fix(twincat): Test Connect probe no longer dials the AmsNetId as an IP (#863)

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
This commit is contained in:
Joseph Doherty
2026-11-11 12:13:55 -05:00
parent e857135e8b
commit fdc9473819
4 changed files with 298 additions and 334 deletions
+4
View File
@@ -268,6 +268,10 @@ host logger factory (wired in `DriverFactoryBootstrap`; the FOCAS and TwinCAT fa
`DriverHealthChanged.Diagnostics` / proto `has_diagnostics`+`diagnostics`; its digest is in the publish fingerprint).
- **TwinCAT (#637):** TIME/TOD publish Float64 ms, DATE/DT a UTC DateTime; a tag authored `UDInt` keeps the raw counter
(and keeps historizing — DateTime is not historized).
- **TwinCAT Test Connect (#863):** the probe opens no socket of its own — `AdsClient.Connect` + `ReadStateAsync`
through the LOCAL AMS router, exactly the driver's path. Never derive a host from an AMS Net ID (it is not an IP;
the retired pre-check dialled its first four octets on the AMS port — a public stranger — and always timed out).
No router ⇒ red "No local AMS router" (no degrade-to-green). `docs/drivers/TestConnectProbes.md`.
- **TwinCAT (#645, #310/#314):** opt-in ADS sum-read (`useSumRead`, `sumReadMaxItems` default 200 / ceiling 500, 32 KiB payload
budget). It reads raw bytes and decodes as the declared type, so only a scalar passing the strict width+kind check
(`CheckNotification`) rides a batch; value-converted pairs (DInt over INT, Real over LREAL/DINT) read singly. Only overall
+24 -17
View File
@@ -46,7 +46,7 @@ with a human-readable explanation rather than a false-green TCP-open tick.
| **S7** | `Plc.OpenAsync` (COTP CR/CC + S7 setup-communication), check `IsConnected`, then `Close`. Wrong rack/slot or a non-S7 server causes `OpenAsync` to throw → handshake fail. | `"S7 connected (CPU …)"` | `10.100.0.35:1102` (python-snap7 sim) |
| **AbCip** | `libplctag` Tag `InitializeAsync` (EIP session + CIP Forward Open). A CIP-level error such as tag-not-found still proves the controller answered CIP → `Ok`. A session/ForwardOpen/connect error → handshake fail. | `"CIP session OK"` | `10.100.0.35:44818` (CIP sim) |
| **AbLegacy** | Same `libplctag` `InitializeAsync` handshake as AbCip, PCCC protocol family. | `"CIP session OK"` (PCCC family) | Deferred — no PLC5/SLC sim |
| **TwinCAT** | `AdsClient.Connect` + `ReadStateAsync`. See [degrade semantics](#twincat-degrade) below. | `"ADS state: {state}"` | Deferred — no ADS target |
| **TwinCAT** | `AdsClient.Connect` + `ReadStateAsync` through the **local AMS router** — no TCP pre-check. See [TwinCAT: no TCP pre-check](#twincat-no-tcp-pre-check) below. | `"ADS state: {state}"` | Z35474 (production TC3 press, read-only, via an in-process TcpRouter) |
| **FOCAS** | `cnc_allclibhndl3` via a direct `DllImport("fwlib32")` in the probe. See [degrade semantics](#focas-degrade) below. | `"FOCAS handle OK"` | Deferred — no CNC + FWLIB |
| **Galaxy** | gRPC unary call to `GalaxyRepository.TestConnection` on the configured mxaccessgw endpoint. See [auth-rejection rule](#galaxy-auth-rejection) below. | `"gateway gRPC OK"` | `http://10.100.0.48:5120` (mxaccessgw) |
| **Heidenhain** | LSV2 `A_LG INSPECT` (password resolved from its `secret:`/`env:` reference) + `R_VR`, through the same read-only allowlist as the driver; a refusal or malformed version record → fail. No logout — the socket close ends the session. | `"LSV2 INSPECT session OK at host:port: <model> NC <version>"` | Deferred — no LSV2 control; simulator-proven |
@@ -61,29 +61,36 @@ backend is now the external HistorianGateway (a gRPC client package, not a probe
## Degrade semantics
Three drivers have environmental constraints that can prevent the handshake
Some drivers have environmental constraints that can prevent the handshake
from running on certain hosts. The **degradation principle** is: the probe
must never produce a result *worse* than today's TCP-only probe. A genuine
protocol rejection from a reachable device is a correct RED; an inability to
*run* the handshake at all (no FWLIB, no managed router) degrades to the
existing TCP-reachability message — still a green tick but annotated.
*run* the handshake at all (no FWLIB) degrades to the
existing TCP-reachability message — still a green tick but annotated. TwinCAT
no longer degrades (see below).
### TwinCAT degrade
### TwinCAT: no TCP pre-check
Where the handshake is available:
TwinCAT is the exception to the "TCP connect fails" row above: it opens **no socket of its own**
(#863). The driver has no IP field — an AMS Net ID is a Beckhoff identifier the AMS router maps to a
host through its route table, and the AMS port (801/851) is not a TCP port. The probe therefore takes
exactly the driver's path (`AdsTwinCATClient.ConnectAsync`): `AdsClient.Connect(netId, port)` through
the local AMS router (TwinCAT XAR, or `Beckhoff.TwinCAT.Ads.TcpRouter` hosted in-process), then one
read-only `ReadStateAsync`, all inside the probe budget. Until #863 a first-phase check dialled the Net
ID's first four octets as an IPv4 address on the AMS port — for Z35474 (`5.118.56.93.1.1`) an
unrelated public host — so Test Connect timed out against every real runtime.
- `AdsClient.Connect(netId, port)` + `ReadStateAsync` → `Ok=true`,
`"ADS state: {state}"` (Run / Config / Stop).
- An ADS **route-table rejection** from a reachable ADS router is a **true RED**:
`"Reachable at {host}:{port} but ADS handshake failed: {detail} — check the
target's ADS route table authorizes this host"`. This is the correct result:
the driver would also be unable to function without an authorized route.
| Outcome | `Ok` | Message |
|---------|------|---------|
| ReadState succeeds | `true` | `"ADS state: {state}"` (Run / Config / Stop) |
| No local AMS router (Beckhoff "Check for a running TwinCAT router instance!", `ClientPortNotOpen`, router not active, ADS client cannot load) | `false` | `"No local AMS router on this host — start TwinCAT XAR or host the Beckhoff TcpRouter; the target was not contacted (…)"` |
| Router answered but cannot reach the target (`TargetMachineNotFound`, `TargetPortNotFound`, port/access refused, host unreachable) | `false` | `"No route to {netId}:{port} or target unreachable: {code} — check the local AMS router has a route …"` |
| Sync/device timeout code, or the budget expired | `false` | `"Probe timed out after {n}s waiting for ADS state from {netId}:{port} …"` |
| Any other ADS error | `false` | `"ADS handshake with {netId}:{port} failed: {code}"` |
Where the handshake is unavailable (headless server, no TwinCAT runtime, the
managed AMS router cannot start):
- Probe degrades to TCP-reachability: `Ok=true`,
`"(ADS handshake unavailable on this host — TCP reachability only)"`.
There is **no degrade-to-green** any more: with no TCP check, a host that cannot run the handshake
has contacted nothing, so it reports the missing router rather than a green tick. A stock admin node
has no AMS router, so TwinCAT Test Connect is red there by design — the driver needs the same router.
### FOCAS degrade
@@ -1,5 +1,4 @@
using System.Diagnostics;
using System.Net.Sockets;
using System.Text.Json;
using System.Text.Json.Serialization;
using TwinCAT.Ads;
@@ -8,59 +7,40 @@ using ZB.MOM.WW.OtOpcUa.Core.Abstractions;
namespace ZB.MOM.WW.OtOpcUa.Driver.TwinCAT;
/// <summary>
/// Two-phase, degrade-guarded Test-Connect probe for the <see cref="TwinCATDriverOptions"/>-shaped
/// driver config. Phase 1: a bare TCP connect to the first device's AMS router host (first four
/// octets of the AMS Net ID) on the AMS port — fast rejection of unreachable targets. Phase 2:
/// a real ADS handshake — <c>AdsClient.Connect(netId, port)</c> + <c>ReadStateAsync</c> — to
/// confirm the endpoint speaks ADS and report the controller's run-state, not just that a TCP
/// socket opened.
/// Test-Connect probe for the <see cref="TwinCATDriverOptions"/>-shaped driver config. It takes
/// exactly the path the driver takes (<see cref="AdsTwinCATClient.ConnectAsync"/>):
/// <c>AdsClient.Connect(netId, port)</c> through the <b>local AMS router</b> (TwinCAT XAR, or
/// <c>Beckhoff.TwinCAT.Ads.TcpRouter</c> hosted in-process), followed by one read-only
/// <c>ReadStateAsync</c>, all inside the caller's time budget.
/// <para>
/// <b>Outcome classification</b> — three cases:
/// <list type="number">
/// <item>
/// <b>ADS connected + ReadState OK</b> → <c>Ok=true</c>, message <c>"ADS state: {AdsState}"</c>
/// (e.g. "Run" / "Config" / "Stop"), with latency.
/// </item>
/// <item>
/// <b>Route/auth rejection from a reachable router</b> — an <see cref="AdsErrorException"/>
/// (or a non-success <c>ReadStateAsync</c> result) whose <see cref="AdsErrorCode"/> means the
/// router answered but won't let us in (e.g. <see cref="AdsErrorCode.TargetPortNotFound"/>,
/// <see cref="AdsErrorCode.TargetMachineNotFound"/>, <see cref="AdsErrorCode.PortNotConnected"/>,
/// <see cref="AdsErrorCode.PortDisabled"/>) → <c>Ok=false</c>, message
/// <c>"Reachable at {host}:{port} but ADS handshake failed: {code} — check the target's ADS
/// route table authorizes this host"</c>. This is a TRUE red: the driver itself also needs
/// the route, so a green tick here would be a false positive.
/// </item>
/// <item>
/// <b>Handshake could not be ATTEMPTED on this host</b> — the managed AMS router cannot run
/// headless (Beckhoff's <c>AdsClient.Connect</c> throws a server exception
/// "Check for a running TwinCAT router instance!"), or a <see cref="PlatformNotSupportedException"/>
/// / <see cref="TypeInitializationException"/> / <see cref="DllNotFoundException"/> /
/// <see cref="NotSupportedException"/> surfaces, or <c>ReadStateAsync</c> reports a client-side
/// port-not-open status → <b>DEGRADE</b>: <c>Ok=true</c>, message
/// <c>"Reachable at {host}:{port} (ADS handshake unavailable on this host — TCP reachability
/// only)"</c>, with latency. The probe NEVER produces a result worse than the old TCP-only
/// probe.
/// </item>
/// <b>No TCP pre-check (#863).</b> The driver has no IP field: an AMS Net ID is a Beckhoff
/// identifier that the router maps to a host through its route table, and the AMS port
/// (801/851/…) is not a TCP port. The probe therefore never opens a socket of its own, and never
/// dials anything derived from the AMS Net ID. The former first-phase check dialled the first four
/// Net ID octets as an IPv4 address on the AMS port — a host unrelated to the target (often a
/// public internet address) on a port ADS does not use — so Test Connect timed out against every
/// real runtime.
/// </para>
/// <para>
/// <b>Outcome classification</b> — <c>Ok=true</c> only when the target answered:
/// <list type="bullet">
/// <item><b>ReadState OK</b> → <c>Ok=true</c>, <c>"ADS state: {AdsState}"</c>, with latency.</item>
/// <item><b>No local AMS router</b> — <c>AdsClient.Connect</c> cannot reach a router on this
/// host (Beckhoff's <c>AdsServerException</c> "Check for a running TwinCAT router
/// instance!"), the client port never opened, or the ADS client cannot load here →
/// <c>Ok=false</c>, "No local AMS router on this host …". The target was not contacted.</item>
/// <item><b>No route / target unreachable</b> — the router answered but has no route to the Net
/// ID, the target refused the route or port, or the router's own connection to the target
/// failed → <c>Ok=false</c>, "No route to … or target unreachable: {code} …".</item>
/// <item><b>Timeout</b> — a sync/device timeout code, or the caller's budget expired →
/// <c>Ok=false</c>, "Probe timed out after {n}s …".</item>
/// <item>Any other ADS error → <c>Ok=false</c>, "ADS handshake with … failed: {code}".</item>
/// </list>
/// </para>
/// </summary>
/// <remarks>
/// The line between case 2 (device-rejected → RED) and case 3 (can't-attempt → DEGRADE) is the
/// crux. Classification rests on the exception's identity: only an <see cref="AdsErrorException"/>
/// (or a result <see cref="AdsErrorCode"/>) carrying a route/target-port code is a RED — that is the
/// ADS router answering and refusing the route. Beckhoff's <c>TwinCAT.Ads.Server.AdsServerException</c>
/// ("running TwinCAT router instance!") derives from plain <see cref="Exception"/>, NOT from
/// <see cref="AdsErrorException"/>, so it is correctly classified as "can't attempt → DEGRADE".
/// When genuinely ambiguous, the probe DEGRADES (Ok=true, TCP-only note) rather than emit a false RED.
/// <para>
/// AMS Net ID format is six dot-separated octets (e.g. <c>192.168.1.10.1.1</c>); the first four
/// are typically the host IPv4 address by Beckhoff convention. The AMS router resolves the real IP
/// route server-side; the probe uses the first-four-octet heuristic for the TCP preflight target,
/// which is reliable for the overwhelming majority of production deployments.
/// </para>
/// The probe is read-only — <c>ReadStateAsync</c> never mutates PLC state — and always disposes the
/// <see cref="AdsClient"/>.
/// The probe is read-only — <c>ReadStateAsync</c> never mutates PLC state, and the probe sends no
/// AddRoute — and always disposes the <see cref="AdsClient"/>.
/// </remarks>
public sealed class TwinCATDriverProbe : IDriverProbe
{
@@ -71,22 +51,57 @@ public sealed class TwinCATDriverProbe : IDriverProbe
Converters = { new JsonStringEnumConverter() },
};
/// <summary>
/// AMS error codes that mean "the router answered but refused the route / target port" — a
/// genuine RED. The driver itself would also be denied, so a green tick would be a false
/// positive. Everything NOT in this set (client-side port/connection errors, sync timeouts,
/// router-not-initialised, etc.) is treated as "couldn't attempt the handshake" → DEGRADE.
/// </summary>
private static readonly HashSet<AdsErrorCode> _routeRejectCodes =
/// <summary>Codes meaning the router on this host is absent, inactive or never gave us a port.</summary>
private static readonly HashSet<AdsErrorCode> _noRouterCodes =
[
AdsErrorCode.ClientPortNotOpen,
AdsErrorCode.ClientNoAmsAddr,
AdsErrorCode.RouterNotInitialized,
AdsErrorCode.RouterNotActive,
];
/// <summary>
/// Codes meaning the router answered but cannot get us to the target: no route, the route or
/// port was refused, or the router's connection to the target host failed.
/// </summary>
private static readonly HashSet<AdsErrorCode> _noRouteCodes =
[
AdsErrorCode.TargetPortNotFound,
AdsErrorCode.TargetMachineNotFound,
AdsErrorCode.TargetPortNotFound,
AdsErrorCode.PortNotConnected,
AdsErrorCode.PortDisabled,
AdsErrorCode.AccessDenied,
AdsErrorCode.DeviceAccessDenied,
AdsErrorCode.HostUnreachable,
AdsErrorCode.WSA_NetworkUnreachable,
AdsErrorCode.WSA_HostDown,
AdsErrorCode.WSA_HostUnreachable,
AdsErrorCode.WSA_HostNotFound,
];
/// <summary>Codes meaning a route exists but the target did not answer in time.</summary>
private static readonly HashSet<AdsErrorCode> _timeoutCodes =
[
AdsErrorCode.ClientSyncTimeOut,
AdsErrorCode.SyncTimeOut,
AdsErrorCode.DeviceTimeOut,
AdsErrorCode.WSA_TimedOut,
AdsErrorCode.RouterFragmentTimeout,
];
/// <summary>
/// The ADS handshake: connect through the local AMS router and read the target's state.
/// Throws <see cref="AdsErrorException"/> for a non-success ReadState result. Swappable for tests.
/// </summary>
private readonly Func<TwinCATAmsAddress, TimeSpan, CancellationToken, Task<AdsState>> _readState;
/// <summary>Initializes the probe with the real ADS handshake through the local AMS router.</summary>
public TwinCATDriverProbe() : this(ReadStateViaLocalRouterAsync) { }
/// <summary>Test seam — substitutes the ADS handshake.</summary>
internal TwinCATDriverProbe(Func<TwinCATAmsAddress, TimeSpan, CancellationToken, Task<AdsState>> readState)
=> _readState = readState;
/// <inheritdoc />
public string DriverType => "TwinCAT";
@@ -98,110 +113,90 @@ public sealed class TwinCATDriverProbe : IDriverProbe
catch (Exception ex) { return new(false, $"Config JSON is invalid: {ex.Message}", null); }
if (opts is null) return new(false, "Config JSON deserialized to null.", null);
var (host, port, parsed) = ExtractTarget(opts);
if (parsed is null || string.IsNullOrWhiteSpace(host) || port <= 0)
return new(false, "Config has no host/port to probe.", null);
var target = TwinCATAmsAddress.TryParse(opts.Devices.FirstOrDefault()?.HostAddress);
if (target is null)
return new(false, "Config has no host/port to probe (expected a device hostAddress ads://{netId}:{port}).", null);
// Phase 1: bare TCP preflight — fast rejection for unreachable hosts. Messages here are
// UNCHANGED from the original TCP-only probe.
var label = $"{target.NetId}:{target.Port}";
var timedOut = $"Probe timed out after {timeout.TotalSeconds:F0}s waiting for ADS state from {label}";
var sw = Stopwatch.StartNew();
try
{
using var socket = new Socket(AddressFamily.InterNetwork, SocketType.Stream, ProtocolType.Tcp);
await socket.ConnectAsync(host, port, ct);
}
catch (SocketException ex)
{
return new(false, $"Connect failed: {ex.SocketErrorCode}", null);
}
catch (OperationCanceledException) when (ct.IsCancellationRequested)
{
return new(false, $"Probe timed out after {timeout.TotalSeconds:F0}s.", null);
}
catch (Exception ex)
{
return new(false, ex.Message, null);
}
// Phase 2: real ADS handshake. Connect + ReadStateAsync. The crux is the three-way
// classification of how this can fail — see the class-doc and ClassifyHandshakeFailure.
var degradeNote = $"Reachable at {host}:{port} (ADS handshake unavailable on this host — TCP reachability only)";
try
{
using var client = new AdsClient();
// Bound the ADS round-trip by the caller's timeout (clamped >=1s, mirrors AdsTwinCATClient).
client.Timeout = (int)Math.Max(1_000, timeout.TotalMilliseconds);
// Connect can throw a server exception ("running TwinCAT router instance!") on a headless
// host with no AMS router — that is a can't-attempt DEGRADE, classified below.
client.Connect(AmsNetId.Parse(parsed.NetId), parsed.Port);
var state = await client.ReadStateAsync(ct).ConfigureAwait(false);
// Task.Run + WaitAsync keep the whole handshake — including the synchronous
// AdsClient.Connect — inside the caller's budget.
var state = await Task.Run(() => _readState(target, timeout, ct), ct)
.WaitAsync(timeout, ct).ConfigureAwait(false);
sw.Stop();
if (state.Succeeded)
return new(true, $"ADS state: {state.State.AdsState}", sw.Elapsed);
// Non-throwing failure carried in the result's error code.
return state.ErrorCode == AdsErrorCode.ClientPortNotOpen
? new(true, degradeNote, sw.Elapsed) // client never opened — DEGRADE
: ClassifyHandshakeFailure(state.ErrorCode, host, port, sw, degradeNote);
return new(true, $"ADS state: {state}", sw.Elapsed);
}
catch (TimeoutException)
{
return new(false, $"{timedOut} — the target did not answer.", null);
}
catch (OperationCanceledException) when (ct.IsCancellationRequested)
{
// Caller timeout — keep the original timed-out message.
return new(false, $"Probe timed out after {timeout.TotalSeconds:F0}s.", null);
return new(false, $"{timedOut}.", null);
}
catch (AdsErrorException ex)
{
// The router answered with an ADS-level error. Route/auth rejection → RED; anything
// else (sync timeout, client port issues, …) → DEGRADE.
return ClassifyHandshakeFailure(ex.ErrorCode, host, port, sw, degradeNote);
return Classify(ex.ErrorCode, label, timedOut);
}
catch (Exception)
catch (Exception ex) when (IsNoRouter(ex))
{
// Everything else — TwinCAT.Ads.Server.AdsServerException ("running TwinCAT router
// instance!"), PlatformNotSupportedException, TypeInitializationException,
// DllNotFoundException, NotSupportedException, etc. — means the handshake could not be
// ATTEMPTED on this host. DEGRADE: never worse than the TCP-only probe.
sw.Stop();
return new(true, degradeNote, sw.Elapsed);
return new(false, NoRouterMessage(ex.Message), null);
}
catch (Exception ex)
{
return new(false, $"ADS handshake with {label} failed: {ex.Message}", null);
}
}
/// <summary>Default handshake — the same calls <see cref="AdsTwinCATClient.ConnectAsync"/> makes.</summary>
private static async Task<AdsState> ReadStateViaLocalRouterAsync(
TwinCATAmsAddress target, TimeSpan timeout, CancellationToken ct)
{
using var client = new AdsClient();
client.Timeout = (int)Math.Max(1_000, timeout.TotalMilliseconds);
client.Connect(AmsNetId.Parse(target.NetId), target.Port);
var state = await client.ReadStateAsync(ct).ConfigureAwait(false);
if (!state.Succeeded)
throw new AdsErrorException($"ReadState failed: {state.ErrorCode}", state.ErrorCode);
return state.State.AdsState;
}
/// <summary>Maps an ADS error code to one of the probe's failure classes.</summary>
internal static DriverProbeResult Classify(AdsErrorCode code, string label, string timedOut)
{
if (_noRouterCodes.Contains(code))
return new(false, NoRouterMessage(code.ToString()), null);
if (_noRouteCodes.Contains(code))
return new(false,
$"No route to {label} or target unreachable: {code} — check the local AMS router has a route to this AMS Net ID and the target's route table authorizes this host.",
null);
if (_timeoutCodes.Contains(code))
return new(false, $"{timedOut} ({code}) — the router has a route but the target did not answer.", null);
return new(false, $"ADS handshake with {label} failed: {code}", null);
}
private static string NoRouterMessage(string detail) =>
$"No local AMS router on this host — start TwinCAT XAR or host the Beckhoff TcpRouter; the target was not contacted ({detail}).";
/// <summary>
/// Classifies an ADS-level failure (from an <see cref="AdsErrorException"/> or a non-success
/// <c>ReadStateAsync</c> result). A route/target-port/access code means the router answered
/// but refused the route → RED. Any other code is treated as "couldn't attempt" → DEGRADE,
/// so the probe never under-reports a host with no usable ADS runtime.
/// True when the ADS client could not reach a router on this host, or cannot load here at all.
/// Beckhoff's <c>AdsServerException</c> / <c>AmsServerException</c> ("Check for a running TwinCAT
/// router instance!") derive from plain <see cref="Exception"/>, so they are matched by type name.
/// </summary>
private static DriverProbeResult ClassifyHandshakeFailure(
AdsErrorCode code, string host, int port, Stopwatch sw, string degradeNote)
private static bool IsNoRouter(Exception ex)
{
if (_routeRejectCodes.Contains(code))
return new(false,
$"Reachable at {host}:{port} but ADS handshake failed: {code} — check the target's ADS route table authorizes this host",
null);
sw.Stop();
return new(true, degradeNote, sw.Elapsed);
}
private static (string host, int port, TwinCATAmsAddress? parsed) ExtractTarget(TwinCATDriverOptions opts)
{
// Parse the first device's ads:// address. AMS Net ID is six-octet; by Beckhoff
// convention the first four octets are the host IPv4. Extract those as the TCP target.
var firstDevice = opts.Devices.FirstOrDefault();
if (firstDevice is null) return (string.Empty, 0, null);
var parsed = TwinCATAmsAddress.TryParse(firstDevice.HostAddress);
if (parsed is null) return (string.Empty, 0, null);
// NetId = "a.b.c.d.e.f" — take the first 4 octets as the host IP.
var parts = parsed.NetId.Split('.');
if (parts.Length < 4) return (string.Empty, 0, null);
var hostIp = string.Join('.', parts[0], parts[1], parts[2], parts[3]);
return (hostIp, parsed.Port, parsed);
for (var e = ex; e is not null; e = e.InnerException)
{
if (e is PlatformNotSupportedException or TypeInitializationException or DllNotFoundException)
return true;
var name = e.GetType().FullName;
if (name is "TwinCAT.Ads.Server.AdsServerException" or "TwinCAT.Ams.AmsServerException")
return true;
}
return false;
}
}
@@ -1,22 +1,17 @@
// Unit tests for TwinCATDriverProbe — the degrade-guarded ADS ReadState Test-Connect probe.
// Unit tests for TwinCATDriverProbe — the ADS ReadState Test-Connect probe (#863).
//
// Coverage here is the set of outcomes that are *deterministic on a CI/dev box with no TwinCAT
// runtime*: invalid JSON, missing host/port, an unreachable TCP target, and — the crux — the
// DEGRADE path. On this box there is no AMS router, so once the TCP preflight succeeds against a
// reachable listener the managed ADS client cannot initialise (it throws a "Check for a running
// TwinCAT router instance!" server exception, NOT an ADS device rejection). The probe MUST treat
// that as "TCP reachability only" (Ok=true) so it never regresses below the old TCP-only probe.
//
// The two paths that REQUIRE a live ADS target are LIVE-VERIFY DEFERRED (no TwinCAT runtime on the
// rig): (a) the happy "ADS state: Run/Config/Stop" path (Ok=true with the AdsState name), and
// (b) the route/auth-reject RED path (a reachable router that refuses the route, surfaced as an
// AdsErrorException carrying a TargetPortNotFound / route AdsErrorCode → Ok=false with the
// "check the target's ADS route table" message). Those are exercised against a real TwinCAT XAR /
// remote runtime, out of scope for this unit suite.
// The probe takes exactly the driver's path: AdsClient.Connect(netId, port) through the LOCAL AMS
// router, then one ReadStateAsync. It must never open a socket of its own, and never to a host derived
// from the AMS Net ID (the old first-phase check dialled the Net ID's first four octets on the AMS
// port — an unrelated host, often a public one, on a port ADS does not use — so Test Connect timed out
// against every real runtime). The classification and budget tests drive the handshake through the
// internal seam; the no-dial and no-router tests use the real AdsClient on a box with no AMS router.
using System.Diagnostics;
using System.Net;
using System.Net.Sockets;
using Shouldly;
using TwinCAT.Ads;
using Xunit;
namespace ZB.MOM.WW.OtOpcUa.Driver.TwinCAT.Tests;
@@ -24,199 +19,162 @@ namespace ZB.MOM.WW.OtOpcUa.Driver.TwinCAT.Tests;
[Trait("Category", "Unit")]
public sealed class TwinCATDriverProbeTests
{
private static readonly TwinCATDriverProbe Probe = new();
private const string Z35474Like = "{\"devices\":[{\"hostAddress\":\"ads://5.118.56.93.1.1:851\"}]}";
// -------------------------------------------------------------------------
// 1. Invalid JSON
// -------------------------------------------------------------------------
private static TwinCATDriverProbe ProbeWith(Func<TwinCATAmsAddress, TimeSpan, CancellationToken, Task<AdsState>> handshake)
=> new(handshake);
// ---- config validation ----------------------------------------------------------------
/// <summary>Invalid JSON returns Ok=false with a message containing "invalid".</summary>
[Fact]
public async Task InvalidJson_Returns_OkFalse_WithInvalidMessage()
{
var result = await Probe.ProbeAsync(
"not-valid-json{{{",
TimeSpan.FromSeconds(3),
TestContext.Current.CancellationToken);
var result = await new TwinCATDriverProbe().ProbeAsync(
"not-valid-json{{{", TimeSpan.FromSeconds(3), TestContext.Current.CancellationToken);
result.Ok.ShouldBeFalse();
result.Message.ShouldNotBeNull();
result.Message!.ToLowerInvariant().ShouldContain("invalid");
result.Latency.ShouldBeNull();
}
// -------------------------------------------------------------------------
// 2. Config with no device / no host
// -------------------------------------------------------------------------
/// <summary>An empty Devices list returns Ok=false with a "no host/port" message.</summary>
[Fact]
public async Task NoDevices_Returns_OkFalse_WithNoHostPortMessage()
[Theory]
[InlineData("{\"devices\":[]}")]
[InlineData("{\"devices\":[{\"hostAddress\":\"not-an-ads-url\"}]}")]
public async Task NoUsableDevice_Returns_OkFalse_WithNoHostPortMessage(string json)
{
var result = await Probe.ProbeAsync(
"{\"devices\":[]}",
TimeSpan.FromSeconds(3),
TestContext.Current.CancellationToken);
var called = false;
var result = await ProbeWith((_, _, _) => { called = true; return Task.FromResult(AdsState.Run); })
.ProbeAsync(json, TimeSpan.FromSeconds(3), TestContext.Current.CancellationToken);
result.Ok.ShouldBeFalse();
result.Message.ShouldNotBeNull();
result.Message!.ShouldContain("no host/port");
result.Latency.ShouldBeNull();
called.ShouldBeFalse();
}
/// <summary>
/// A first device with a malformed (non-<c>ads://</c>) host address returns Ok=false with a
/// "no host/port" message because <see cref="TwinCATAmsAddress.TryParse"/> returns null.
/// </summary>
[Fact]
public async Task MalformedHostAddress_Returns_OkFalse_WithNoHostPortMessage()
{
var result = await Probe.ProbeAsync(
"{\"devices\":[{\"hostAddress\":\"not-an-ads-url\"}]}",
TimeSpan.FromSeconds(3),
TestContext.Current.CancellationToken);
result.Ok.ShouldBeFalse();
result.Message.ShouldNotBeNull();
result.Message!.ShouldContain("no host/port");
result.Latency.ShouldBeNull();
}
// -------------------------------------------------------------------------
// 3. Unreachable target — TCP preflight fails first
// -------------------------------------------------------------------------
// ---- #863: never dial a host derived from the AMS Net ID --------------------------------
/// <summary>
/// A closed port causes the TCP preflight <c>ConnectAsync</c> to throw
/// <see cref="SocketException"/>; the probe returns Ok=false with "Connect failed".
/// The port is reserved then released so the OS refuses the connection immediately.
/// Regression for #863. A listener sits on exactly the endpoint the old probe dialled
/// (Net ID octets 1–4 as IPv4, AMS port as TCP port). The real probe must never connect to
/// it — the old TCP pre-check connected here first, so this fails on the old behaviour.
/// </summary>
[Fact]
public async Task ClosedPort_Returns_OkFalse_WithConnectFailedMessage()
public async Task Probe_NeverConnects_ToAmsNetIdDerivedEndpoint()
{
// Reserve an ephemeral port then release it so the port is definitely closed.
var reserved = new TcpListener(IPAddress.Loopback, 0);
reserved.Start();
var port = ((IPEndPoint)reserved.LocalEndpoint).Port;
reserved.Stop();
// 127.0.0.1 as the first four AMS-Net-ID octets → the probe's TCP target is 127.0.0.1.
var configJson = $"{{\"devices\":[{{\"hostAddress\":\"ads://127.0.0.1.1.1:{port}\"}}]}}";
using var cts = CancellationTokenSource.CreateLinkedTokenSource(
TestContext.Current.CancellationToken);
cts.CancelAfter(TimeSpan.FromSeconds(5));
var result = await Probe.ProbeAsync(
configJson,
TimeSpan.FromSeconds(3),
cts.Token);
result.Ok.ShouldBeFalse();
result.Message.ShouldNotBeNull();
// Either a fast SocketException ("Connect failed") or, if the OS stalls, a timeout.
var lower = result.Message!.ToLowerInvariant();
(lower.Contains("connect failed") || lower.Contains("timed out")).ShouldBeTrue(
$"Expected 'Connect failed' or 'timed out' but got: {result.Message}");
result.Latency.ShouldBeNull();
}
/// <summary>
/// A target on a non-routable IP (192.0.2.1 — TEST-NET-1, RFC 5737) with a short
/// cancellation window times out (or fails fast on EHOSTUNREACH) at the TCP preflight.
/// </summary>
[Fact]
public async Task BlackHoleTarget_Returns_OkFalse_WithTimedOutOrConnectFailed()
{
using var cts = CancellationTokenSource.CreateLinkedTokenSource(
TestContext.Current.CancellationToken);
cts.CancelAfter(TimeSpan.FromSeconds(1));
var configJson = "{\"devices\":[{\"hostAddress\":\"ads://192.0.2.1.1.1:851\"}]}";
var result = await Probe.ProbeAsync(
configJson,
TimeSpan.FromSeconds(1),
cts.Token);
result.Ok.ShouldBeFalse();
result.Message.ShouldNotBeNull();
var lower = result.Message!.ToLowerInvariant();
(lower.Contains("timed out") || lower.Contains("connect failed")).ShouldBeTrue(
$"Expected 'timed out' or 'Connect failed' but got: {result.Message}");
result.Latency.ShouldBeNull();
}
// -------------------------------------------------------------------------
// 4. DEGRADE path — TCP reachable, but the ADS handshake can't be attempted here
// -------------------------------------------------------------------------
/// <summary>
/// The crux. A reachable TCP listener satisfies the preflight, but on this box (no TwinCAT
/// AMS router) the managed ADS client cannot establish a session — <c>AdsClient.Connect</c>
/// throws an environment/router exception (NOT an ADS device rejection). The probe MUST
/// DEGRADE: <c>Ok=true</c> with the "TCP reachability only" note, never a false RED. This
/// guarantees the new probe never reports worse than the old TCP-only probe on a host with
/// no ADS runtime.
/// </summary>
[Fact]
public async Task ReachableTcp_NoAdsRuntime_Degrades_To_OkTrue_WithTcpOnlyNote()
{
// Stand up a real TCP listener and accept (then drop) connections so the preflight
// ConnectAsync succeeds. We never speak ADS — the ADS handshake is what must degrade.
var listener = new TcpListener(IPAddress.Loopback, 0);
listener.Start();
var port = ((IPEndPoint)listener.LocalEndpoint).Port;
using var acceptCts = CancellationTokenSource.CreateLinkedTokenSource(
TestContext.Current.CancellationToken);
var acceptLoop = Task.Run(async () =>
{
try
{
while (true)
{
var client = await listener.AcceptTcpClientAsync(acceptCts.Token);
// Hold the socket briefly so the preflight sees a clean accept, then close.
_ = Task.Run(async () =>
{
await Task.Delay(200);
client.Dispose();
});
}
}
catch (OperationCanceledException) { /* accept cancelled — expected on teardown */ }
catch (ObjectDisposedException) { /* listener stopped — expected on teardown */ }
catch (SocketException) { /* listener stopped — expected on teardown */ }
}, TestContext.Current.CancellationToken);
try
{
var configJson = $"{{\"devices\":[{{\"hostAddress\":\"ads://127.0.0.1.1.1:{port}\"}}]}}";
using var cts = CancellationTokenSource.CreateLinkedTokenSource(
TestContext.Current.CancellationToken);
using var cts = CancellationTokenSource.CreateLinkedTokenSource(TestContext.Current.CancellationToken);
cts.CancelAfter(TimeSpan.FromSeconds(10));
await new TwinCATDriverProbe().ProbeAsync(
$"{{\"devices\":[{{\"hostAddress\":\"ads://127.0.0.1.1.1:{port}\"}}]}}",
TimeSpan.FromSeconds(3), cts.Token);
var result = await Probe.ProbeAsync(
configJson,
TimeSpan.FromSeconds(3),
cts.Token);
// DEGRADE guard: must be Ok=true (never worse than TCP-only), with the explicit note
// and a latency. If on some box AdsClient instead threw synchronously at construction,
// the probe still degrades the same way — this assertion holds either way.
result.Ok.ShouldBeTrue(
$"Degrade guard violated — probe returned a worse-than-TCP result: {result.Message}");
result.Message.ShouldNotBeNull();
result.Message!.ShouldContain("ADS handshake unavailable on this host");
result.Message!.ShouldContain("TCP reachability only");
result.Latency.ShouldNotBeNull();
listener.Pending().ShouldBeFalse(
"the probe opened a TCP connection to the AmsNetId-derived address — it must only use the local AMS router");
}
finally
{
await acceptCts.CancelAsync();
listener.Stop();
try { await acceptLoop; } catch { /* ignore teardown races */ }
}
}
/// <summary>The handshake receives the AMS address unchanged — the probe resolves no host itself.</summary>
[Fact]
public async Task Handshake_IsHandedTheAmsAddress_Unchanged()
{
TwinCATAmsAddress? seen = null;
var result = await ProbeWith((t, _, _) => { seen = t; return Task.FromResult(AdsState.Run); })
.ProbeAsync(Z35474Like, TimeSpan.FromSeconds(3), TestContext.Current.CancellationToken);
result.Ok.ShouldBeTrue();
seen.ShouldBe(new TwinCATAmsAddress("5.118.56.93.1.1", 851));
}
// ---- outcome classification -------------------------------------------------------------
[Fact]
public async Task ReadStateOk_Returns_OkTrue_WithStateAndLatency()
{
var result = await ProbeWith((_, _, _) => Task.FromResult(AdsState.Run))
.ProbeAsync(Z35474Like, TimeSpan.FromSeconds(3), TestContext.Current.CancellationToken);
result.Ok.ShouldBeTrue();
result.Message.ShouldBe("ADS state: Run");
result.Latency.ShouldNotBeNull();
}
/// <summary>
/// On a box with no AMS router the real AdsClient cannot connect. That is a RED naming the
/// missing router — no longer a green "TCP reachability only", since nothing was contacted.
/// </summary>
[Fact]
public async Task NoLocalRouter_RealClient_Returns_OkFalse_NamingTheRouter()
{
var result = await new TwinCATDriverProbe().ProbeAsync(
Z35474Like, TimeSpan.FromSeconds(3), TestContext.Current.CancellationToken);
result.Ok.ShouldBeFalse();
result.Message!.ShouldContain("No local AMS router");
result.Message!.ShouldContain("not contacted");
result.Latency.ShouldBeNull();
}
[Fact]
public async Task AmsServerException_Classified_AsNoLocalRouter()
{
var result = await ProbeWith((_, _, _) => throw new global::TwinCAT.Ams.AmsServerException("Connect to Server failed!"))
.ProbeAsync(Z35474Like, TimeSpan.FromSeconds(3), TestContext.Current.CancellationToken);
result.Ok.ShouldBeFalse();
result.Message!.ShouldContain("No local AMS router");
}
[Theory]
[InlineData(AdsErrorCode.ClientPortNotOpen, "No local AMS router")]
[InlineData(AdsErrorCode.TargetMachineNotFound, "No route to 5.118.56.93.1.1:851 or target unreachable")]
[InlineData(AdsErrorCode.TargetPortNotFound, "No route to 5.118.56.93.1.1:851 or target unreachable")]
[InlineData(AdsErrorCode.WSA_HostUnreachable, "No route to 5.118.56.93.1.1:851 or target unreachable")]
[InlineData(AdsErrorCode.ClientSyncTimeOut, "timed out")]
[InlineData(AdsErrorCode.DeviceInvalidGroup, "ADS handshake with 5.118.56.93.1.1:851 failed")]
public async Task AdsErrorCodes_AreClassified(AdsErrorCode code, string expected)
{
var result = await ProbeWith((_, _, _) => throw new AdsErrorException("x", code))
.ProbeAsync(Z35474Like, TimeSpan.FromSeconds(3), TestContext.Current.CancellationToken);
result.Ok.ShouldBeFalse();
result.Message!.ShouldContain(expected);
result.Message!.ShouldContain(code.ToString());
result.Latency.ShouldBeNull();
}
// ---- time budget --------------------------------------------------------------------------
/// <summary>A handshake that never returns (and ignores its token) is cut off at the budget.</summary>
[Fact]
public async Task HangingHandshake_IsBoundedByTimeout()
{
var never = new TaskCompletionSource<AdsState>();
var sw = Stopwatch.StartNew();
var result = await ProbeWith((_, _, _) => never.Task)
.ProbeAsync(Z35474Like, TimeSpan.FromSeconds(1), TestContext.Current.CancellationToken);
sw.Stop();
result.Ok.ShouldBeFalse();
result.Message!.ShouldContain("timed out after 1s");
sw.Elapsed.ShouldBeLessThan(TimeSpan.FromSeconds(5));
}
[Fact]
public async Task CallerCancellation_Returns_TimedOut()
{
using var cts = new CancellationTokenSource(TimeSpan.FromMilliseconds(200));
var result = await ProbeWith(async (_, _, ct) => { await Task.Delay(Timeout.Infinite, ct); return AdsState.Run; })
.ProbeAsync(Z35474Like, TimeSpan.FromSeconds(10), cts.Token);
result.Ok.ShouldBeFalse();
result.Message!.ShouldContain("Probe timed out after 10s");
}
}