fix(abcip): final review I4 — a type mismatch is a refusal, not a comms fault

A native TypeMismatch (read status, write gate, or CIP 0xFF/0x2107) used
to evict the runtime and degrade health on every poll; for a sole lease
holder that meant Forward Close + Forward Open per poll, and an all-
mismatched driver tripped the comms-loss watchdog. It now reports
BadTypeMismatch, keeps the handle, leaves health alone and warns once per
tag per instance — reads, whole-UDT reads and writes, plus the libplctag
width check. Tests (red against the pre-fix driver): single create over N
polls, unchanged health, one warning, and one Forward Open / zero Forward
Close over the real native runtime against FakeCipTarget. AbLegacy has no
TypeMismatch path (its native runtime has no type check).

Claude-Session: https://claude.ai/code/session_01Ef2ozxS7QPanGT8uoATbTe
This commit is contained in:
Joseph Doherty
2026-10-07 21:40:04 -04:00
parent 2cea1a0687
commit 352ea449e9
5 changed files with 247 additions and 8 deletions
+6 -2
View File
@@ -123,7 +123,8 @@ The driver has two interchangeable wire transports, selected by one `DriverConfi
`Native`; libplctag remains available as the fallback.
- ⚠️ **Client-visible:** native reports **`BadTypeMismatch`** for a mis-authored same-width type (for example a
tag authored `DINT` where the controller holds a `REAL`), where libplctag returned Good garbage. Correct the
authored type; see [Native transport behaviour](#native-transport-behaviour).
authored type; see [Native transport behaviour](#native-transport-behaviour). It does not degrade driver
health or churn the connection; one Warning per tag names it.
- **To fall back** one driver, set `"transport": "Libplctag"` in its driver config (or pick "libplctag
(fallback)" in the AdminUI dropdown). To fall back a whole node, set `Drivers:AbCip:DefaultTransport` to
`Libplctag` (below).
@@ -176,7 +177,10 @@ The driver has two interchangeable wire transports, selected by one `DriverConfi
holds a `REAL` returns the REAL's bits as an integer with a Good status under libplctag, and
`BadTypeMismatch` natively. That is the fix working, but a consumer will see Bad quality on tags that
previously looked healthy: correct the authored type. Same-width signed/unsigned pairs are exempt, and
writes are gated on a mismatch too.
writes are gated on a mismatch too. A mismatch is a **deterministic configuration refusal**, handled like
an Unsupported refusal (final review I4): the tag's handle is kept (no re-create, no reconnect), driver
health is unchanged (no `Degraded`, nothing fed to the comms-loss watchdog), and the Warning is logged
**once per tag** per driver instance, not every poll. The same applies to the libplctag-path width check.
- Malformed tag names put the device into backoff, as under libplctag (#570). STRING capacity includes the
alignment pad (84, not 82) (#570).
- **A refused operation is not a comms fault.** Any operation a runtime refuses as Unsupported (in AbCip, e.g.
@@ -195,6 +195,18 @@ public sealed class AbCipDriver : IDriver, IReadable, IWritable, ITagDiscovery,
// Configured-vs-effective rendering of _transport for the init log, e.g. "Native (default)".
private readonly string _transportDescription;
// Tags whose type-mismatch Warning has already been logged (final review I4): a mis-authored tag is
// polled forever, so the Warning is once per tag per driver instance, not once per poll.
private readonly System.Collections.Concurrent.ConcurrentDictionary<string, byte> _typeMismatchWarned =
new(StringComparer.OrdinalIgnoreCase);
/// <summary>Logs a type-mismatch Warning the first time <paramref name="tagName"/> reports one on this instance.</summary>
private void WarnTypeMismatchOnce(string tagName, string message, params object?[] args)
{
if (_typeMismatchWarned.TryAdd(tagName, 0))
_logger.LogWarning(message, args);
}
/// <summary>Upper bound on the poll-loop failure backoff — adopts the S7-proven 30 s cap fleet-wide (05/STAB-8).</summary>
private static readonly TimeSpan PollBackoffCap = TimeSpan.FromSeconds(30);
@@ -807,6 +819,19 @@ public sealed class AbCipDriver : IDriver, IReadable, IWritable, ITagDiscovery,
await runtime.ReadAsync(ct).ConfigureAwait(false);
var status = runtime.GetStatus();
if (status == (int)AbStatus.TypeMismatch)
{
// Final review I4: the controller ANSWERED with a type that disagrees with the authored
// one — a deterministic config error, not a comms fault. Keep the handle (evicting would
// re-create it — and, for a sole lease holder, re-Forward-Open — on every poll), leave
// health alone, warn once per tag.
results[fb.OriginalIndex] = new DataValueSnapshot(null, AbCipStatusMapper.BadTypeMismatch, null, now);
WarnTypeMismatchOnce(def.Name,
"AbCip read of {Tag} on device {Device}: the controller's type does not match the authored " +
"{DataType}; reporting BadTypeMismatch until the tag's dataType is corrected (logged once per tag)",
reference, def.DeviceHostAddress, def.DataType);
return ReadOutcome.None;
}
if (status != 0)
{
// Evict the stale handle so the next call re-creates it.
@@ -834,11 +859,13 @@ public sealed class AbCipDriver : IDriver, IReadable, IWritable, ITagDiscovery,
&& ExpectedScalarWidth(def.DataType) is { } width
&& runtime.GetSize() is var size && size >= 0 && size != width)
{
// Same class as the native TypeMismatch above (final review I4): deterministic, device
// answered — no Degraded, warn once per tag.
results[fb.OriginalIndex] = new DataValueSnapshot(null, AbCipStatusMapper.BadTypeMismatch, null, now);
_logger.LogWarning(
"AbCip read of {Tag} on device {Device} returned {Size} bytes but its authored type {DataType} is {Width} bytes wide; check the tag's dataType",
WarnTypeMismatchOnce(def.Name,
"AbCip read of {Tag} on device {Device} returned {Size} bytes but its authored type {DataType} is {Width} bytes wide; check the tag's dataType (logged once per tag)",
reference, def.DeviceHostAddress, size, def.DataType, width);
return ReadOutcome.Degraded($"type mismatch reading {reference}: {size} bytes for {def.DataType}");
return ReadOutcome.None;
}
// Review I-1 — an array tag (the EXPLICIT IsArray flag) decodes the whole buffer into an
// element-typed CLR array (int[]/float[]/bool[]/string[]…), INCLUDING a 1-element array
@@ -865,6 +892,14 @@ public sealed class AbCipDriver : IDriver, IReadable, IWritable, ITagDiscovery,
results[fb.OriginalIndex] = new DataValueSnapshot(null, AbCipStatusMapper.BadNotSupported, null, now);
return ReadOutcome.None;
}
catch (AbTransportException ate) when (ate.Status == AbStatus.TypeMismatch)
{
// Final review I4: deterministic, keep the handle, no health signal, warn once per tag.
results[fb.OriginalIndex] = new DataValueSnapshot(null, AbCipStatusMapper.BadTypeMismatch, null, now);
WarnTypeMismatchOnce(def.Name, "AbCip read of {Tag} on device {Device} refused: {Reason} (logged once per tag)",
reference, def.DeviceHostAddress, ate.Message);
return ReadOutcome.None;
}
catch (Exception ex)
{
// Transport exception — evict so the next read creates a fresh handle. An AbTransportException
@@ -911,6 +946,16 @@ public sealed class AbCipDriver : IDriver, IReadable, IWritable, ITagDiscovery,
await runtime.ReadAsync(ct).ConfigureAwait(false);
var status = runtime.GetStatus();
if (status == (int)AbStatus.TypeMismatch)
{
// Final review I4: deterministic — keep the handle, no health signal, warn once.
StampGroupStatus(group, results, now, AbCipStatusMapper.BadTypeMismatch);
WarnTypeMismatchOnce(parent.Name,
"AbCip whole-UDT read of {Parent} on device {Device}: the controller's type does not match " +
"the authored type; {MemberCount} members report BadTypeMismatch (logged once per tag)",
group.ParentName, parent.DeviceHostAddress, group.Members.Count);
return ReadOutcome.None;
}
if (status != 0)
{
EvictRuntime(device, parent.Name);
@@ -944,6 +989,14 @@ public sealed class AbCipDriver : IDriver, IReadable, IWritable, ITagDiscovery,
StampGroupStatus(group, results, now, AbCipStatusMapper.BadNotSupported);
return ReadOutcome.None;
}
catch (AbTransportException ate) when (ate.Status == AbStatus.TypeMismatch)
{
StampGroupStatus(group, results, now, AbCipStatusMapper.BadTypeMismatch);
WarnTypeMismatchOnce(parent.Name,
"AbCip whole-UDT read of {Parent} on device {Device} refused: {Reason} (logged once per tag)",
group.ParentName, parent.DeviceHostAddress, ate.Message);
return ReadOutcome.None;
}
catch (Exception ex)
{
EvictRuntime(device, parent.Name);
@@ -1027,7 +1080,15 @@ public sealed class AbCipDriver : IDriver, IReadable, IWritable, ITagDiscovery,
opLock.Release();
}
if (status != 0)
if (status == (int)AbStatus.TypeMismatch)
{
// Final review I4: deterministic — keep the handle, leave health, warn once per tag.
results[i] = new WriteResult(AbCipStatusMapper.BadTypeMismatch);
WarnTypeMismatchOnce(def.Name,
"AbCip write to {Tag} on device {Device} refused: the controller's type does not match the authored {DataType} (logged once per tag)",
w.FullReference, def.DeviceHostAddress, def.DataType);
}
else if (status != 0)
{
EvictRuntime(device, def.Name);
results[i] = new WriteResult(AbCipStatusMapper.MapStatus(status));
@@ -1089,6 +1150,16 @@ public sealed class AbCipDriver : IDriver, IReadable, IWritable, ITagDiscovery,
"AbCip write refused (not supported) for tag {Tag} on device {Device}",
w.FullReference, def.DeviceHostAddress);
}
catch (AbTransportException ate) when (ate.Status == AbStatus.TypeMismatch)
{
// Final review I4: the controller's type disagrees with the authored one (native type gate, or
// CIP 0xFF/0x2107 from the controller) — deterministic: BadTypeMismatch, keep the handle, leave
// health alone, warn once per tag. Retrying cannot succeed until the config is corrected.
results[i] = new WriteResult(AbCipStatusMapper.BadTypeMismatch);
WarnTypeMismatchOnce(def.Name,
"AbCip write to {Tag} on device {Device} refused: {Reason} (logged once per tag)",
w.FullReference, def.DeviceHostAddress, ate.Message);
}
catch (Exception ex)
{
// Transport / wire error — evict so the next write creates a fresh handle.
@@ -7,9 +7,12 @@ namespace ZB.MOM.WW.OtOpcUa.Driver.AbCip;
/// connection from <see cref="DeviceSessionPool.Shared"/>, so all tags on one device share one CIP
/// connection and one request scheduler. Stateless; safe to share across devices.
/// </summary>
internal sealed class NativeAbCipTagFactory : IAbCipTagFactory
internal sealed class NativeAbCipTagFactory(DeviceSessionPool? pool = null) : IAbCipTagFactory
{
// Tests pass a private pool (and a fake target); production uses the process-wide one.
private readonly DeviceSessionPool _pool = pool ?? DeviceSessionPool.Shared;
/// <inheritdoc />
public IAbCipTagRuntime Create(AbCipTagCreateParams createParams) =>
new NativeAbCipTagRuntime(createParams, DeviceSessionPool.Shared);
new NativeAbCipTagRuntime(createParams, _pool);
}
@@ -0,0 +1,159 @@
using Microsoft.Extensions.Logging;
using Shouldly;
using Xunit;
using ZB.MOM.WW.OtOpcUa.Core.Abstractions;
using ZB.MOM.WW.OtOpcUa.Driver.Ab.Transport;
using ZB.MOM.WW.OtOpcUa.Driver.Ab.Transport.Tests.Fakes;
using CapturingLogger = ZB.MOM.WW.OtOpcUa.Driver.AbCip.Tests.AbCipLoggingTests.CapturingLogger<ZB.MOM.WW.OtOpcUa.Driver.AbCip.AbCipDriver>;
namespace ZB.MOM.WW.OtOpcUa.Driver.AbCip.Tests;
/// <summary>
/// Final review I4: a type mismatch (the controller answered with a type that disagrees with the authored
/// one) is a deterministic configuration refusal, handled like <see cref="AbStatus.Unsupported"/>:
/// <c>BadTypeMismatch</c>, the handle is KEPT (one create across N polls), health is unchanged, and the
/// Warning is logged once per tag — not per poll. Before the fix each poll evicted the runtime, so a sole
/// lease holder cost a Forward Close + Forward Open on every poll and the driver sat Degraded. The
/// control tests (a genuine transport failure still evicts and degrades) live in
/// <see cref="AbCipLocalRefusalTests"/> and below.
/// </summary>
[Trait("Category", "Unit")]
public sealed class AbCipTypeMismatchTests
{
private const string Host = "ab://10.0.0.5/1,0";
private const int Polls = 4;
/// <summary>A fake whose read completes but reports <see cref="AbStatus.TypeMismatch"/> (the native
/// runtime's same-width check), and whose write throws it (the native write gate / CIP 0xFF 0x2107).</summary>
private sealed class MismatchTag(AbCipTagCreateParams p) : FakeAbCipTag(p)
{
public override async Task ReadAsync(CancellationToken cancellationToken)
{
await base.ReadAsync(cancellationToken);
Status = (int)AbStatus.TypeMismatch;
}
public override Task WriteAsync(CancellationToken cancellationToken) =>
Task.FromException(new AbTransportException(AbStatus.TypeMismatch, "REAL on the controller, DINT authored"));
}
private static async Task<(AbCipDriver Driver, FakeAbCipTagFactory Factory, Func<int> Creates, CapturingLogger Log)>
NewAsync()
{
var creates = 0;
var factory = new FakeAbCipTagFactory
{
Customise = p => { Interlocked.Increment(ref creates); return new MismatchTag(p) { Value = 1 }; },
};
var log = new CapturingLogger();
var drv = new AbCipDriver(new AbCipDriverOptions
{
Devices = [new AbCipDeviceOptions(Host)],
RawTags = AbCipRawTags.From(new AbCipTagDefinition("T", Host, "T", AbCipDataType.DInt, Writable: true)),
Probe = new AbCipProbeOptions { Enabled = false },
}, "drv-i4", factory, logger: log);
await drv.InitializeAsync("{}", CancellationToken.None);
return (drv, factory, () => Volatile.Read(ref creates), log);
}
private static int MismatchWarnings(CapturingLogger log) =>
log.Entries.Count(e => e.Level == LogLevel.Warning && e.Message.Contains("logged once per tag"));
#pragma warning disable OTOPCUA0001 // the driver's own status handling is under test, not resilience wrapping
[Fact]
public async Task A_type_mismatched_read_keeps_the_handle_leaves_health_alone_and_warns_once()
{
var (drv, factory, creates, log) = await NewAsync();
var health = drv.GetHealth();
for (var n = 0; n < Polls; n++)
(await drv.ReadAsync(["T"], CancellationToken.None)).Single().StatusCode
.ShouldBe(AbCipStatusMapper.BadTypeMismatch);
creates().ShouldBe(1);
factory.Tags["T"].Disposed.ShouldBeFalse();
factory.Tags["T"].ReadCount.ShouldBe(Polls); // positive control: every poll did reach the runtime
drv.GetHealth().ShouldBe(health);
MismatchWarnings(log).ShouldBe(1);
}
[Fact]
public async Task A_type_mismatched_write_keeps_the_handle_leaves_health_alone_and_warns_once()
{
var (drv, factory, creates, log) = await NewAsync();
var health = drv.GetHealth();
for (var n = 0; n < Polls; n++)
(await drv.WriteAsync([new WriteRequest("T", 5)], CancellationToken.None)).Single().StatusCode
.ShouldBe(AbCipStatusMapper.BadTypeMismatch);
creates().ShouldBe(1);
factory.Tags["T"].Disposed.ShouldBeFalse();
drv.GetHealth().ShouldBe(health);
MismatchWarnings(log).ShouldBe(1);
}
/// <summary>Control: a genuine transport failure on a READ still evicts and degrades.</summary>
[Fact]
public async Task A_transport_failure_read_still_evicts_and_degrades()
{
var factory = new FakeAbCipTagFactory { Customise = p => new FakeAbCipTag(p) { Status = (int)AbStatus.BadConnection } };
var drv = new AbCipDriver(new AbCipDriverOptions
{
Devices = [new AbCipDeviceOptions(Host)],
RawTags = AbCipRawTags.From(new AbCipTagDefinition("T", Host, "T", AbCipDataType.DInt)),
Probe = new AbCipProbeOptions { Enabled = false },
}, "drv-i4", factory);
await drv.InitializeAsync("{}", CancellationToken.None);
(await drv.ReadAsync(["T"], CancellationToken.None)).Single().StatusCode
.ShouldNotBe(AbCipStatusMapper.BadTypeMismatch);
drv.GetHealth().State.ShouldBe(DriverState.Degraded);
factory.Tags["T"].Disposed.ShouldBeTrue();
}
/// <summary>
/// End to end over the REAL native runtime + a private <see cref="DeviceSessionPool"/> against a fake
/// CIP target that answers every read with a REAL for a DINT-authored tag. The tag is the only lease
/// holder (probe off), so an evict would be the LAST release: Forward Close, then a throttled Forward
/// Open on the next poll. One Forward Open across N polls and no Forward Close proves there is no
/// connection churn.
/// </summary>
[Fact]
public async Task A_native_type_mismatch_causes_no_connection_churn()
{
await using var target = new FakeCipTarget();
// Read Tag Fragmented (0x52) → success reply typed REAL (0x00CA) with 4 data bytes.
target.ConnectedHandler = req => Task.FromResult<byte[]?>(
req[0] == 0x52 ? [0xD2, 0x00, 0x00, 0x00, 0xCA, 0x00, 0x00, 0x00, 0x80, 0x3F] : target.ServiceReply(req));
var pool = new DeviceSessionPool();
var host = $"ab://127.0.0.1:{target.Port}/1,0";
var log = new CapturingLogger();
var drv = new AbCipDriver(new AbCipDriverOptions
{
Devices = [new AbCipDeviceOptions(host, AbCipPlcFamily.ControlLogix)],
RawTags = AbCipRawTags.From(new AbCipTagDefinition("T", host, "T", AbCipDataType.DInt)),
Probe = new AbCipProbeOptions { Enabled = false },
Timeout = TimeSpan.FromSeconds(5),
}, "drv-i4-native", new NativeAbCipTagFactory(pool), logger: log);
await drv.InitializeAsync("{}", CancellationToken.None);
var health = drv.GetHealth();
try
{
for (var n = 0; n < Polls; n++)
(await drv.ReadAsync(["T"], CancellationToken.None)).Single().StatusCode
.ShouldBe(AbCipStatusMapper.BadTypeMismatch);
target.ForwardOpens.ShouldBe(1);
target.ForwardCloses.ShouldBe(0);
target.ConnectedRequests.ShouldBeGreaterThanOrEqualTo(Polls); // positive control: polls hit the wire
drv.GetHealth().ShouldBe(health);
MismatchWarnings(log).ShouldBe(1);
}
finally
{
await drv.ShutdownAsync(CancellationToken.None);
}
}
#pragma warning restore OTOPCUA0001
}
@@ -34,6 +34,8 @@
<ItemGroup>
<!-- Native transport test fake, source-linked so it stays internal (Task 12). -->
<Compile Include="..\ZB.MOM.WW.OtOpcUa.Driver.Ab.Transport.Tests\Fakes\FakeCipChannel.cs" Link="Fakes\FakeCipChannel.cs" />
<Compile Include="..\ZB.MOM.WW.OtOpcUa.Driver.Ab.Transport.Tests\Fakes\FakeEipServer.cs" Link="Fakes\FakeEipServer.cs" />
<Compile Include="..\ZB.MOM.WW.OtOpcUa.Driver.Ab.Transport.Tests\Fakes\FakeCipTarget.cs" Link="Fakes\FakeCipTarget.cs" />
</ItemGroup>
</Project>