feat(siteeventlog): emit deployment + instance_lifecycle events (M1.6)

DeploymentManagerActor now fire-and-forgets a 'deployment' site operational event on deploy/enable/disable/delete outcomes (Info on success, Error on failure), source 'DeploymentManagerActor'. The disable/delete events are emitted from the existing PipeTo continuations (safe: reads only the immutable _serviceProvider and fire-and-forgets). InstanceActor now emits an 'instance_lifecycle' Info event in PreStart (started) and a new PostStop (stopped) — covering start/stop/enable/disable/redeploy/ failover transitions from the instance's own vantage point. Both actors already hold _serviceProvider; no ctor change. Resolution is optional and LogEventAsync is fire-and-forget so a logging failure never affects the deployment pipeline or instance lifecycle.
2026-06-15 12:26:54 -04:00
parent a00e43c4f9
commit 09b9e8f259
4 changed files with 229 additions and 2 deletions
@@ -1,4 +1,5 @@
 using Akka.Actor;
+using Microsoft.Extensions.DependencyInjection;
 using Microsoft.Extensions.Logging;
 using ZB.MOM.WW.ScadaBridge.Commons.Messages.Artifacts;
 using ZB.MOM.WW.ScadaBridge.Commons.Messages.DebugView;
@@ -10,6 +11,7 @@ using ZB.MOM.WW.ScadaBridge.Commons.Messages.Management;
 using ZB.MOM.WW.ScadaBridge.Commons.Messages.ScriptExecution;
 using ZB.MOM.WW.ScadaBridge.Commons.Types.Enums;
 using ZB.MOM.WW.ScadaBridge.HealthMonitoring;
+using ZB.MOM.WW.ScadaBridge.SiteEventLogging;
 using ZB.MOM.WW.ScadaBridge.SiteRuntime.Messages;
 using ZB.MOM.WW.ScadaBridge.SiteRuntime.Persistence;
 using ZB.MOM.WW.ScadaBridge.SiteRuntime.Scripts;
@@ -456,6 +458,10 @@ public class DeploymentManagerActor : ReceiveActor, IWithTimers
    {
        if (result.Success)
        {
+            // M1.6: operational `deployment` event — deploy succeeded.
+            LogDeploymentEvent("Info", result.InstanceName,
+                $"Instance {result.InstanceName} deployed (deploymentId={result.DeploymentId})");
+
            result.OriginalSender.Tell(new DeploymentStatusResponse(
                result.DeploymentId,
                result.InstanceName,
@@ -469,6 +475,11 @@ public class DeploymentManagerActor : ReceiveActor, IWithTimers
            "Failed to persist deployment {DeploymentId} for {Instance}: {Error}",
            result.DeploymentId, result.InstanceName, result.Error);

+        // M1.6: operational `deployment` event — deploy failed.
+        LogDeploymentEvent("Error", result.InstanceName,
+            $"Instance {result.InstanceName} deploy failed (deploymentId={result.DeploymentId})",
+            result.Error);
+
        // Persistence failed — undo the optimistic actor creation and counter bump so
        // the site does not advertise an instance it cannot durably recover.
        if (_instanceActors.Remove(result.InstanceName, out var orphan))
@@ -504,7 +515,17 @@ public class DeploymentManagerActor : ReceiveActor, IWithTimers
        _storage.SetInstanceEnabledAsync(instanceName, false).ContinueWith(t =>
        {
            if (t.IsCompletedSuccessfully)
+            {
                _replicationActor?.Tell(new ReplicateConfigSetEnabled(instanceName, false));
+                // M1.6: operational `deployment` event — disable succeeded.
+                LogDeploymentEvent("Info", instanceName, $"Instance {instanceName} disabled");
+            }
+            else
+            {
+                LogDeploymentEvent("Error", instanceName,
+                    $"Instance {instanceName} disable failed",
+                    t.Exception?.GetBaseException().Message);
+            }

            return new InstanceLifecycleResponse(
                command.CommandId,
@@ -551,6 +572,9 @@ public class DeploymentManagerActor : ReceiveActor, IWithTimers
        if (result.Error != null || result.Config == null)
        {
            var error = result.Error ?? $"No deployed config found for {instanceName}";
+            // M1.6: operational `deployment` event — enable failed.
+            LogDeploymentEvent("Error", instanceName,
+                $"Instance {instanceName} enable failed", error);
            result.OriginalSender.Tell(new InstanceLifecycleResponse(
                result.Command.CommandId, instanceName, false, error, DateTimeOffset.UtcNow));
            return;
@@ -562,6 +586,9 @@ public class DeploymentManagerActor : ReceiveActor, IWithTimers
        }
        UpdateInstanceCounts();

+        // M1.6: operational `deployment` event — enable succeeded.
+        LogDeploymentEvent("Info", instanceName, $"Instance {instanceName} enabled");
+
        result.OriginalSender.Tell(new InstanceLifecycleResponse(
            result.Command.CommandId, instanceName, true, null, DateTimeOffset.UtcNow));

@@ -588,7 +615,17 @@ public class DeploymentManagerActor : ReceiveActor, IWithTimers
        _storage.RemoveDeployedConfigAsync(instanceName).ContinueWith(t =>
        {
            if (t.IsCompletedSuccessfully)
+            {
                _replicationActor?.Tell(new ReplicateConfigRemove(instanceName));
+                // M1.6: operational `deployment` event — delete succeeded.
+                LogDeploymentEvent("Info", instanceName, $"Instance {instanceName} deleted");
+            }
+            else
+            {
+                LogDeploymentEvent("Error", instanceName,
+                    $"Instance {instanceName} delete failed",
+                    t.Exception?.GetBaseException().Message);
+            }

            return new InstanceLifecycleResponse(
                command.CommandId,
@@ -601,6 +638,21 @@ public class DeploymentManagerActor : ReceiveActor, IWithTimers
        _logger.LogInformation("Instance {Instance} deleted", instanceName);
    }

+    /// <summary>
+    /// M1.6: fire-and-forget a <c>deployment</c> operational event to the optional
+    /// <see cref="ISiteEventLogger"/> on a deploy/enable/disable/delete outcome.
+    /// Resolved optionally and never awaited so a logging failure cannot affect the
+    /// deployment pipeline (matching the established ScriptActor/ScriptExecutionActor
+    /// pattern). Only reads the immutable <c>_serviceProvider</c> field, so it is
+    /// safe to call from the PipeTo continuations that report disable/delete
+    /// outcomes off the actor thread.
+    /// </summary>
+    private void LogDeploymentEvent(string severity, string instanceName, string message, string? details = null)
+    {
+        _ = _serviceProvider?.GetService<ISiteEventLogger>()?.LogEventAsync(
+            "deployment", severity, instanceName, "DeploymentManagerActor", message, details);
+    }
+
    /// <summary>
    /// DeploymentManager-006: answers a central query for the instance's
    /// currently-applied deployment identity. The site's deployed-config store
@@ -1,5 +1,6 @@
 using Akka.Actor;
 using Microsoft.CodeAnalysis.Scripting;
+using Microsoft.Extensions.DependencyInjection;
 using Microsoft.Extensions.Logging;
 using ZB.MOM.WW.ScadaBridge.Commons.Messages.DataConnection;
 using ZB.MOM.WW.ScadaBridge.Commons.Messages.DebugView;
@@ -9,6 +10,7 @@ using ZB.MOM.WW.ScadaBridge.Commons.Messages.Streaming;
 using ZB.MOM.WW.ScadaBridge.Commons.Types.Enums;
 using ZB.MOM.WW.ScadaBridge.Commons.Types.Flattening;
 using ZB.MOM.WW.ScadaBridge.HealthMonitoring;
+using ZB.MOM.WW.ScadaBridge.SiteEventLogging;
 using ZB.MOM.WW.ScadaBridge.SiteRuntime.Persistence;
 using ZB.MOM.WW.ScadaBridge.SiteRuntime.Scripts;
 using ZB.MOM.WW.ScadaBridge.SiteRuntime.Streaming;
@@ -164,6 +166,11 @@ public class InstanceActor : ReceiveActor
        base.PreStart();
        _logger.LogInformation("InstanceActor started for {Instance}", _instanceUniqueName);

+        // M1.6: operational `instance_lifecycle` event — instance started.
+        // An instance starts on deploy, on enable (DeploymentManager re-creates
+        // the actor), and on failover/restart; this single point covers them all.
+        LogLifecycleEvent($"Instance {_instanceUniqueName} started");
+
        // Asynchronously load static overrides from SQLite and pipe to self
        var self = Self;
        _storage.GetStaticOverridesAsync(_instanceUniqueName).ContinueWith(t =>
@@ -180,6 +187,29 @@ public class InstanceActor : ReceiveActor
        SubscribeToDcl();
    }

+    /// <inheritdoc />
+    protected override void PostStop()
+    {
+        // M1.6: operational `instance_lifecycle` event — instance stopped. An
+        // instance stops on disable, delete, redeployment, and graceful shutdown;
+        // this single point covers them all.
+        LogLifecycleEvent($"Instance {_instanceUniqueName} stopped");
+        base.PostStop();
+    }
+
+    /// <summary>
+    /// M1.6: fire-and-forget an <c>instance_lifecycle</c> operational event to the
+    /// optional <see cref="ISiteEventLogger"/>. Resolved optionally and never
+    /// awaited so a logging failure cannot affect the instance lifecycle
+    /// (matching the established ScriptActor/ScriptExecutionActor pattern).
+    /// </summary>
+    private void LogLifecycleEvent(string message)
+    {
+        _ = _serviceProvider?.GetService<ISiteEventLogger>()?.LogEventAsync(
+            "instance_lifecycle", "Info", _instanceUniqueName,
+            $"InstanceActor:{_instanceUniqueName}", message);
+    }
+
    /// <inheritdoc />
    protected override SupervisorStrategy SupervisorStrategy()
    {
@@ -10,6 +10,7 @@ using ZB.MOM.WW.ScadaBridge.Commons.Types.Flattening;
 using ZB.MOM.WW.ScadaBridge.SiteRuntime.Actors;
 using ZB.MOM.WW.ScadaBridge.SiteRuntime.Persistence;
 using ZB.MOM.WW.ScadaBridge.SiteRuntime.Scripts;
+using ZB.MOM.WW.ScadaBridge.SiteRuntime.Tests.TestSupport;
 using System.Text.Json;

 namespace ZB.MOM.WW.ScadaBridge.SiteRuntime.Tests.Actors;
@@ -44,7 +45,8 @@ public class DeploymentManagerActorTests : TestKit, IDisposable
        try { File.Delete(_dbFile); } catch { /* cleanup */ }
    }

-    private IActorRef CreateDeploymentManager(SiteRuntimeOptions? options = null)
+    private IActorRef CreateDeploymentManager(
+        SiteRuntimeOptions? options = null, IServiceProvider? serviceProvider = null)
    {
        options ??= new SiteRuntimeOptions();
        return ActorOf(Props.Create(() => new DeploymentManagerActor(
@@ -53,7 +55,12 @@ public class DeploymentManagerActorTests : TestKit, IDisposable
            _sharedScriptLibrary,
            null, // no stream manager in tests
            options,
-            NullLogger<DeploymentManagerActor>.Instance)));
+            NullLogger<DeploymentManagerActor>.Instance,
+            null,
+            null,
+            null,
+            serviceProvider,
+            null)));
    }

    private static string MakeConfigJson(string instanceName)
@@ -171,6 +178,67 @@ public class DeploymentManagerActorTests : TestKit, IDisposable
        Assert.Equal("NewPump", response.InstanceUniqueName);
    }

+    // ── M1.6: site event log `deployment` category ─────────────────────────
+
+    [Fact]
+    public async Task DeploymentManager_DeploySuccess_EmitsDeploymentSiteEvent()
+    {
+        var siteLog = new FakeSiteEventLogger();
+        var actor = CreateDeploymentManager(serviceProvider: new SingleServiceProvider(siteLog));
+
+        await Task.Delay(500); // wait for empty startup
+
+        actor.Tell(new DeployInstanceCommand(
+            "dep-evt-1", "AuditedPump", "sha256:xyz",
+            MakeConfigJson("AuditedPump"), "admin", DateTimeOffset.UtcNow));
+        var response = ExpectMsg<DeploymentStatusResponse>(TimeSpan.FromSeconds(5));
+        Assert.Equal(DeploymentStatus.Success, response.Status);
+
+        AwaitAssert(() =>
+        {
+            var rows = siteLog.OfType("deployment");
+            Assert.Contains(rows, r =>
+                r.Severity == "Info" &&
+                r.InstanceId == "AuditedPump" &&
+                r.Source == "DeploymentManagerActor" &&
+                r.Message.Contains("deploy", StringComparison.OrdinalIgnoreCase));
+        }, TimeSpan.FromSeconds(2));
+    }
+
+    [Fact]
+    public async Task DeploymentManager_DisableEnableDelete_EmitDeploymentSiteEvents()
+    {
+        var siteLog = new FakeSiteEventLogger();
+        var actor = CreateDeploymentManager(serviceProvider: new SingleServiceProvider(siteLog));
+
+        await Task.Delay(500);
+
+        actor.Tell(new DeployInstanceCommand(
+            "dep-evt-2", "EvtPump", "sha256:abc",
+            MakeConfigJson("EvtPump"), "admin", DateTimeOffset.UtcNow));
+        ExpectMsg<DeploymentStatusResponse>(TimeSpan.FromSeconds(5));
+        await Task.Delay(1000);
+
+        actor.Tell(new DisableInstanceCommand("cmd-de1", "EvtPump", DateTimeOffset.UtcNow));
+        Assert.True(ExpectMsg<InstanceLifecycleResponse>(TimeSpan.FromSeconds(5)).Success);
+        await Task.Delay(300);
+
+        actor.Tell(new EnableInstanceCommand("cmd-en1", "EvtPump", DateTimeOffset.UtcNow));
+        Assert.True(ExpectMsg<InstanceLifecycleResponse>(TimeSpan.FromSeconds(5)).Success);
+        await Task.Delay(300);
+
+        actor.Tell(new DeleteInstanceCommand("cmd-del-evt", "EvtPump", DateTimeOffset.UtcNow));
+        Assert.True(ExpectMsg<InstanceLifecycleResponse>(TimeSpan.FromSeconds(5)).Success);
+
+        AwaitAssert(() =>
+        {
+            var rows = siteLog.OfType("deployment");
+            Assert.Contains(rows, r => r.Message.Contains("disabled", StringComparison.OrdinalIgnoreCase));
+            Assert.Contains(rows, r => r.Message.Contains("enabled", StringComparison.OrdinalIgnoreCase));
+            Assert.Contains(rows, r => r.Message.Contains("deleted", StringComparison.OrdinalIgnoreCase));
+        }, TimeSpan.FromSeconds(2));
+    }
+
    [Fact]
    public async Task DeploymentManager_Lifecycle_DisableEnableDelete()
    {
@@ -10,6 +10,7 @@ using ZB.MOM.WW.ScadaBridge.Commons.Types.Flattening;
 using ZB.MOM.WW.ScadaBridge.SiteRuntime.Actors;
 using ZB.MOM.WW.ScadaBridge.SiteRuntime.Persistence;
 using ZB.MOM.WW.ScadaBridge.SiteRuntime.Scripts;
+using ZB.MOM.WW.ScadaBridge.SiteRuntime.Tests.TestSupport;
 using System.Text.Json;

 namespace ZB.MOM.WW.ScadaBridge.SiteRuntime.Tests.Actors;
@@ -58,6 +59,82 @@ public class InstanceActorTests : TestKit, IDisposable
        try { File.Delete(_dbFile); } catch { /* cleanup */ }
    }

+    // ── M1.6: site event log `instance_lifecycle` category ──────────────────
+
+    [Fact]
+    public void InstanceActor_Start_EmitsInstanceLifecycleSiteEvent()
+    {
+        var siteLog = new FakeSiteEventLogger();
+        var config = new FlattenedConfiguration
+        {
+            InstanceUniqueName = "LifecyclePump",
+            Attributes = [new ResolvedAttribute { CanonicalName = "T", Value = "1", DataType = "Int32" }]
+        };
+
+        ActorOf(Props.Create(() => new InstanceActor(
+            "LifecyclePump",
+            JsonSerializer.Serialize(config),
+            _storage,
+            _compilationService,
+            _sharedScriptLibrary,
+            null,
+            _options,
+            NullLogger<InstanceActor>.Instance,
+            null,
+            null,
+            new SingleServiceProvider(siteLog))));
+
+        AwaitAssert(() =>
+        {
+            var rows = siteLog.OfType("instance_lifecycle");
+            Assert.Contains(rows, r =>
+                r.Severity == "Info" &&
+                r.InstanceId == "LifecyclePump" &&
+                r.Source == "InstanceActor:LifecyclePump" &&
+                r.Message.Contains("started", StringComparison.OrdinalIgnoreCase));
+        }, TimeSpan.FromSeconds(2));
+    }
+
+    [Fact]
+    public void InstanceActor_Stop_EmitsInstanceLifecycleSiteEvent()
+    {
+        var siteLog = new FakeSiteEventLogger();
+        var config = new FlattenedConfiguration
+        {
+            InstanceUniqueName = "StoppedPump",
+            Attributes = [new ResolvedAttribute { CanonicalName = "T", Value = "1", DataType = "Int32" }]
+        };
+
+        var actor = ActorOf(Props.Create(() => new InstanceActor(
+            "StoppedPump",
+            JsonSerializer.Serialize(config),
+            _storage,
+            _compilationService,
+            _sharedScriptLibrary,
+            null,
+            _options,
+            NullLogger<InstanceActor>.Instance,
+            null,
+            null,
+            new SingleServiceProvider(siteLog))));
+
+        // Let PreStart land its started event, then stop the actor.
+        AwaitAssert(() => Assert.NotEmpty(siteLog.OfType("instance_lifecycle")),
+            TimeSpan.FromSeconds(2));
+        Watch(actor);
+        actor.Tell(PoisonPill.Instance);
+        ExpectTerminated(actor, TimeSpan.FromSeconds(5));
+
+        AwaitAssert(() =>
+        {
+            var rows = siteLog.OfType("instance_lifecycle");
+            Assert.Contains(rows, r =>
+                r.Severity == "Info" &&
+                r.InstanceId == "StoppedPump" &&
+                r.Message.Contains("stopped", StringComparison.OrdinalIgnoreCase));
+        }, TimeSpan.FromSeconds(2));
+    }
+
    [Fact]
    public void InstanceActor_LoadsAttributesFromConfig()
    {