as Infinity.
- String / array write encoding factored into AdsTwinCATClient.PlanStringWrite
/ PlanArrayWrite; the fake runs them so a driver-level test covers the real
encode path against a fake resolved symbol.
- Timing-fragile tests use condition waits / measured probe-tick windows; the
dispose-during-shutdown test asserts ordering with a 1 s bound.
Refs #739
Claude-Session: https://claude.ai/code/session_01Ub9hGmYLQvn9XETzw7CMsT
The poll engine already builds a per-cycle batch; TwinCAT now forwards it as
OnDataChangeBatch, so DriverInstanceActor gets one message per poll cycle
instead of one per changed tag. Because the actor subscribes ONLY the batch
event of a batch-capable driver, every native-path data change (pushes, the
BadNodeIdUnknown / refusal / unavailable snapshots) is raised through a single
RaiseDataChange that fires both the per-tag event and a fresh one-item batch.
Refs #739
Claude-Session: https://claude.ai/code/session_01Ub9hGmYLQvn9XETzw7CMsT
- Device and tag tables are build-and-swap: InitializeAsync builds a fresh
generation and publishes it by one swap under a lifecycle lock, then
releases the replaced generation; teardown swaps in empty tables. A
retained poll loop never sees a table mid-Clear (it used to read
_tagsByRawPath while a re-init only ever added to it), and each device
set is released by exactly one caller.
- A lifecycle generation makes an InitializeAsync overtaken by a shutdown
(or a newer init) release what it built instead of installing it - the
Lane A abandoned-init-overlaps-shutdown case.
- ShutdownAsync joins all probe loops together within 2 s (it awaited each
unbounded - a probe wedged in the blocking ADS Connect hung the actor's
teardown) in parallel with the poll-engine dispose, the whole bounded at
4.5 s; DisposeAsync may run concurrently and returns at once. A connect
that completes on a released device disposes its client.
- Releasing a generation disposes its native registrations, so a direct
re-init no longer leaves live notifications bound to dead DeviceStates.
- Connect gates and RMW semaphores are no longer disposed under in-flight
users (no kernel handle is ever allocated); a fault on a released device
is not recorded as driver health.
- The notification channel counts DropWrite overflows
(ITwinCATClient.DroppedNotificationCount); the probe tick warns with the
delta. A queued value whose registration was disposed is purged in the
client dispatch loop and dropped driver-side for any client.
Refs #739
Claude-Session: https://claude.ai/code/session_01Ub9hGmYLQvn9XETzw7CMsT
Array tags were advertised writable but every write threw InvalidCastException
in ConvertForWrite (reported BadTypeMismatch); STRING writes went through the
Beckhoff marshaller with no length check against STRING(n) and the host's
ANSI code page, which the read side deliberately avoids.
- Array write: TwinCATArrayCodec.ToWriteElements converts exactly ArrayLength
elements to the declared type (CLR lists convert - ScadaBridge writes int[]
to Int16 nodes; a length mismatch or out-of-range element is BadOutOfRange, a
scalar value BadTypeMismatch) before any wire traffic; the ADS client plans
the layout from the resolved PLC array (TwinCATTypeGuard.CheckArray, as the
read does), encodes with TwinCATArrayCodec.Encode - the exact inverse of
Decode - and writes the raw bytes (new ITwinCATClient.WriteArrayAsync).
- STRING / WSTRING write: encoded at the PLC's declared width by
TwinCATStringCodec.EncodeString/EncodeWString (Windows-1252 or UTF-8 per
target, NUL-padded, strict: an over-long value or unrepresentable character
is OverflowException -> BadOutOfRange) and written as raw bytes.
- TwinCAT CLI read/write gain --array-length (+ --separator) for the hardware
gate.
Refs #739
Claude-Session: https://claude.ai/code/session_01Ub9hGmYLQvn9XETzw7CMsT
The bit write read and wrote the parent word as a 32-bit UDInt with no parent
check: a bit beyond a 16-bit parent was written into neighbouring bytes, and a
signed parent with its sign bit set read back >= 2^31, so the checked
conversion turned any bit write into BadOutOfRange.
The write now goes through ITwinCATClient.WriteBitAsync: the ADS client
resolves the parent symbol, checks the bit against it (TwinCATBitAccess.
CheckParent, as the notification path already did), reads exactly
parent.ByteSize raw bytes, flips one bit (TwinCATBitAccess.ApplyBitWrite) and
writes the bytes back by index group/offset or variable handle. A bit beyond
the parent is BadOutOfRange, a non-integer parent BadTypeMismatch, a short
parent read BadCommunicationError (never written back), all without a health
write. The per-parent RMW lock stays in the driver.
Refs #739
Claude-Session: https://claude.ai/code/session_01Ub9hGmYLQvn9XETzw7CMsT
CompleteRead logged a Warning for every non-Good status on every poll and
latched Degraded for all of them. Now a bad tag warns once per bad episode
(re-armed when it reads Good, and on re-initialise), a comms-class status
(BadCommunicationError / BadTimeout) degrades with LastSuccessfulRead
preserved, and any other status - the device answered and refused this tag -
advances LastSuccessfulRead with no state change. Read-path exception
handlers now go through MarkCommsFault, so they never downgrade Faulted.
Refs #739
Claude-Session: https://claude.ai/code/session_01Ub9hGmYLQvn9XETzw7CMsT
In native-notification mode LastSuccessfulRead never moved past the init
stamp, so one Degraded write on a connection older than 90 s latched the
comms-loss watchdog into a full reconnect (every notification on every device
deleted and re-registered). Conversely an unreachable AMS router was swallowed
by the probe and never left Healthy.
- a native push and a successful probe now stamp LastSuccessfulRead (throttled
to one health record per 250 ms) and restore Healthy unless a device is still
probe-unreachable;
- 3 consecutive failed probe ticks write Degraded with the stamp preserved;
- a NotSupportedException write is BadNotSupported with no health write (#611);
- a reconnect replay the PLC refuses republishes the refusal's status without
degrading (it used to report BadCommunicationError + Degraded on every
reconnect for a correctly-isolated type-mismatch tag).
Refs #739
Claude-Session: https://claude.ai/code/session_01Ub9hGmYLQvn9XETzw7CMsT
Native subscribe was all-or-nothing: any registration failure other than a
type refusal disposed every registration and rethrew, so one typo'd symbol or
one unreachable PLC in a multi-device driver served zero data, churned
register/delete every reconcile tick, and left health Healthy.
Now each registration fails on its own: a refusal (the device answered: symbol
not found, PLC not ready, type mismatch) publishes its status with no health
write; any other per-tag failure publishes BadCommunicationError and degrades;
a connect failure marks the device unreachable for the rest of the call and
stores its tags as dead intents. Every failed tag is kept as a dead intent that
the probe-tick retry and the reconnect replay re-register. The ADS client now
classifies a failed symbol resolve / AddDeviceNotification by status instead of
throwing InvalidOperationException for all of them. Cancellation still tears
the partial subscription down.
Refs #739
Claude-Session: https://claude.ai/code/session_01Ub9hGmYLQvn9XETzw7CMsT
The native subscribe path silently skipped a reference it could not resolve
(unknown RawPath, or a tag routed to an unconfigured device), leaving the node
on whatever status the server seeded. It now publishes one BadNodeIdUnknown
snapshot, matching the poll path.
Refs #739
Claude-Session: https://claude.ai/code/session_01Ub9hGmYLQvn9XETzw7CMsT
Clears CommentChecker findings in src/ (3,280 -> 209, the remainder being false
positives like PLC-5/IEEE-754/M0.0) by adding missing summaries/params/returns,
normalising inheritdoc usage, and removing Gitea/review tracking IDs while keeping
the rationale they annotated. Comment-only: comment-stripped sources are identical
to the parent commit.
Since #516 a DriverConfig change is stop + respawn, so DriverSpawnPlan.ToApplyDelta
was always empty and the in-place path had no producer. #663 and #664 were both
latent bugs in it. This deletes it:
- host: ApplyChildDelta, _pendingDelta, HandleApplyResult and HandleDeltaApplied
- plan: DriverSpawnPlan.ToApplyDelta
- actor: the ApplyDelta, ApplyResult and DeltaApplied messages, the Connected
reinit handler and the mid-connect AdoptConfigDuringInit
- tests: the actor tests that drove the seam
IDriver.ReinitializeAsync and each driver's implementation and unit tests stay.
Comments and docs that cited AdoptConfigDuringInit as the reason a driver must
release prior state at init now cite the IDriver contract and overlapping
retries, which still apply.
Claude-Session: https://claude.ai/code/session_01Ef2ozxS7QPanGT8uoATbTe
- ReadBucketAsync: the once-per-device "sum-read chunk failed" Warning fires only after a successful
connect; a connect/backoff failure is the connect path's to report. TwinCATSumChunkWarningTests pins
it (no warning across connect failures, exactly one across two later failing sum polls; red without
the guard) plus a source guard that ReadGroupAsync builds its SumRead only via CreateSumRead.
- Docs: diagnostics live-gated via the driver harness (/hosts not exercised); TC2 costs 8 reads on the
first tick and after each reconnect, then <= 4; read counts attributed to the unit tests.
Claude-Session: https://claude.ai/code/session_01Ef2ozxS7QPanGT8uoATbTe
Read-only live gate on Elma presses Z35474 (TC3 3.1.4024) and Z34834 (TC2 2.11):
- Sum-read: every chunk came back BadCommunicationError. Beckhoff's
SumRead(IList<IgIoSumReadEntity>) ctor casts to IList<SumDataEntity>, which
throws InvalidCastException for a List<> (only an array survives via
covariance). The fake-client tests never built a real SumRead. New
AdsTwinCATClient.CreateSumRead passes an array; TwinCATSumReadConstructionTests
pins it (red on the List form) plus a falsifiability control on the library.
After the fix: 59 tags, status identical everywhere, values identical on every
non-moving tag, ~0.35-0.66 s vs ~2.7 s per poll; TC2 CE 2.11 answers sum reads.
- An unexpected sum chunk failure now logs one Warning per device (it was
invisible: the health message was overwritten by the next good read).
- Diagnostics: TC2 has no TwinCAT_SystemInfoVarList; fall back to .SYSTEMINFO /
.SYSTEMTASKINFOARR[1] once the TC3 name answers not-found (live: all six keys).
- #612 browse: Flat verified on TC3 (10,770 leaves; VirtualTree 1,459 strict
subset) — comment and docs updated, no load-mode option needed.
Claude-Session: https://claude.ai/code/session_01Ef2ozxS7QPanGT8uoATbTe
Strict width+kind eligibility for sum-batched scalars (others read singly); latch only on ADS
1793/1794, split 1792/1797/1798 without latching; 32 KiB per-command budget and 500-item ceiling;
short sum response handled per item. Diagnostics: opt-in diagnosticsEnabled (default off), reuse the
probe's ReadState, sample after the host-state transition, JitterMs renamed HeadroomMs, compact
per-device line so many devices fit the 16-entry cap, unavailable-symbol latch cleared on reconnect.
AdminUI: min/max on sum-read max, diagnostics checkbox. Pure planner extracted and unit-tested.
Claude-Session: https://claude.ai/code/session_01Ef2ozxS7QPanGT8uoATbTe
Sum-read (#310): batches a poll's reads per device into sum commands capped at sumReadMaxItems,
per-item ADS errors mapped per tag, falls back to individual reads when the target refuses the
command. Diagnostics (#314): ADS state, app name, online-change count, task cycle/exec/jitter via
IDriverDiagnostics. Re-implemented against the current raw-byte read path (stranded auto/driver-gaps
commits predate it). Sum-write not ported. Unit tests only; live gate pending a route.
Claude-Session: https://claude.ai/code/session_01Ef2ozxS7QPanGT8uoATbTe
FOCAS, AbLegacy, AbCip, TwinCAT, S7 and OpcUaClient orphaned the previous
probe/recycle/fixed-tree loops and clients when re-initialised on a live driver.
Claude-Session: https://claude.ai/code/session_01Ef2ozxS7QPanGT8uoATbTe
Removes the 'use UseNativeNotifications=false for arrays' workaround and
documents the raw-bytes array model shared by poll and notification: element
checks, STRING arrays, multi-dim flattening, 0-based publication, leading
sub-range for a shorter arrayLength, refusals (struct/UDT arrays, too long,
scalar over array), the 128 KiB native cap (BadOutOfRange), the bit-path fix,
and the 2026-10-07 read-only live gate on Z33856 / Z33733.
Live on Z33856 a UInt tag over .ARRFAULTCODE (ARRAY [0..12] OF UINT) made the
Beckhoff read throw - BadCommunicationError and a Degraded driver - while the
native path reported BadTypeMismatch. CheckRead now refuses a scalar declared
over an array (by category, so an alias of an array type is caught), pointing
at the array-tag / element-path alternatives.
A native notification for MAIN.Flags.3 / .ADSSTATE_RUN.0 was registered on the
literal WORD.N path, which ADS has no symbol for: live on the TC 2.11 Elma
cleaner Z33856 it failed with DeviceSymbolNotFound, and that registration
error tore down the whole subscribe call. The read path already reads the
parent word; the notification now does the same - it registers the parent as
raw bytes at its resolved width and reads the bit (TwinCATBitAccess), refusing
a bit outside the parent or a non-integer parent with BadTypeMismatch.
An array tag subscribed natively used to register only its declared scalar
(publishing element 0 at Good) and was then refused wholesale with
BadTypeMismatch. The driver now hands the authored ArrayLength to the client,
which registers the whole array as raw bytes and decodes every notification.
- TwinCATTypeGuard.CheckArray plans the layout from the RESOLVED array type:
element checked like a scalar notification (width / float kind / string
kind); STRING(n)/WSTRING(n) elements stride n+1 / 2(n+1); multi-dim arrays
flattened in PLC memory order; an authored ArrayLength shorter than the PLC
array reads the leading sub-range, a longer one is refused; struct / pointer
/ nested-array elements, non-array symbols and VAR_IN_OUT refs are refused.
- CheckNotificationSize caps a native array at 128 KiB (largest registration
proven on a live TC 2.11 target); above it the tag is refused BadOutOfRange.
- TwinCATArrayCodec decodes the bytes into the element-typed CLR array.
- The poll read uses the same plan + codec instead of the Beckhoff typed array
read, which failed live for most real arrays (exact CLR element type, rank-N
for multi-dim, member arrays sliced by the under-reported TC2 symbol size).
- TwinCATNotificationRefusedException carries the refusal's status; the type
mismatch derives from it and the driver isolates either per tag.
Found by the read-only live gate against the TwinCAT 2.11 Elma cleaners
Z33856 / Z33733:
- The Beckhoff library reports a TC2 dotted MEMBER symbol's ByteSize as if the
PLC's byte count were a bit count (STRING(80) -> 11, INT -> 1, UDINT -> 1)
while DataType.ByteSize is right (81/2/4, raw SYM_INFOBYNAMEEX says 81). The
previous commit's typeof(byte[]) read sliced by the symbol size, so
.OPC_MACHINEDATA.PROCESSSTATUS read "ready for o" at Good; the notification
width check would also have refused every numeric member. The shape now takes
the resolved type's size, and string reads go by index group/offset (or a
fresh handle for 0xF005) at that width.
- P_INITINPUT_ACTION.SLOTID is "STRING(20) (VAR_IN_OUT)" - a 4-byte reference -
but resolves to a plain STRING(20); a String tag read 21 bytes at the pointer
as text at Good. The symbol's own VAR_IN_OUT marker now classifies it
NonScalar -> BadTypeMismatch.
Port of the ignitionoee TwinCAT string + type-reconciliation fixes onto the
Beckhoff-library client:
- STRING/WSTRING subscriptions never worked: AddDeviceNotificationExAsync was
called with typeof(string) and no length, and the Beckhoff any-type
marshaller throws ArgumentNullException for a string without one, so one
string tag tore down its whole subscribe call. A string now registers the
PLC's own byte width as raw bytes (0.19.2: STRING(n) of any n).
- Strings are read as raw bytes and decoded by TwinCATStringCodec. The
Beckhoff marshaller already cut a read at the first NUL (0.19.3 / TwinCAT-007
were not defects on the read path); the codec keeps that rule for both STRING
and WSTRING and extends it to the new notification path. A STRING now decodes
as Windows-1252 (TwinCAT-010) - or UTF-8 when the target reports UTF-8
strings - instead of the host culture's ANSI code page the Beckhoff
marshaller falls back to (1250 on a Polish host, 1251 Cyrillic, ...).
- TwinCATTypeGuard (TwinCAT-002/-008): a read refuses a numeric tag over a PLC
string (and vice versa / STRING vs WSTRING / string over a container) with
BadTypeMismatch instead of handing back a string object at Good; a
notification - raw bytes decoded as the DECLARED type - additionally refuses
width and float/integer mismatches (DINT over REAL reinterpreted the bits).
- A refused notification is isolated: that tag publishes BadTypeMismatch with
one WARN, keeps a dead intent for the probe-tick retry, and the rest of the
subscription registers.
A TC2 PLC reports top-level integer globals as INT16/UINT16 (354 symbols on
the Elma cleaner Z33856) and the Beckhoff type system keeps that spelling as
a built-in primitive, so MapSymbolTypeName returned null and discovery dropped
them silently. INT16/UINT16/INT32/UINT32/UINT8/FLOAT/DOUBLE now map like their
IEC twins; padded member names ("REAL ") are trimmed; a parenthesised
non-string form ("INT (VAR_IN_OUT)", a reference) stays unsupported.
A TC2 global is named with ONE leading dot (.OPC_HEARTBEAT) and the PLC
requires it (SYM_HNDBYNAME answers 0x710 without it). TryParse rejected the
dot, so the driver fell back to the raw string and lost bit-index parsing:
.X.WORD.3 went to ADS verbatim instead of reading .X.WORD and extracting bit 3.
The dot now stays on the root segment and round-trips through ToAdsSymbolName;
.., ., .[1] and .9x stay invalid.
Found on the docker-dev live gate, not in tests. Two drivers that faulted
during initialise (rtu-gate, opcua1) each rendered "0 tags" on /hosts against
two authored tags apiece, because a ledger that had never committed returned
SkippedTagReport.Empty.
A zero report is a positive claim -- "this driver skipped nothing" -- and
making it on behalf of a driver that never got as far as looking is exactly
the quiet inaccuracy #541 exists to remove. GetSkippedTags() and
SkippedTagLedger.Snapshot() now return null until the first committed build,
so the column renders the same em-dash it already used for a driver that
cannot report at all. Both are 'no answer'; neither is 'nothing skipped'.
Knock-on: a FOCAS driver whose only init threw now reports null rather than a
zero report, which is strictly more honest than the assertion the test
originally made.
A driver builds its RawPath -> definition table at init and tolerates a
per-tag miss: an unmappable TagConfig, an unparseable address, a
capability-matrix rejection. That tolerance is right -- one bad tag must not
take down the other few thousand. What was missing is any statement of the
fact afterwards. A skip is not an error and happens before any read, so the
driver stays Healthy, errors/5 min stays 0, Last error stays empty, and Last
read advances live off the tags that did map. On the wonder Makino in #540
that reading held while focas-z-34184 ignored 20 of its 22 tags.
New optional capability ISkippedTagReporter, backed by a shared
SkippedTagLedger so the reset/commit/cap rules are one implementation rather
than one per driver. The report -- authored count, exact skipped count, and a
capped sample of RawPath + reason -- rides DriverHealthChanged to a new Tags
column on /hosts, over the Phase 5 telemetry stream. No table: a driver-only
node has no ConfigDb to write to, the lesson from #521.
The issue named eight drivers. It is twelve -- every driver in
DriverTypeNames.All has an authored-tag skip path. Calculation (no scriptId),
MTConnect (no readable DataItem id), Galaxy (no attributeRef) and OpcUaClient
(no nodeId) were unlisted, and OpcUaClient's was the quietest of the twelve:
NamespaceMap drops such an entry with a bare `continue` and not even a log
line. SkippedTagReporterFleetTests pins the invariant reflectively over every
IDriver in src/Drivers/, so a thirteenth cannot re-introduce the silence.
Sql drops a tag at TWO points and its report is the union of both: a blob
that does not map, AND an identifier the 8.1 catalog gate rejects against the
live catalog. The second is what schema drift produces -- a renamed column, a
dropped view -- months after a deployment sealed green, and reporting only
the first would have that driver claim "0 skipped" while polling half its
tags, which is worse than reporting nothing because the column would then be
actively reassuring.
Deliberately additive to driver state, never a change of it. Latching
non-Healthy for a config problem would satisfy both halves of the #527
comms-loss conjunction -- 90 s without a successful read AND non-Healthy --
and a driver whose tags were ALL skipped has no successful reads by
construction, so it would reconnect-loop forever against a healthy device.
Three details that are load-bearing rather than incidental:
- The skip digest is part of the publish fingerprint. A skipped tag leaves
every other fingerprint component identical to a clean driver's, so without
it PublishHealthSnapshot's dedup swallows exactly the publish carrying the
signal -- the trap that already bit the rediscovery and host-status
channels. The digest covers the sample, not just the counts: a redeploy
that fixes one tag and breaks another leaves both counts unchanged.
- The count is exact; only the sample is capped, and the tooltip says so when
it is showing fewer than it counted. A truncated list that looked complete
would have an operator fix twenty tags and believe the job done.
- The ledger stages and commits, so an init that throws part-way leaves the
previous report standing. A partial snapshot would read as "we now skip
fewer tags", the opposite of the truth.
The proto field needs no companion presence flag, unlike host_statuses: a
singular message field carries real presence in proto3, and a has_ flag would
collide with protoc's generated HasSkippedTags.
Verified red-first three ways: dropping SkippedTagDigest from the fingerprint
fails the two dedup tests, removing the seam from one driver fails the fleet
guard, and dropping the catalog-gate leg fails the Sql union test. Full
solution builds with zero new warnings against 4e22f024.
A browse-committed AB CIP / TwinCAT / FOCAS tag bound with its model's DEFAULT
data type, not the one the device reported. A browsed REAL committed as
`{"tagPath":"Pump_1_Speed","dataType":"DInt"}` — and those three drivers decode
the wire bytes by that field, not by the raw row's OPC UA DataType — so the tag
read four bytes of a float through an integer decoder and published a plausible
wrong number with no error raised anywhere. FOCAS was missed when #536 was filed.
The type is CARRIED, not derived. Reverse-mapping the leaf's driver-agnostic
DriverDataType looks like the smaller fix and is wrong: SInt/Int/DInt/USInt/UInt
all collapse into Int32 on the way out of the driver, so that version would bind
a 16-bit INT as a 32-bit DINT while appearing to fix the REAL. Drivers now state
their own type name on DriverAttributeInfo.NativeDataType, which reaches the
picker as AttributeInfo.AddressFields["dataType"] and the commit mapper from
there. A leaf that states nothing keeps the model default — the pre-#536
behaviour, kept deliberately as the fallback rather than a guess.
Numeric strings are refused before parsing: Enum.TryParse accepts "9" for any
enum and Enum.IsDefined then confirms it, so an unguarded parse would bind a tag
to a type nobody named. A test caught this in the first version of the guard.
--- and the other half, which the fixture was hiding ---
Test Connect on a device with NO authored tags — every AB CIP device at the
moment an operator first presses the button — reported "CIP handshake failed"
against a healthy controller. That was written down as a fixture limitation
("ab_server doesn't implement @raw_cpu_type"). Both halves of that were wrong:
* @raw_cpu_type is not a libplctag special tag at all (only @raw, @tags and
@udt/ are), so it goes out as an ordinary symbolic name that no controller
defines, and the controller's "tag does not exist" IS the successful answer.
* ab_server answered an unresolvable tag path with CIP 0x03 where a Logix
controller answers 0x04 — ErrorBadParam vs ErrorNotFound, which the probe
classifies oppositely. Fixed in vendored patch 0002.
* With the fixture then telling the truth, the real defect surfaced:
AbCipDriverProbe.IsReachableException matched the native C status spellings
("STATUS_NOT_FOUND"), but libplctag.NET throws with the MANAGED enum name
("ErrorNotFound"). Those substrings can never match, so the reachable
classification could not fire on the exception path — which is the path every
no-authored-tags probe takes. This was failing on real hardware too, not just
against the sim. It now recovers the Status and defers to IsReachableStatus,
so there is one classification rather than two that can drift.
Rejected: switching the fallback to @tags. libplctag builds a tag-listing tag
with NO network I/O, so it reports OK against any TCP listener — measured green
against a Mosquitto broker.
Verification: full solution suite green. Falsification runs confirm exactly the
intended tests go red — 9 for the type carry, 3 (the reachable statuses only) for
the probe classification. Live on docker-dev: browse-committed TestString/
Motor_Run/TestSINT now store String/Bool/SInt, alongside last session's
Pump_1_Speed still holding the wrong DInt. TestSINT is the falsifier — its row
DataType is Int32, identical to TestDINT's, so no reverse map could have produced
"SInt".
Known consequence: pre-existing browse-committed tags keep their wrong dataType;
the fix applies to new commits only.
Claude-Session: https://claude.ai/code/session_015p7wGqy3YpZNCpDzTpGMKo
Both `IDriver` members had no production call site. Their only consumer was
`MemoryTracking`, deleted with the driver-tier recycle machinery in #522, and
`GetMemoryFootprint`'s doc-comment claimed "polled every 30s by Core" — which
had not been true since v3, even before #522 removed the poller.
Owner decision on #525 was (a) delete rather than (b) build the process-wide
watchdog that would consume them. The driver-side implementations are in git
history and were mostly reusable; what was always missing is the trigger, and a
member nobody calls is how the misleading doc-comment survived this long.
Removed from `IDriver`, all 12 driver implementations, and every test stub —
plus the support code that existed only to feed them, which is the part a
member-only deletion would have left behind as dead weight:
- MTConnect's three per-entry byte-weight constants, its cached
`ProbeModelDataItemCount`, and `CountDataItems` — a full device-tree walk run
on every cache fill to populate a number with no remaining reader.
- OpcUaClient's `_discoveredNodeCount` field and the discovery-pass write to it.
- MQTT's `ApproxBytesPerAuthoredTag`.
Two behavioural notes, neither a change in what the software does:
- `AbCipTemplateCache` now has no production reset path. Its only caller was
AB CIP's flush, which itself had no caller — so nothing changes except how
visible the fact is. Recorded on the type, along with the correction that its
comment's claim about a device `ReinitializeAsync` also flushing it was never
true. A UDT redefined on a live controller keeps its stale shape until the
driver instance is rebuilt.
- MTConnect's `/probe` cache is still dropped, by `InvalidateProbeModel()` on an
Agent restart. That is now `internal` rather than `private` so the tests which
used the flush purely to empty the cache drive the real path instead — they
now cover a case that actually occurs in production. Several unrelated doc
comments cited `FlushOptionalCachesAsync` as the canonical CAS example and are
repointed to it.
Tests: the ones that existed only to assert on the removed members are deleted
(3 OpcUaClient, 3 Galaxy, 2 MQTT, 1 MTConnect, 1 AbCip). Worth noting one was
vacuous — `FlushOptionalCachesAsync_resets_footprint_counter` asserted a
footprint of 0 before the flush and 0 after with no discovery in between, so it
would have passed against a flush that did nothing at all.
Full suite green except 7 failures verified identical on master before and
after (4 AbLegacy integration with the compose profile down, 3
Host.IntegrationTests), checked by name rather than assumed.
Docs corrected where they described this as live: `DriverLifecycle.md` (which
carried the #525 placeholder), `drivers/AbCip.md`, `drivers/AbLegacy.md`, and
`v2/driver-stability.md` — whose "polls this every 30 s" + per-driver
`DriverConfig` memory budget were never built as written and are now banner-ed
as design intent rather than description.
Claude-Session: https://claude.ai/code/session_015p7wGqy3YpZNCpDzTpGMKo
Removes the tag-set drift detector shipped in e77c8a35. The design was sound
-- the tautology drivers were correctly gated out, and the structurally
undetectable "vanished" direction was declared rather than faked -- but none of
that is the objection. The feature browsed real plant equipment on a recurring
timer, at a frequency no operator could configure (the interval was a
constructor parameter with no appsettings key), and that outbound traffic was
never once observed against a real PLC, CNC or MTConnect Agent.
Removed: TagSetDrift/TagSetDriftDetector; DriverInstanceActor's drift timer,
tick message, Props/ctor parameter, gate, handler and leaf-collector;
ITagDiscovery.AuthoredDiscoveryRefs + .DiscoveryStreamIncludesAuthoredTags and
their four implementations; 14 tests.
Kept: ITagDiscovery.SupportsOnlineDiscovery (it predates this and drives the
universal browse picker), and the whole IRediscoverable consumption from
09a401b8 -- the driver tells us, we do not poll.
CONSEQUENCE, deliberately accepted: AbCip and FOCAS browse a live backend but
implement no IRediscoverable, so they now have NO tag-change signal at all -- a
PLC re-download is invisible until someone re-browses by hand. Recorded in
CLAUDE.md so the next reader does not have to rediscover it, along with the
shape to reach for if that coverage is wanted back (event-driven, or triggered
by a reconnect/deploy, not a wall-clock timer).
Runtime.Tests 476 pass / 13 skipped (skip set unchanged); FOCAS 275, TwinCAT
192, MTConnect 491, AbCip 342, Core.Abstractions 266 all green.
Claude-Session: https://claude.ai/code/session_015p7wGqy3YpZNCpDzTpGMKo
The second thing #516/§8.2 deliberately did not build. The original objection
stands — Modbus, S7, MQTT, AbLegacy and Sql echo authored config from
DiscoverAsync, so diffing discovered against authored is a tautology for them
— but it is now ENCODED rather than used as a reason not to build.
The discriminator already existed: ITagDiscovery.SupportsOnlineDiscovery is
documented as "enumerates the tag set from the live backend rather than
replaying pre-declared/authored tags", and it separates the fleet cleanly —
FOCAS, TwinCAT, MTConnect and AbCip true; the five config-echo drivers
explicitly false. The check runs only for a driver that is
SupportsOnlineDiscovery AND reports the new AuthoredDiscoveryRefs. That is
nullable rather than empty-by-default on purpose: an empty collection
legitimately means "nothing authored, everything discovered is new", so
conflating the two would report total drift for a driver that never opted in.
Where the value is: AbCip and FOCAS browse a live backend but implement no
IRediscoverable, so before this they had no change signal at all. A periodic
compare is the only way to notice a PLC re-download.
A finding that changed the design: FOCAS, TwinCAT and AbCip all re-emit their
authored tags into the same DiscoverAsync stream as device-derived ones, so
the browse picker can show an operator their existing tags. Discovered is
therefore always a superset of authored, and a VANISHED tag can never appear
missing — one whole direction is structurally undetectable for three of the
four. Shipping a Vanished list that is permanently empty for them would have
been the half-inert seam this whole exercise removes. So the driver declares
DiscoveryStreamIncludesAuthoredTags, the detector does not compute the
undetectable half, and the operator message says "missing-tag detection
unavailable for this driver" rather than letting the absence of a missing-tag
clause read as reassurance. MTConnect is the only driver where both directions
work — its stream is purely probe-model-derived.
Behaviour: a failed browse is not drift (an unreachable device would otherwise
report every authored tag as vanished); a truncated capture is skipped for the
same reason; an unchanged drift is reported once rather than re-stamping its
timestamp; drift that resolves clears the prompt. Interval defaults to 5
minutes — it browses a real device, and a tag set changing is an engineering
event, not a runtime one. It reuses the Stage-2 surfacing, raising the same
/hosts re-browse prompt, and never rebuilds the served address space.
Comparison logic is a pure, separately-tested type rather than actor-inline.
14 new tests. The gate was verified load-bearing by deleting the
SupportsOnlineDiscovery half: the tautology-driver test goes red, and it
asserts discovery was never invoked rather than merely "no prompt raised",
which would have passed for the wrong reason if the timer never fired.
Full suite green except the same three pre-existing fixture-gated integration
suites, identical counts.
Every value verified against Opc.Ua.StatusCodes in the pinned SDK (1.5.378.106) by
reflection, not by copying siblings. All are client-visible: OPC UA clients branch on
status.
The three named in #497:
- Historian.Gateway SampleMapper "BadNoData" 0x800E0000 -> 0x809B0000
(0x800E0000 is BadServerHalted)
- Galaxy "BadTimeout" 0x800B0000 -> 0x800A0000
(0x800B0000 is BadServiceUnsupported)
- TwinCAT BadTypeMismatch 0x80730000 -> 0x80740000
(0x80730000 is BadWriteNotSupported)
BadTypeMismatch was wrong in three MORE drivers the issue did not name — FOCAS,
AbLegacy and AbCip carried the same 0x80730000. Every call site is a
FormatException/InvalidCastException conversion failure, so the name was right and
the value was wrong in all four.
Adding the reflection guard then surfaced nine further defects offline tests had
never checked:
- Galaxy GoodLocalOverride 0x00D80000 -> 0x00960000 (not a UA code at all)
- Galaxy UncertainLastUsableValue 0x40A40000 -> 0x40900000
- Galaxy UncertainSensorNotAccurate 0x408D0000 -> 0x40930000
- Galaxy UncertainEngineeringUnitsExceeded 0x408E0000 -> 0x40940000
- Galaxy UncertainSubNormal 0x408F0000 -> 0x40950000
- TwinCAT BadOutOfService 0x80BE0000 -> 0x808D0000 (was BadProtocolVersionUnsupported)
- TwinCAT BadInvalidState 0x80350000 -> 0x80AF0000 (was BadAttributeIdInvalid)
- AbCip/AbLegacy GoodMoreData 0x00A70000 -> 0x00A60000 (was GoodCommunicationEvent)
- Historian.Gateway GatewayQualityMapper Good_LocalOverride 0x00D80000 -> 0x00960000
Galaxy's whole Uncertain block read as though the OPC DA quality byte could be
shifted into the UA substatus position; it cannot — the substatuses are an unrelated
enumeration. GatewayQualityMapper already held the correct table, which is what the
Galaxy values are now reconciled against.
Guard: StatusCodeParityTests (Core.Abstractions.Tests, which already project-
references every driver) reflects over every `const uint` in the deployed
ZB.MOM.WW.OtOpcUa.Driver.*.dll whose name reads as a status code, and asserts it
equals Opc.Ua.StatusCodes.<name>. Discovery is by convention, so a new driver
assembly referenced by that project is covered with no edit here. It went red on all
nine defects above before the fixes and now checks 109 constants green.
Because reflection can only see a NAMED constant, two inline literals were hoisted so
the guard can reach them: Galaxy's BadTimeout/Bad into StatusCodeMap, and
GatewayQualityMapper's 15-entry table into a new HistorianStatusCodes. An inline
`0x800B0000u, // BadTimeout` at a call site is exactly how the Galaxy defect survived.
The drivers stay SDK-free by design; only the test project references Opc.Ua.Core,
as the oracle.
Also corrected: tests that pinned the wrong values (Galaxy StatusCodeMapTests,
SampleMapperTests, GatewayQualityMapperTests — all written from the same bad source),
a mislabelled comment in DriverInstanceActorTests, and stale constants in two design
docs, one of them the unexecuted MTConnect driver design that would have propagated
BadNoData's wrong value into a new driver.
The CLI's SnapshotFormatter value->name table was checked against the SDK and is
correct; no change needed.
Task 3/4/5: SupportsOnlineDiscovery => true on AbCipDriver, TwinCATDriver, FocasDriver
(their device enumeration is real, config-gated on EnableControllerBrowse / FixedTree.Enabled
which PatchForBrowse guarantees at open). One gate fact per driver test project.