Files
UA-.NETStandard/docs/RuntimeNodeSets.md
5899354441 Clarify documentation, migration guidance, and SDK selection (#4823)
# Description

Remove stock phrasing and correct misleading or inconsistent technical
documentation. The initial two passes covered 367 Markdown files,
followed by an interactive review of 169 editorial flags. The U5
interpretation originally deferred in that review has since received the
upstream Mantis answer recorded by #4843; remaining server/spec
follow-ups stay explicit in the CTT log.

- Correct package responsibilities, migration steps, API names,
ownership, protocol descriptions and links. Keep bundled migration
guides synchronized. Preserve historical conclusions as historical
records with separate re-check notes.
- Provide at least two rendered TOC levels throughout `docs/`. Convert
appropriate architecture figures to Mermaid without changing code
examples, transcripts or historical quotations.
- Address the October 19 feedback with concrete setup/migration
instructions, corrected certificate paths and PubSub names, shorter
repeated caveats, consistent XML comments, and a full review of the
Concepts guide. Retain #4839's API corrections.
- Integrate #4841's runtime/performance guidance and review
incoming/recent prose. Preserve dated measurements rather than claiming
new benchmark runs; clarify allocation ratios, startup-only processor
settings, JIT warm-up and the benchmark commit's scope. Missing raw
measurement artifacts and repeat ranges remain tracked by #4842.
- Merge master through `e8c5d800a` in `32c6394b7`. Resolve the CTT log
conflict using the updated Mantis states and OPC answers, retaining
historical framing and re-checks for C28/C32/C43.
- Address the latest naming/index feedback in `adb5bf693`: rename the
guides to `docs/WotCon.md` and `docs/WotNodeSetConversion.md`; move the
provider and DataType-declaration contracts into the conversion guide
and remove `docs/WoT.md`. Rename `samples/DI` to `samples/Di` and
`samples/Isa95` to `samples/ISA95`. Add content-index READMEs to Di,
FluentApi, ISA95, Lds and PubSub; update solution entries, project
references, Docker/CI paths, test fixture paths and documentation links
with exact casing.

**WoT guide scope.** This branch retains the detailed conversion guide
that #4758 had replaced with an anchor stub. Conversion limits, verified
document sets, preservation rules and provider ownership now live
together in `WotNodeSetConversion.md`; binding/runtime guidance remains
in `WotBindings.md` and `WotCon.md`. The design-rationale/test-ledger
prose was shortened, the oversized argument-table cell moved into its
section, and existing anchors preserved.

**SDK policy is a separate, explicit user-directed change.**
`global.json` sets a stable minimum of `10.0.400`, `rollForward:
latestMajor`, and `allowPrerelease: false`. Newer major SDKs are
permitted; CI's explicit SDK pins are unchanged. README, GettingStarted,
DeveloperGuide and Recycling instructions reflect the possible analyzer
differences. The image-contract regression fix accepts compatible stable
SDK images instead of requiring exact equality with the floor.

The latest four rename/index threads are resolved after publication.
Three older threads remain open: universal obsolete-removal timing,
unconditional nonce-strength wording, and a blanket Vision NativeAOT
claim. Existing implementation/evidence constraints are explained in
their inline replies. The two-level TOCs remain as requested.

## Validation

- `git diff --check` and staged diff checks: passed. The untouched
upstream AutoID NodeSet retains its provenance bytes, including two
existing license-header trailing spaces.
- `.\.azurepipelines\validate-migration-plugin.ps1`: passed; 15 bundled
migration documents and manifests synchronized.
- Documentation/path validation: **101** documents retain two rendered
TOC levels (**2,228** entries), with idempotent formatting. **52** new
relative links and their anchors passed exact-case checks. No tracked
references to the old guide/sample paths remain. Both relocated
provider-contract sections are preserved in the destination.
- `dotnet sln UA.slnx list`: passed; shows the renamed Di/ISA95 projects
and the existing FluentApi/Lds/PubSub projects. Updated solution and
pump ProjectReference paths also passed exact-case existence checks.
- Latest focused commands, run from the repository root:

```powershell
# Passed: 61 channel send-order and reconnect tests.
dotnet test tests\Opc.Ua.Core.Tests\Opc.Ua.Core.Tests.csproj -p:CustomTestTarget=net10.0 --filter 'FullyQualifiedName~UaSCBinaryChannelSendOrderTests|FullyQualifiedName~TcpReconnectKeyAgreementTests' --verbosity minimal

# Passed: 5 pump datasheet tests; builds the renamed pump sample.
dotnet test tests\Opc.Ua.Di.Tests\Opc.Ua.Di.Tests.csproj -p:CustomTestTarget=net10.0 --filter 'FullyQualifiedName~PumpDatasheetConformanceTests' --verbosity minimal

# Passed: 4 repository NodeSet path/round-trip tests, including the moved pump models.
dotnet test tests\Opc.Ua.Types.Tests\Opc.Ua.Types.Tests.csproj -p:CustomTestTarget=net10.0 --filter 'FullyQualifiedName~RepositorySourceSurvivesSerializedLinkedPartitionsAsync' --verbosity minimal

# Passed: build, zero warnings/errors.
dotnet build samples\ISA95\MinimalIsa95Server\MinimalIsa95Server.csproj -p:CustomTestTarget=net10.0 -f net10.0 --verbosity minimal

# Passed: dependent sample-test project build, zero warnings/errors.
dotnet build tests\Opc.Ua.WotCon.Samples.Tests\Opc.Ua.WotCon.Samples.Tests.csproj -p:CustomTestTarget=net10.0 -f net10.0 --verbosity minimal
```

Earlier focused integration checks passed 2,620 tests across AutoID,
event queues/scheduling, source generation, MCP/package profiles and
WoT. Both reviewed Mermaid diagrams rendered locally without network
access. These checks are not a full solution run.

The full solution has **not** been rerun for this review. Earlier
`dotnet test UA.slnx -p:CustomTestTarget=net10.0 --verbosity quiet` did
not pass: three AOT projects rejected VSTest and 20 Vision tests failed
during graphics initialization with `OutOfMemoryException`. Remaining
long-running tests were stopped at the user's request. A focused Vision
rerun passed 558 tests and reproduced those 20 failures. The earlier
`dotnet test UA.slnx -p:CustomTestTarget=net48 --verbosity quiet` was
blocked by the former SDK pin and was not retried after the policy
change. The user approved proceeding with those limitations documented;
current remote CI is not claimed green.

## Related Issues

- Follow-up #4842 tracks the requested 2.0.2+ benchmark/tuning and
diagnostic-tool documentation refresh. #4841 supplies documentation and
earlier results, not all newly requested measurements.
- #4758 explains the conversion guide's earlier consolidation; the
current provider contracts are now consolidated into
`WotNodeSetConversion.md`.
- #4839 supplies the reviewed API corrections; #4843 supplies the
updated CTT issue record.
- The original prose cleanup and SDK policy were user-requested without
a source issue. No new runtime architecture is introduced by these
documentation edits.

## Checklist

_Put an `x` in the boxes that apply. You can complete these step by step
after opening the PR._

- [ ] I have signed the
[CLA](https://opcfoundation.org/license/cla/ContributorLicenseAgreementv1.0.pdf)
and read the
[CONTRIBUTING](https://github.com/OPCFoundation/UA-.NETStandard/blob/master/CONTRIBUTING.md)
doc.
- [ ] I have added tests that prove my fix is effective or that my
feature works and increased code coverage.
- [x] I have added all necessary documentation.
- [ ] I have verified that my changes do not introduce (new) build or
analyzer warnings.
- [ ] I ran **all** tests locally using the **UA.slnx** solution against
at least .net **framework** and .net **10**, and all passed.
- [ ] I fixed **all** failing and flaky tests in the CI pipelines and
**all** CodeQL warnings.
- [ ] I have addressed **all** PR feedback received.

---------

Co-authored-by: Roman Ettlinger <romanett98@gmail.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-21 10:10:55 +02:00

336 lines
18 KiB
Markdown

# Runtime NodeSets
This guide explains how to import one or more NodeSet2 XML documents into a
server's address space at startup, without writing a source-generated or
hand-coded NodeManager. Configure the files or streams to load. The server
imports them in dependency order and registers the resulting nodes.
Within each imported set, types are registered in inheritance order rather than
XML record order. A derived type may precede its supertype in the document;
an inheritance cycle is rejected before registration. Imported namespace-zero
`InputArguments` and `OutputArguments` Properties populate their Method's typed
signature, retaining the declared Property NodeIds, argument order, and values.
## Contents
- [When to use the runtime NodeSet path](#when-to-use-the-runtime-nodeset-path)
- [Imported security policy](#imported-security-policy)
- [Startup and live lifecycle semantics](#startup-and-live-lifecycle-semantics)
- [Shadow reload](#shadow-reload)
- [Immediate reload](#immediate-reload)
- [Quick-start examples](#quick-start-examples)
- [Single file](#single-file)
- [Single file with a fluent callback](#single-file-with-a-fluent-callback)
- [Group of dependent NodeSets](#group-of-dependent-nodesets)
- [Custom stream source](#custom-stream-source)
- [Direct factory registration](#direct-factory-registration)
- [Stream ownership contract](#stream-ownership-contract)
- [Default namespace for unqualified browse paths](#default-namespace-for-unqualified-browse-paths)
- [Dependency sorting](#dependency-sorting)
- [Referencing a Node another NodeManager owns](#referencing-a-node-another-nodemanager-owns)
- [Complex types](#complex-types)
- [Comparison with source-generated NodeManagers](#comparison-with-source-generated-nodemanagers)
- [Related documentation](#related-documentation)
## When to use the runtime NodeSet path
Use `AddRuntimeNodeSet` when:
- You receive a NodeSet2 XML from a companion-specification vendor and want to host its nodes without regenerating source.
- Your information model changes frequently enough that rebuilding the source-generated manager for every XML update would be disruptive.
- You are prototyping or testing a new NodeSet2 design.
Use the [source-generated path](NodeManagers.md#source-generated-node-managers) when you want compile-time safety, strong typing, and AOT-safe named constants for every node in your model. The runtime path gives you generic `NodeState` objects and untyped browse-path wiring.
## Imported security policy
Runtime import retains node-local `RolePermissions` and `AccessRestrictions`.
When a node omits either attribute, import applies the corresponding default
from its owning `Models/Model` entry. Ownership follows the node's namespace
URI, not its BrowseName, the order of model entries, or the server's namespace
index. Defaults are materialized before the nodes enter the address space, so
ordinary Read, Write, Call and subscription authorization use the same policy.
An explicit empty node ACL or `AccessRestrictions="0"` remains an explicit
local override; it does not inherit the model setting. An omitted model
restriction has the NodeSet schema's default of zero and does not manufacture
a node-local restriction. Normal namespace defaults and application permission
callbacks still apply. Type-hierarchy visibility exemptions are enforced by the
server, not by removing imported policy: metadata remains available for
operations that require authorization.
## Startup and live lifecycle semantics
`AddRuntimeNodeSet` on `IOpcUaServerBuilder` remains the startup path: its factory is created before the server starts and its NodeSet is imported during `CreateAddressSpaceAsync`. After startup, the resulting generation-1 registration is available from `INodeManagerLifecycle.Registrations`, so a model composed at startup can use the same reload and removal APIs as one added while the server is already running.
For example, a startup task can locate the registration by an owned model namespace:
```csharp
public sealed class ModelRegistration(INodeManagerLifecycle lifecycle)
: IServerStartupTask
{
public NodeManagerRegistration? Registration { get; private set; }
public ValueTask OnServerStartedAsync(
IServerContext server,
CancellationToken ct)
{
Registration = lifecycle.Registrations.Find(registration =>
registration.NamespaceUris.Contains("urn:example:MyMachine"));
return default;
}
}
```
Running servers also expose `INodeManagerLifecycle`. Resolve it from dependency injection in a hosted server, or use `StandardServer.NodeManagerLifecycle` when constructing the server directly. The lifecycle provider can add, reload, shadow-reload, and remove runtime NodeSets without restarting the server.
```csharp
public sealed class ModelLoader(INodeManagerLifecycle lifecycle)
{
private NodeManagerRegistration? m_registration;
public async ValueTask LoadAsync(CancellationToken ct)
{
m_registration = await lifecycle.AddRuntimeNodeSetAsync(
new RuntimeNodeSetOptions
{
Sources = [RuntimeNodeSetSource.FromFile("Models/MyMachine.NodeSet2.xml")]
},
callerContext: null,
ct);
}
public async ValueTask ReloadAsync(CancellationToken ct)
{
m_registration = await lifecycle.ReloadRuntimeNodeSetAsync(
m_registration!,
new RuntimeNodeSetOptions
{
Sources = [RuntimeNodeSetSource.FromFile("Models/MyMachine.NodeSet2.xml")]
},
callerContext: null,
ct);
}
public ValueTask RemoveAsync(CancellationToken ct)
{
return lifecycle.RemoveAsync(m_registration!, callerContext: null, ct);
}
}
```
Each add returns an immutable `NodeManagerRegistration`, and reload returns the next generation while invalidating the previous handle.
A registration obtained from `Registrations` for an `AddRuntimeNodeSet` model is already the first
live generation; pass it directly to `ReloadRuntimeNodeSetAsync`,
`ShadowReloadRuntimeNodeSetAsync`, `ImmediateReloadRuntimeNodeSetAsync`, or `RemoveAsync`.
`AddRuntimeNodeSetAsync`, `ReloadRuntimeNodeSetAsync`, and `RemoveAsync` take the operation the caller is running under. Pass `context.GetOperationContext()` when calling from a NodeManager or Method callback: a lifecycle operation drains the requests that are in flight, so one started from inside a request would wait for itself and is rejected with an `InvalidOperationException`. A control-plane caller such as the `ModelLoader` above is not serving a request and passes `null`. See [Registering NodeManagers](NodeManagers.md#runtime-registration).
The rules that apply to every NodeManager registered at runtime -- what happens to MonitoredItems, Browse continuation points, namespace indexes, DataTypes, and change notifications, and which NodeManagers may be reloaded at all -- are described once in [Registering NodeManagers](NodeManagers.md#registering-node-managers). Runtime NodeSets follow those rules, and the built-in runtime NodeSet manager already implements the `INodeManagerReloadParticipant` contract that reload requires.
### Shadow reload
`ShadowReloadRuntimeNodeSetAsync` permits replacing a generation that still owns
active monitored items. Those items stay on the retired generation and continue
to be served there until they are deleted or otherwise released; new requests use
the replacement generation. Normal reload instead migrates compatible monitored items.
```csharp
public async ValueTask ShadowReloadAsync(CancellationToken ct)
{
m_registration = await lifecycle.ShadowReloadRuntimeNodeSetAsync(
m_registration!,
new RuntimeNodeSetOptions
{
Sources = [RuntimeNodeSetSource.FromFile("Models/MyMachine.NodeSet2.xml")]
},
ct);
}
```
The returned registration replaces the old handle for lifecycle operations.
Choose shadow reload only when retaining the old model for existing subscriptions
is intentional; its resources remain allocated until those subscriptions drain.
### Immediate reload
`ImmediateReloadRuntimeNodeSetAsync` does not migrate existing monitored items.
Affected data-change items receive `BadNodeIdUnknown`, and clients may recreate
items against the replacement model.
The [node-manager reload matrix](NodeManagers.md#reload-modes) is authoritative for
all three modes, including request draining, continuation points, and generation
cleanup. Runtime NodeSet reload methods use those same lifecycle operations.
## Quick-start examples
### Single file
```csharp
services.AddOpcUa()
.AddServer(o => { /* … */ })
.AddRuntimeNodeSet("Models/MyMachine.NodeSet2.xml");
```
### Single file with a fluent callback
Wire read/write/method handlers on top of the imported nodes using the existing untyped `INodeManagerBuilder` surface:
```csharp
services.AddOpcUa()
.AddServer(o => { /* … */ })
.AddRuntimeNodeSet(
"Models/MyMachine.NodeSet2.xml",
nodes =>
{
nodes.Node("Machines/Machine1/Start")
.OnCall(StartMachineAsync);
nodes.Variable<double>("Machines/Machine1/Temperature")
.OnRead(ReadTemperature);
});
```
Standard Method argument children are bound to their typed `NodeState` properties during import.
A NodeSet2 Method's `InputArguments` and `OutputArguments` Variables populate
`MethodState.InputArguments` and `MethodState.OutputArguments`, so normal Call argument validation
uses the declarations from the document.
Binding uses the authored `HasProperty` relationship, declared either on the Method or
as an inverse reference on the argument Property; `ParentNodeId` is optional. Local
namespace-URI reference targets are resolved against the import context's namespace table.
The Property must use the namespace-zero standard BrowseName, `PropertyType`, the `Argument`
DataType, and a one-dimensional value rank. Ambiguous or conflicting declarations and
custom or malformed properties do not bind the Method's typed signature; their authored
nodes and references are retained. Other authored reference types are preserved.
### Group of dependent NodeSets
Register multiple NodeSet2 sources that depend on each other. The factory resolves the import order automatically from the `RequiredModel` declarations in each document.
```csharp
services.AddOpcUa()
.AddServer(o => { /* … */ })
.AddRuntimeNodeSet(options =>
{
options.Sources =
[
RuntimeNodeSetSource.FromFile("Models/Opc.Ua.Di.NodeSet2.xml"),
RuntimeNodeSetSource.FromFile("Models/MyMachine.NodeSet2.xml")
];
options.DefaultNamespaceUri = "urn:example:MyMachine";
options.Configure = nodes =>
{
nodes.Node("Machines/Machine1/Start").OnCall(StartMachineAsync);
};
});
```
### Custom stream source
Use `RuntimeNodeSetSource.FromStream` when you want to open the NodeSet2 document lazily — for example from a database, a blob store, or an assembly resource. The delegate is called for each startup load or live generation and must return a fresh, readable `Stream` each time.
```csharp
services.AddOpcUa()
.AddServer(o => { /* … */ })
.AddRuntimeNodeSet(options =>
{
options.Sources =
[
RuntimeNodeSetSource.FromStream(
name: "MyMachine",
openStream: ct => OpenNodeSetStreamAsync(ct),
modelNamespaceUris: ["urn:example:MyMachine"])
];
});
```
### Direct factory registration
If you prefer to create the factory manually — for example to inject it into an existing `StandardServer` subclass — construct a `RuntimeNodeSetNodeManagerFactory` directly:
```csharp
var factory = new RuntimeNodeSetNodeManagerFactory(new RuntimeNodeSetOptions
{
Sources = [RuntimeNodeSetSource.FromFile("Models/MyMachine.NodeSet2.xml")]
});
// In a StandardServer subclass constructor:
AddNodeManager(factory);
```
## Stream ownership contract
When `FromStream` is used, the runtime loader calls the `openStream` delegate while each NodeManager generation is created and closes the returned stream after deserialization. You must ensure that:
1. Each call to `openStream` returns a **new** stream positioned at the beginning.
2. The stream is **readable** and contains a valid NodeSet2 XML document.
3. You do not close the stream yourself; the factory disposes it after `UANodeSet.Read`.
If the delegate returns `null` or the stream does not contain valid NodeSet2 XML, server startup fails with a clear `InvalidOperationException` that names the source.
## Default namespace for unqualified browse paths
When the `Configure` callback uses browse paths without an explicit `ns=N;` prefix (for example `"Machines/Machine1/Start"`), the runtime loader must know which namespace index to apply for the first path segment. The resolution is:
1. If `RuntimeNodeSetOptions.DefaultNamespaceUri` is set, that URI is used.
2. Otherwise the factory infers the **unique leaf model** — the one model in the loaded group that is not required by any other included source.
3. If inference is ambiguous (multiple leaf models and a `Configure` callback is present), startup fails with an error message that lists the candidates. In this case, set `DefaultNamespaceUri` explicitly.
When no `Configure` callback is registered, `DefaultNamespaceUri` has no effect and may be omitted.
## Dependency sorting
The factory reads the `Models/Model/RequiredModel` entries from each parsed NodeSet document and performs a topological sort (Kahn's algorithm) before importing. Import order guarantees that a required model's nodes are in the address space before any document that depends on them imports its nodes.
All documents of one manager are parsed into a single batch whose parent-child relationships are linked exactly once, after the last document has been parsed. A node may therefore declare a `ParentNodeId` that lives in another document of the group regardless of the order the two were parsed in; a duplicate NodeId across documents is rejected while parsing. A fluent `Configure` callback can add further documents to the manager with `builder.Import` — see [Importing a NodeSet2 overlay at runtime](NodeManagers.md#importing-a-nodeset2-overlay-at-runtime--builderimport).
Dependencies on models **not included in the group** — for example the OPC UA base namespace or a third-party model hosted by a generated NodeManager — are silently allowed and treated as external. The server resolves cross-manager references through the normal `AddReverseReferencesAsync` mechanism.
### Referencing a Node another NodeManager owns
A NodeSet may declare a Reference whose other endpoint belongs to a different NodeManager — an inverse `Organizes` placing an Object under a folder some other model defines, for instance. OPC 10000-3 requires such a Reference to be visible from both ends, so the forward edge has to appear on a Node this NodeManager does not own.
That works in both registration orders. At startup the master collects every NodeManager's external references first and applies them afterwards, so creation order does not matter. A NodeManager added **after** startup has no such second phase, so the master retains the startup references and each dynamic NodeManager's references, and **replays** them to any NodeManager registered later. Without that replay a Reference into a NodeManager that did not exist yet would simply be dropped — a reference to an absent Node is discarded rather than queued — leaving the two ends permanently disagreeing: the target browses to the source, the source does not list the target.
Both orders are pinned by `RuntimeNodeSetCrossSourceReferenceTests`.
Cycles among the included sources cause `InvalidOperationException` at startup with an error message that lists the participating sources.
## Complex types
Runtime complex type loading (structures, enumerations, union types) is **on by default** and requires no extra configuration. After all NodeManagers have built their address spaces, `StandardServer.OnNodeManagerStartedAsync` scans every DataType node whose `DataTypeDefinition` attribute is populated and registers a NativeAOT-safe stand-in encodeable in the server's factory. The same stand-ins are reused for client decode/encode via the OPC UA Part 6 binary protocol.
You can tune this behaviour through the existing `ServerComplexTypeOptions`:
```csharp
services.AddOpcUa()
.AddServer(o => { /* … */ })
.AddRuntimeNodeSet("Models/MyMachine.NodeSet2.xml")
.AddComplexTypeSystem(o => o.Enabled = false); // disable if not needed
```
Setting `StandardServer.LoadComplexTypes = false` or disabling the complex type system entirely suppresses the stand-in loading without producing a warning. No second complex-type loading path is introduced by the runtime NodeSet feature.
For a complete description of the server-side complex type system, see [ComplexTypes.md](ComplexTypes.md).
## Comparison with source-generated NodeManagers
| Aspect | Runtime NodeSet (`AddRuntimeNodeSet`) | Source-generated (`[NodeManager]`) |
|--------|--------------------------------------|-------------------------------------|
| Node access in callbacks | Generic `NodeState` / `BaseVariableState` via untyped browse paths | Strongly typed, compiler-checked fluent accessors per node |
| Compilation required on model change | No — reload through `INodeManagerLifecycle` | Yes — regenerate and rebuild |
| AOT / trimming compatibility | Full (uses the existing `UANodeSet.Read` XmlSerializer path) | Full (generated code is static) |
| Named NodeId constants | Not generated | Generated (`Variables.*`, `Objects.*`, etc.) |
| Multiple namespaces in one manager | Yes — group multiple sources | One namespace per generator run |
| DI registration | `AddRuntimeNodeSet(...)` | `AddNodeManager<TFactory>()` |
| Stream / file input | Files and custom stream factories | MSBuild `AdditionalFiles` only |
Source-generated managers are the recommended path for production code where type safety and compile-time validation matter. Runtime NodeSets are the recommended path for rapid prototyping, model-file delivery scenarios, and cases where the XML content changes independently of the server binary.
## Related documentation
- [Source-Generated NodeManagers](NodeManagers.md#source-generated-node-managers) — strongly typed alternative.
- [Dependency Injection](DependencyInjection.md) — `IOpcUaServerBuilder` and service registration.
- [ComplexTypes.md](ComplexTypes.md) — server-side complex type loading and client-side decoding.