Phase 6.1 Stream B.4 — wire ScheduledRecycleHostedService into bootstrap

Task #125 / #137. The hosted service + scheduler classes already shipped;
this commit connects them to the published-generation driver list so a
Tier C driver with `RecycleIntervalSeconds` in its `ResilienceConfig`
actually gets an armed scheduler at bootstrap.

Wiring:

- `DriverFactoryRegistry.Register` gains an optional `DriverTier`
  parameter (default Tier.A). Existing call sites unchanged —
  `GalaxyProxyDriverFactoryExtensions.Register` explicitly passes
  Tier.C so the bootstrapper can identify out-of-process drivers
  without a per-driver-type allow-list.
- `DriverResilienceOptions` + parser grow `RecycleIntervalSeconds`.
  Tier A/B values are rejected with a diagnostic (decision #74 —
  recycling an in-process driver would kill every OPC UA session).
  Non-positive values are rejected the same way.
- `DriverInstanceBootstrapper` auto-arms a `ScheduledRecycleScheduler`
  after a successful driver register when: (1) the registered tier is
  C, (2) the row's ResilienceConfig carries a positive recycle interval,
  (3) DI has an `IDriverSupervisor` keyed by that `DriverInstanceId`.
  Missing supervisor → warn + skip (no crash). That keeps the wiring
  harmless by default: no driver ships a supervisor today, so the
  hosted service runs with zero schedulers out of the box.
- `Program.cs` registers `ScheduledRecycleHostedService` as singleton
  (shared with `DriverInstanceBootstrapper`) + hosted service (drives
  the tick loop). Constructor changes on the bootstrapper ripple into
  DI resolution automatically.

Tests: 4 new parser tests covering RecycleIntervalSeconds on Tier C
happy path, null default, Tier A/B rejection, non-positive rejection.
Existing 283 Server.Tests + 200 Core.Tests all still green.

No behavioural change for existing deployments: Galaxy driver + any
future Tier C driver gain the opt-in automatically; Tier A/B drivers
(FOCAS, Modbus, S7, AB CIP, AB Legacy, TwinCAT) are structurally
excluded.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
Joseph Doherty
2026-04-24 18:58:13 -04:00
parent a52086efc5
commit bd6568bcbd
7 changed files with 152 additions and 2 deletions

View File

@@ -19,6 +19,8 @@ public sealed class DriverFactoryRegistry
{
private readonly Dictionary<string, Func<string, string, IDriver>> _factories
= new(StringComparer.OrdinalIgnoreCase);
private readonly Dictionary<string, DriverTier> _tiers
= new(StringComparer.OrdinalIgnoreCase);
private readonly object _lock = new();
/// <summary>
@@ -33,7 +35,15 @@ public sealed class DriverFactoryRegistry
/// 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)
/// <param name="tier">
/// Stability tier per <c>docs/v2/driver-stability.md</c>. Defaults to
/// <see cref="DriverTier.A"/> (in-process, lowest recycle footprint); Tier C
/// drivers SHOULD pass this explicitly so the scheduled-recycle wiring in
/// <c>DriverInstanceBootstrapper</c> can skip in-process drivers by tier check
/// rather than by driver-type allow-list.
/// </param>
public void Register(string driverType, Func<string, string, IDriver> factory,
DriverTier tier = DriverTier.A)
{
ArgumentException.ThrowIfNullOrWhiteSpace(driverType);
ArgumentNullException.ThrowIfNull(factory);
@@ -43,6 +53,7 @@ public sealed class DriverFactoryRegistry
throw new InvalidOperationException(
$"DriverType '{driverType}' factory already registered for this process");
_factories[driverType] = factory;
_tiers[driverType] = tier;
}
}
@@ -57,6 +68,17 @@ public sealed class DriverFactoryRegistry
lock (_lock) return _factories.GetValueOrDefault(driverType);
}
/// <summary>
/// Look up the tier recorded alongside the factory. Returns <see cref="DriverTier.A"/>
/// for unknown driver types — a missing registration is already a skipped-bootstrap
/// case upstream; we don't double-surface that failure here.
/// </summary>
public DriverTier GetTier(string driverType)
{
ArgumentException.ThrowIfNullOrWhiteSpace(driverType);
lock (_lock) return _tiers.GetValueOrDefault(driverType, DriverTier.A);
}
public IReadOnlyCollection<string> RegisteredTypes
{
get { lock (_lock) return [.. _factories.Keys]; }

View File

@@ -28,6 +28,16 @@ public sealed record DriverResilienceOptions
/// </summary>
public int BulkheadMaxQueue { get; init; } = 64;
/// <summary>
/// Periodic scheduled recycle interval for Tier C out-of-process hosts, in seconds.
/// Null (the default) = no scheduled recycle; the driver's Host process keeps running
/// indefinitely unless a memory breach or operator action triggers a recycle. Only
/// respected for <see cref="DriverTier.C"/>; Tier A/B recycle would tear down every
/// OPC UA session, so the loader ignores non-null values for those tiers + logs a
/// warning (per decisions #74 / #145).
/// </summary>
public int? RecycleIntervalSeconds { get; init; }
/// <summary>
/// Look up the effective policy for a capability, falling back to tier defaults when no
/// override is configured. Never returns null.

View File

@@ -91,12 +91,27 @@ public static class DriverResilienceOptionsParser
}
}
// Scheduled recycle is Tier C only — reject a configured interval on Tier A/B as a
// misconfiguration surface rather than silently honouring it (recycling an in-process
// driver would kill every OPC UA session + every co-hosted driver, per decision #74).
int? recycleIntervalSeconds = null;
if (shape.RecycleIntervalSeconds is int secs)
{
if (secs <= 0)
parseDiagnostic ??= $"RecycleIntervalSeconds must be positive; got {secs} — ignoring.";
else if (tier != DriverTier.C)
parseDiagnostic ??= $"RecycleIntervalSeconds is Tier C only; Tier {tier} driver will not scheduled-recycle.";
else
recycleIntervalSeconds = secs;
}
return new DriverResilienceOptions
{
Tier = tier,
CapabilityPolicies = merged,
BulkheadMaxConcurrent = shape.BulkheadMaxConcurrent ?? baseOptions.BulkheadMaxConcurrent,
BulkheadMaxQueue = shape.BulkheadMaxQueue ?? baseOptions.BulkheadMaxQueue,
RecycleIntervalSeconds = recycleIntervalSeconds,
};
}
@@ -104,6 +119,7 @@ public static class DriverResilienceOptionsParser
{
public int? BulkheadMaxConcurrent { get; set; }
public int? BulkheadMaxQueue { get; set; }
public int? RecycleIntervalSeconds { get; set; }
public Dictionary<string, CapabilityPolicyShape>? CapabilityPolicies { get; set; }
}

View File

@@ -1,4 +1,5 @@
using System.Text.Json;
using ZB.MOM.WW.OtOpcUa.Core.Abstractions;
using ZB.MOM.WW.OtOpcUa.Core.Hosting;
namespace ZB.MOM.WW.OtOpcUa.Driver.Galaxy.Proxy;
@@ -21,7 +22,8 @@ public static class GalaxyProxyDriverFactoryExtensions
public static void Register(DriverFactoryRegistry registry)
{
ArgumentNullException.ThrowIfNull(registry);
registry.Register(DriverTypeName, CreateInstance);
// Galaxy is Tier C — out-of-process MXAccess Host, scheduled recycle is allowed.
registry.Register(DriverTypeName, CreateInstance, DriverTier.C);
}
internal static GalaxyProxyDriver CreateInstance(string driverInstanceId, string driverConfigJson)

View File

@@ -2,7 +2,11 @@ using Microsoft.EntityFrameworkCore;
using Microsoft.Extensions.DependencyInjection;
using Microsoft.Extensions.Logging;
using ZB.MOM.WW.OtOpcUa.Configuration;
using ZB.MOM.WW.OtOpcUa.Core.Abstractions;
using ZB.MOM.WW.OtOpcUa.Core.Hosting;
using ZB.MOM.WW.OtOpcUa.Core.Resilience;
using ZB.MOM.WW.OtOpcUa.Core.Stability;
using ZB.MOM.WW.OtOpcUa.Server.Hosting;
namespace ZB.MOM.WW.OtOpcUa.Server;
@@ -33,9 +37,20 @@ namespace ZB.MOM.WW.OtOpcUa.Server;
public sealed class DriverInstanceBootstrapper(
DriverFactoryRegistry factories,
DriverHost driverHost,
ScheduledRecycleHostedService recycleHost,
ILoggerFactory loggerFactory,
IServiceScopeFactory scopeFactory,
ILogger<DriverInstanceBootstrapper> logger)
{
// IDriverSupervisor instances, looked up by DriverInstanceId. The bootstrapper
// consults DI at run time because no driver ships a supervisor today — the
// dictionary is built from optional DI registrations; Tier C drivers that
// register one via `services.AddKeyedSingleton<IDriverSupervisor>(instanceId, ...)`
// become eligible for scheduled recycle. Others silently skip.
private readonly IReadOnlyDictionary<string, IDriverSupervisor> _supervisors =
scopeFactory.CreateScope().ServiceProvider
.GetServices<IDriverSupervisor>()
.ToDictionary(s => s.DriverInstanceId, StringComparer.OrdinalIgnoreCase);
public async Task<int> RegisterDriversFromGenerationAsync(long generationId, CancellationToken ct)
{
using var scope = scopeFactory.CreateScope();
@@ -68,6 +83,13 @@ public sealed class DriverInstanceBootstrapper(
registered++;
logger.LogInformation(
"DriverInstance {Id} ({Type}) registered + initialized", row.DriverInstanceId, row.DriverType);
// Scheduled-recycle opt-in — only meaningful for Tier C out-of-process hosts,
// and only when the row's ResilienceConfig carries a positive
// RecycleIntervalSeconds AND the deployment wired an IDriverSupervisor for
// this DriverInstanceId. Silently skipping when any of those is absent is the
// intended zero-config-default behaviour.
TryRegisterScheduledRecycle(row.DriverInstanceId, row.DriverType, row.ResilienceConfig);
}
catch (Exception ex)
{
@@ -85,4 +107,32 @@ public sealed class DriverInstanceBootstrapper(
generationId, registered, skippedUnknownType, failedInit);
return registered;
}
private void TryRegisterScheduledRecycle(string driverInstanceId, string driverType, string? resilienceJson)
{
var tier = factories.GetTier(driverType);
if (tier != DriverTier.C) return;
var options = DriverResilienceOptionsParser.ParseOrDefaults(tier, resilienceJson, out _);
if (options.RecycleIntervalSeconds is not int secs) return;
if (!_supervisors.TryGetValue(driverInstanceId, out var supervisor))
{
logger.LogWarning(
"DriverInstance {Id} ({Type}) has RecycleIntervalSeconds={Secs} in ResilienceConfig but no IDriverSupervisor registered; scheduled recycle will not fire",
driverInstanceId, driverType, secs);
return;
}
var scheduler = new ScheduledRecycleScheduler(
tier,
TimeSpan.FromSeconds(secs),
DateTime.UtcNow,
supervisor,
loggerFactory.CreateLogger<ScheduledRecycleScheduler>());
recycleHost.AddScheduler(scheduler);
logger.LogInformation(
"Scheduled recycle armed for Tier C driver {Id} ({Type}) — interval {Interval}, first fire at {Next:o}",
driverInstanceId, driverType, TimeSpan.FromSeconds(secs), scheduler.NextRecycleUtc);
}
}

View File

@@ -117,6 +117,14 @@ builder.Services.AddSingleton<DriverFactoryRegistry>(_ =>
});
builder.Services.AddSingleton<DriverInstanceBootstrapper>();
// Phase 6.1 Stream B.4 (task #137) — ScheduledRecycleHostedService. Empty scheduler
// list by default; DriverInstanceBootstrapper calls AddScheduler for any Tier C driver
// whose ResilienceConfig carries a RecycleIntervalSeconds AND has an IDriverSupervisor
// registered in DI. Registered as singleton so DriverInstanceBootstrapper can inject
// the same instance that the BackgroundService loop drives.
builder.Services.AddSingleton<ScheduledRecycleHostedService>();
builder.Services.AddHostedService(sp => sp.GetRequiredService<ScheduledRecycleHostedService>());
// 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

View File

@@ -163,4 +163,46 @@ public sealed class DriverResilienceOptionsParserTests
foreach (var cap in Enum.GetValues<DriverCapability>())
options.Resolve(cap).ShouldBe(DriverResilienceOptions.GetTierDefaults(tier)[cap]);
}
[Fact]
public void RecycleIntervalSeconds_TierC_PositiveValue_ParsesAndSurfaces()
{
var options = DriverResilienceOptionsParser.ParseOrDefaults(
DriverTier.C, "{\"recycleIntervalSeconds\":3600}", out var diag);
diag.ShouldBeNull();
options.RecycleIntervalSeconds.ShouldBe(3600);
}
[Fact]
public void RecycleIntervalSeconds_Null_DefaultsToNull()
{
var options = DriverResilienceOptionsParser.ParseOrDefaults(DriverTier.C, "{}", out _);
options.RecycleIntervalSeconds.ShouldBeNull();
}
[Theory]
[InlineData(DriverTier.A)]
[InlineData(DriverTier.B)]
public void RecycleIntervalSeconds_OnTierAorB_Rejected_With_Diagnostic(DriverTier tier)
{
// Decision #74 — in-process drivers must not scheduled-recycle because it would
// tear down every OPC UA session. The parser surfaces a diagnostic rather than
// silently honouring the value.
var options = DriverResilienceOptionsParser.ParseOrDefaults(
tier, "{\"recycleIntervalSeconds\":3600}", out var diag);
options.RecycleIntervalSeconds.ShouldBeNull();
diag.ShouldContain("Tier C only");
}
[Fact]
public void RecycleIntervalSeconds_NonPositive_Rejected_With_Diagnostic()
{
var options = DriverResilienceOptionsParser.ParseOrDefaults(
DriverTier.C, "{\"recycleIntervalSeconds\":0}", out var diag);
options.RecycleIntervalSeconds.ShouldBeNull();
diag.ShouldContain("must be positive");
}
}