Merge pull request 'Driver-instance bootstrap pipeline (#248) — DriverInstance rows materialise as live IDriver instances' (#196) from phase-7-fu-248-driver-bootstrap into v2

.gitignore

@@ -30,3 +30,6 @@ packages/
 .claude/
 
 .local/
+
+# LiteDB local config cache (Phase 6.1 Stream D — runtime artifact, not source)
+src/ZB.MOM.WW.OtOpcUa.Server/config_cache.db

src/ZB.MOM.WW.OtOpcUa.Core/Hosting/DriverFactoryRegistry.cs

@@ -0,0 +1,64 @@
+using ZB.MOM.WW.OtOpcUa.Core.Abstractions;
+
+namespace ZB.MOM.WW.OtOpcUa.Core.Hosting;
+
+/// <summary>
+///     Process-singleton registry of <see cref="IDriver"/> factories keyed by
+///     <c>DriverInstance.DriverType</c> string. Each driver project ships a DI
+///     extension (e.g. <c>services.AddGalaxyProxyDriverFactory()</c>) that registers
+///     its factory at startup; the bootstrapper looks up the factory by
+///     <c>DriverInstance.DriverType</c> + invokes it with the row's
+///     <c>DriverInstanceId</c> + <c>DriverConfig</c> JSON.
+/// </summary>
+/// <remarks>
+///     Closes the gap surfaced by task #240 live smoke — DriverInstance rows in
+///     the central config DB had no path to materialise as registered <see cref="IDriver"/>
+///     instances. The factory registry is the seam.
+/// </remarks>
+public sealed class DriverFactoryRegistry
+{
+    private readonly Dictionary<string, Func<string, string, IDriver>> _factories
+        = new(StringComparer.OrdinalIgnoreCase);
+    private readonly object _lock = new();
+
+    /// <summary>
+    ///     Register a factory for <paramref name="driverType"/>. Throws if a factory is
+    ///     already registered for that type — drivers are singletons by type-name in
+    ///     this process.
+    /// </summary>
+    /// <param name="driverType">Matches <c>DriverInstance.DriverType</c>.</param>
+    /// <param name="factory">
+    ///     Receives <c>(driverInstanceId, driverConfigJson)</c>; returns a new
+    ///     <see cref="IDriver"/>. Must NOT call <see cref="IDriver.InitializeAsync"/>
+    ///     itself — the bootstrapper calls it via <see cref="DriverHost.RegisterAsync"/>
+    ///     so the host's per-driver retry semantics apply uniformly.
+    /// </param>
+    public void Register(string driverType, Func<string, string, IDriver> factory)
+    {
+        ArgumentException.ThrowIfNullOrWhiteSpace(driverType);
+        ArgumentNullException.ThrowIfNull(factory);
+        lock (_lock)
+        {
+            if (_factories.ContainsKey(driverType))
+                throw new InvalidOperationException(
+                    $"DriverType '{driverType}' factory already registered for this process");
+            _factories[driverType] = factory;
+        }
+    }
+
+    /// <summary>
+    ///     Try to look up the factory for <paramref name="driverType"/>. Returns null
+    ///     if no driver assembly registered one — bootstrapper logs + skips so a
+    ///     missing-assembly deployment doesn't take down the whole server.
+    /// </summary>
+    public Func<string, string, IDriver>? TryGet(string driverType)
+    {
+        ArgumentException.ThrowIfNullOrWhiteSpace(driverType);
+        lock (_lock) return _factories.GetValueOrDefault(driverType);
+    }
+
+    public IReadOnlyCollection<string> RegisteredTypes
+    {
+        get { lock (_lock) return [.. _factories.Keys]; }
+    }
+}

src/ZB.MOM.WW.OtOpcUa.Driver.Galaxy.Proxy/GalaxyProxyDriverFactoryExtensions.cs

@@ -0,0 +1,59 @@
+using System.Text.Json;
+using ZB.MOM.WW.OtOpcUa.Core.Hosting;
+
+namespace ZB.MOM.WW.OtOpcUa.Driver.Galaxy.Proxy;
+
+/// <summary>
+///     Static factory registration helper for <see cref="GalaxyProxyDriver"/>. Server's
+///     Program.cs calls <see cref="Register"/> once at startup; the bootstrapper (task #248)
+///     then materialises Galaxy DriverInstance rows from the central config DB into live
+///     driver instances. No dependency on Microsoft.Extensions.DependencyInjection so the
+///     driver project stays free of DI machinery.
+/// </summary>
+public static class GalaxyProxyDriverFactoryExtensions
+{
+    public const string DriverTypeName = "Galaxy";
+
+    /// <summary>
+    ///     Register the Galaxy driver factory in the supplied <see cref="DriverFactoryRegistry"/>.
+    ///     Throws if 'Galaxy' is already registered — single-instance per process.
+    /// </summary>
+    public static void Register(DriverFactoryRegistry registry)
+    {
+        ArgumentNullException.ThrowIfNull(registry);
+        registry.Register(DriverTypeName, CreateInstance);
+    }
+
+    internal static GalaxyProxyDriver CreateInstance(string driverInstanceId, string driverConfigJson)
+    {
+        ArgumentException.ThrowIfNullOrWhiteSpace(driverInstanceId);
+        ArgumentException.ThrowIfNullOrWhiteSpace(driverConfigJson);
+
+        // DriverConfig column is a JSON object that mirrors GalaxyProxyOptions.
+        // Required: PipeName, SharedSecret. Optional: ConnectTimeoutMs (defaults to 10s).
+        // The DriverInstanceId from the row wins over any value in the JSON — the row
+        // is the authoritative identity per the schema's UX_DriverInstance_Generation_LogicalId.
+        using var doc = JsonDocument.Parse(driverConfigJson);
+        var root = doc.RootElement;
+
+        string pipeName = root.TryGetProperty("PipeName", out var p) && p.ValueKind == JsonValueKind.String
+            ? p.GetString()!
+            : throw new InvalidOperationException(
+                $"GalaxyProxyDriver config for '{driverInstanceId}' missing required PipeName");
+        string sharedSecret = root.TryGetProperty("SharedSecret", out var s) && s.ValueKind == JsonValueKind.String
+            ? s.GetString()!
+            : throw new InvalidOperationException(
+                $"GalaxyProxyDriver config for '{driverInstanceId}' missing required SharedSecret");
+        var connectTimeout = root.TryGetProperty("ConnectTimeoutMs", out var t) && t.ValueKind == JsonValueKind.Number
+            ? TimeSpan.FromMilliseconds(t.GetInt32())
+            : TimeSpan.FromSeconds(10);
+
+        return new GalaxyProxyDriver(new GalaxyProxyOptions
+        {
+            DriverInstanceId = driverInstanceId,
+            PipeName = pipeName,
+            SharedSecret = sharedSecret,
+            ConnectTimeout = connectTimeout,
+        });
+    }
+}

src/ZB.MOM.WW.OtOpcUa.Driver.Galaxy.Proxy/ZB.MOM.WW.OtOpcUa.Driver.Galaxy.Proxy.csproj

@@ -13,6 +13,7 @@
 
     <ItemGroup>
         <ProjectReference Include="..\ZB.MOM.WW.OtOpcUa.Core.Abstractions\ZB.MOM.WW.OtOpcUa.Core.Abstractions.csproj"/>
+        <ProjectReference Include="..\ZB.MOM.WW.OtOpcUa.Core\ZB.MOM.WW.OtOpcUa.Core.csproj"/>
         <ProjectReference Include="..\ZB.MOM.WW.OtOpcUa.Driver.Galaxy.Shared\ZB.MOM.WW.OtOpcUa.Driver.Galaxy.Shared.csproj"/>
         <ProjectReference Include="..\ZB.MOM.WW.OtOpcUa.Core.AlarmHistorian\ZB.MOM.WW.OtOpcUa.Core.AlarmHistorian.csproj"/>
     </ItemGroup>

src/ZB.MOM.WW.OtOpcUa.Server/DriverInstanceBootstrapper.cs

@@ -0,0 +1,88 @@
+using Microsoft.EntityFrameworkCore;
+using Microsoft.Extensions.DependencyInjection;
+using Microsoft.Extensions.Logging;
+using ZB.MOM.WW.OtOpcUa.Configuration;
+using ZB.MOM.WW.OtOpcUa.Core.Hosting;
+
+namespace ZB.MOM.WW.OtOpcUa.Server;
+
+/// <summary>
+///     Task #248 — bridges the gap surfaced by the Phase 7 live smoke (#240) where
+///     <c>DriverInstance</c> rows in the central config DB had no path to materialise
+///     as live <see cref="Core.Abstractions.IDriver"/> instances in <see cref="DriverHost"/>.
+///     Called from <c>OpcUaServerService.ExecuteAsync</c> after the bootstrap loads
+///     the published generation, before address-space build.
+/// </summary>
+/// <remarks>
+///     <para>
+///         Per row: looks up the <c>DriverType</c> string in
+///         <see cref="DriverFactoryRegistry"/>, calls the factory with the row's
+///         <c>DriverInstanceId</c> + <c>DriverConfig</c> JSON to construct an
+///         <see cref="Core.Abstractions.IDriver"/>, then registers via
+///         <see cref="DriverHost.RegisterAsync"/> which invokes <c>InitializeAsync</c>
+///         under the host's lifecycle semantics.
+///     </para>
+///     <para>
+///         Unknown <c>DriverType</c> = factory not registered = log a warning and skip.
+///         Per plan decision #12 (driver isolation), failure to construct or initialize
+///         one driver doesn't prevent the rest from coming up — the Server keeps serving
+///         the others' subtrees + the operator can fix the misconfigured row + republish
+///         to retry.
+///     </para>
+/// </remarks>
+public sealed class DriverInstanceBootstrapper(
+    DriverFactoryRegistry factories,
+    DriverHost driverHost,
+    IServiceScopeFactory scopeFactory,
+    ILogger<DriverInstanceBootstrapper> logger)
+{
+    public async Task<int> RegisterDriversFromGenerationAsync(long generationId, CancellationToken ct)
+    {
+        using var scope = scopeFactory.CreateScope();
+        var db = scope.ServiceProvider.GetRequiredService<OtOpcUaConfigDbContext>();
+
+        var rows = await db.DriverInstances.AsNoTracking()
+            .Where(d => d.GenerationId == generationId && d.Enabled)
+            .ToListAsync(ct).ConfigureAwait(false);
+
+        var registered = 0;
+        var skippedUnknownType = 0;
+        var failedInit = 0;
+
+        foreach (var row in rows)
+        {
+            var factory = factories.TryGet(row.DriverType);
+            if (factory is null)
+            {
+                logger.LogWarning(
+                    "DriverInstance {Id} skipped — DriverType '{Type}' has no registered factory (known: {Known})",
+                    row.DriverInstanceId, row.DriverType, string.Join(",", factories.RegisteredTypes));
+                skippedUnknownType++;
+                continue;
+            }
+
+            try
+            {
+                var driver = factory(row.DriverInstanceId, row.DriverConfig);
+                await driverHost.RegisterAsync(driver, row.DriverConfig, ct).ConfigureAwait(false);
+                registered++;
+                logger.LogInformation(
+                    "DriverInstance {Id} ({Type}) registered + initialized", row.DriverInstanceId, row.DriverType);
+            }
+            catch (Exception ex)
+            {
+                // Plan decision #12 — driver isolation. Log + continue so one bad row
+                // doesn't deny the OPC UA endpoint to the rest of the fleet.
+                logger.LogError(ex,
+                    "DriverInstance {Id} ({Type}) failed to initialize — driver state will reflect Faulted; operator can republish to retry",
+                    row.DriverInstanceId, row.DriverType);
+                failedInit++;
+            }
+        }
+
+        logger.LogInformation(
+            "DriverInstanceBootstrapper: gen={Gen} registered={Registered} skippedUnknownType={Skipped} failedInit={Failed}",
+            generationId, registered, skippedUnknownType, failedInit);
+        return registered;
+    }
+}

src/ZB.MOM.WW.OtOpcUa.Server/OpcUaServerService.cs

@@ -18,6 +18,7 @@ public sealed class OpcUaServerService(
     DriverHost driverHost,
     OpcUaApplicationHost applicationHost,
     DriverEquipmentContentRegistry equipmentContentRegistry,
+    DriverInstanceBootstrapper driverBootstrapper,
     Phase7Composer phase7Composer,
     IServiceScopeFactory scopeFactory,
     ILogger<OpcUaServerService> logger) : BackgroundService
@@ -37,6 +38,13 @@ public sealed class OpcUaServerService(
         // address space until the first publish, then the registry fills on next restart.
         if (result.GenerationId is { } gen)
         {
+            // Task #248 — register IDriver instances from the published DriverInstance
+            // rows BEFORE the equipment-content load + Phase 7 compose, so the rest of
+            // the pipeline sees a populated DriverHost. Without this step Phase 7's
+            // CachedTagUpstreamSource has no upstream feed + virtual-tag scripts read
+            // BadNodeIdUnknown for every tag path (gap surfaced by task #240 smoke).
+            await driverBootstrapper.RegisterDriversFromGenerationAsync(gen, stoppingToken);
+
             await PopulateEquipmentContentAsync(gen, stoppingToken);
 
             // Phase 7 follow-up #246 — load Script + VirtualTag + ScriptedAlarm rows,

src/ZB.MOM.WW.OtOpcUa.Server/Program.cs

@@ -9,6 +9,7 @@ using ZB.MOM.WW.OtOpcUa.Configuration;
 using ZB.MOM.WW.OtOpcUa.Configuration.LocalCache;
 using ZB.MOM.WW.OtOpcUa.Core.Hosting;
 using ZB.MOM.WW.OtOpcUa.Core.AlarmHistorian;
+using ZB.MOM.WW.OtOpcUa.Driver.Galaxy.Proxy;
 using ZB.MOM.WW.OtOpcUa.Server;
 using ZB.MOM.WW.OtOpcUa.Server.OpcUa;
 using ZB.MOM.WW.OtOpcUa.Server.Phase7;
@@ -89,6 +90,18 @@ builder.Services.AddSingleton<ILocalConfigCache>(_ => new LiteDbConfigCache(opti
 builder.Services.AddSingleton<DriverHost>();
 builder.Services.AddSingleton<NodeBootstrap>();
 
+// Task #248 — driver-instance bootstrap pipeline. DriverFactoryRegistry is the
+// type-name → factory map; each driver project's static Register call pre-loads
+// its factory so the bootstrapper can materialise DriverInstance rows from the
+// central DB into live IDriver instances.
+builder.Services.AddSingleton<DriverFactoryRegistry>(_ =>
+{
+    var registry = new DriverFactoryRegistry();
+    GalaxyProxyDriverFactoryExtensions.Register(registry);
+    return registry;
+});
+builder.Services.AddSingleton<DriverInstanceBootstrapper>();
+
 // ADR-001 Option A wiring — the registry is the handoff between OpcUaServerService's
 // bootstrap-time population pass + OpcUaApplicationHost's StartAsync walker invocation.
 // DriverEquipmentContentRegistry.Get is the equipmentContentLookup delegate that PR #155

src/ZB.MOM.WW.OtOpcUa.Server/ZB.MOM.WW.OtOpcUa.Server.csproj

@@ -34,6 +34,7 @@
         <ProjectReference Include="..\ZB.MOM.WW.OtOpcUa.Core.VirtualTags\ZB.MOM.WW.OtOpcUa.Core.VirtualTags.csproj"/>
         <ProjectReference Include="..\ZB.MOM.WW.OtOpcUa.Core.ScriptedAlarms\ZB.MOM.WW.OtOpcUa.Core.ScriptedAlarms.csproj"/>
         <ProjectReference Include="..\ZB.MOM.WW.OtOpcUa.Core.AlarmHistorian\ZB.MOM.WW.OtOpcUa.Core.AlarmHistorian.csproj"/>
+        <ProjectReference Include="..\ZB.MOM.WW.OtOpcUa.Driver.Galaxy.Proxy\ZB.MOM.WW.OtOpcUa.Driver.Galaxy.Proxy.csproj"/>
         <ProjectReference Include="..\ZB.MOM.WW.OtOpcUa.Analyzers\ZB.MOM.WW.OtOpcUa.Analyzers.csproj"
                           OutputItemType="Analyzer" ReferenceOutputAssembly="false"/>
     </ItemGroup>

tests/ZB.MOM.WW.OtOpcUa.Server.Tests/DriverFactoryRegistryTests.cs

@@ -0,0 +1,73 @@
+using Shouldly;
+using Xunit;
+using ZB.MOM.WW.OtOpcUa.Core.Abstractions;
+using ZB.MOM.WW.OtOpcUa.Core.Hosting;
+
+namespace ZB.MOM.WW.OtOpcUa.Server.Tests;
+
+/// <summary>
+///     Task #248 — covers the <see cref="DriverFactoryRegistry"/> contract that
+///     <see cref="DriverInstanceBootstrapper"/> consumes.
+/// </summary>
+[Trait("Category", "Unit")]
+public sealed class DriverFactoryRegistryTests
+{
+    private static IDriver FakeDriver(string id, string config) => new FakeIDriver(id);
+
+    [Fact]
+    public void Register_then_TryGet_returns_factory()
+    {
+        var r = new DriverFactoryRegistry();
+        r.Register("MyDriver", FakeDriver);
+
+        r.TryGet("MyDriver").ShouldNotBeNull();
+        r.TryGet("Nope").ShouldBeNull();
+    }
+
+    [Fact]
+    public void Register_is_case_insensitive()
+    {
+        var r = new DriverFactoryRegistry();
+        r.Register("Galaxy", FakeDriver);
+        r.TryGet("galaxy").ShouldNotBeNull();
+        r.TryGet("GALAXY").ShouldNotBeNull();
+    }
+
+    [Fact]
+    public void Register_duplicate_type_throws()
+    {
+        var r = new DriverFactoryRegistry();
+        r.Register("Galaxy", FakeDriver);
+        Should.Throw<InvalidOperationException>(() => r.Register("Galaxy", FakeDriver));
+    }
+
+    [Fact]
+    public void Register_null_args_rejected()
+    {
+        var r = new DriverFactoryRegistry();
+        Should.Throw<ArgumentException>(() => r.Register("", FakeDriver));
+        Should.Throw<ArgumentNullException>(() => r.Register("X", null!));
+    }
+
+    [Fact]
+    public void RegisteredTypes_returns_snapshot()
+    {
+        var r = new DriverFactoryRegistry();
+        r.Register("A", FakeDriver);
+        r.Register("B", FakeDriver);
+        r.RegisteredTypes.ShouldContain("A");
+        r.RegisteredTypes.ShouldContain("B");
+    }
+
+    private sealed class FakeIDriver(string id) : IDriver
+    {
+        public string DriverInstanceId => id;
+        public string DriverType => "Fake";
+        public Task InitializeAsync(string _, CancellationToken __) => Task.CompletedTask;
+        public Task ReinitializeAsync(string _, CancellationToken __) => Task.CompletedTask;
+        public Task ShutdownAsync(CancellationToken _) => Task.CompletedTask;
+        public Task FlushOptionalCachesAsync(CancellationToken _) => Task.CompletedTask;
+        public DriverHealth GetHealth() => new(DriverState.Healthy, null, null);
+        public long GetMemoryFootprint() => 0;
+    }
+}