feat(deploy): additive degraded flag on ApplyAck + publish-actor apply outcome reply (#601)

ApplyAck gains trailing Degraded/DegradedReason (defaulted, so an old node's
DPS JSON or gRPC ack decodes exactly as before). mesh_transport.proto adds
ApplyAckMessage.degraded = 6 / optional degraded_reason = 7. OpcUaPublishActor
replies RebuildCompleted(Degraded, Reason) on every handler path when the
RebuildAddressSpace carries ReportOutcome.

Claude-Session: https://claude.ai/code/session_01Ef2ozxS7QPanGT8uoATbTe
This commit is contained in:
Joseph Doherty
2026-10-08 17:19:57 -04:00
parent 9b74a65b25
commit 521e1eb89b
6 changed files with 286 additions and 11 deletions
@@ -0,0 +1,154 @@
using System.Text.Json;
using Akka.Actor;
using Microsoft.Extensions.Logging.Abstractions;
using Shouldly;
using Xunit;
using ZB.MOM.WW.OtOpcUa.Commons.OpcUa;
using ZB.MOM.WW.OtOpcUa.Commons.Types;
using ZB.MOM.WW.OtOpcUa.OpcUaServer;
using ZB.MOM.WW.OtOpcUa.Runtime.OpcUa;
using ZB.MOM.WW.OtOpcUa.Runtime.Tests.Harness;
namespace ZB.MOM.WW.OtOpcUa.Runtime.Tests.OpcUa;
/// <summary>
/// Gitea #601 — the deploy ack is sent only once the address-space apply outcome is known, so the
/// publish actor must REPORT that outcome when asked (<see cref="OpcUaPublishActor.RebuildAddressSpace.ReportOutcome"/>):
/// clean for a clean apply, degraded for swallowed per-node sink faults or a pipeline throw, and nothing
/// at all when the caller did not ask (boot/restore paths re-ack nothing).
/// </summary>
/// <remarks>
/// In the <c>opcua-publish-meters</c> collection because the degraded cases increment
/// <c>otopcua.opcua.apply.failed</c>, which the sibling classes assert stays at zero.
/// </remarks>
[Collection("opcua-publish-meters")]
public sealed class OpcUaPublishActorApplyOutcomeTests : RuntimeActorTestBase
{
[Fact]
public void A_clean_apply_reports_a_clean_outcome_to_the_requester()
{
var sink = new FaultingSink(throwOnEnsureFolder: false);
var applier = new AddressSpaceApplier(sink, NullLogger<AddressSpaceApplier>.Instance);
var actor = Sys.ActorOf(OpcUaPublishActor.PropsForTests(sink: sink, applier: applier));
var correlation = CorrelationId.NewId();
var deployment = DeploymentId.NewId();
actor.Tell(new OpcUaPublishActor.RebuildAddressSpace(
correlation, deployment, BuildEquipmentArtifact("eq-1"), ReportOutcome: true));
var done = ExpectMsg<OpcUaPublishActor.RebuildCompleted>(PresenceBudget);
done.Correlation.ShouldBe(correlation);
done.DeploymentId.ShouldBe(deployment);
done.Degraded.ShouldBeFalse();
done.DegradedReason.ShouldBeNull();
}
[Fact]
public void A_degraded_non_rebuild_apply_reports_degraded_with_the_failed_node_count()
{
// Every EnsureFolder throws — the applier swallows each and counts it, the classic degraded apply
// that leaves _lastApplied un-advanced and used to seal the deployment green.
var sink = new FaultingSink(throwOnEnsureFolder: true);
var applier = new AddressSpaceApplier(sink, NullLogger<AddressSpaceApplier>.Instance);
var actor = Sys.ActorOf(OpcUaPublishActor.PropsForTests(sink: sink, applier: applier));
actor.Tell(new OpcUaPublishActor.RebuildAddressSpace(
CorrelationId.NewId(), DeploymentId.NewId(), BuildEquipmentArtifact("eq-1"), ReportOutcome: true));
var done = ExpectMsg<OpcUaPublishActor.RebuildCompleted>(PresenceBudget);
done.Degraded.ShouldBeTrue();
done.DegradedReason.ShouldNotBeNull();
done.DegradedReason.ShouldContain("failedNodes=");
}
[Fact]
public void A_pipeline_that_throws_reports_degraded()
{
var sink = new FaultingSink(throwOnEnsureFolder: false);
var applier = new AddressSpaceApplier(sink, NullLogger<AddressSpaceApplier>.Instance);
var actor = Sys.ActorOf(OpcUaPublishActor.PropsForTests(sink: sink, applier: applier));
// Non-empty bytes that are not JSON: DeploymentArtifact.Open throws inside the rebuild pipeline.
actor.Tell(new OpcUaPublishActor.RebuildAddressSpace(
CorrelationId.NewId(), DeploymentId.NewId(), "not json"u8.ToArray(), ReportOutcome: true));
var done = ExpectMsg<OpcUaPublishActor.RebuildCompleted>(PresenceBudget);
done.Degraded.ShouldBeTrue();
done.DegradedReason.ShouldNotBeNull();
}
[Fact]
public void An_unchanged_composition_reports_clean()
{
var sink = new FaultingSink(throwOnEnsureFolder: false);
var applier = new AddressSpaceApplier(sink, NullLogger<AddressSpaceApplier>.Instance);
var actor = Sys.ActorOf(OpcUaPublishActor.PropsForTests(sink: sink, applier: applier));
var artifact = BuildEquipmentArtifact("eq-1");
actor.Tell(new OpcUaPublishActor.RebuildAddressSpace(CorrelationId.NewId(), DeploymentId.NewId(), artifact, ReportOutcome: true));
ExpectMsg<OpcUaPublishActor.RebuildCompleted>(PresenceBudget).Degraded.ShouldBeFalse();
// Same composition again ⇒ empty plan ⇒ nothing to apply, nothing degraded.
actor.Tell(new OpcUaPublishActor.RebuildAddressSpace(CorrelationId.NewId(), DeploymentId.NewId(), artifact, ReportOutcome: true));
ExpectMsg<OpcUaPublishActor.RebuildCompleted>(PresenceBudget).Degraded.ShouldBeFalse();
}
[Fact]
public void Without_ReportOutcome_nothing_is_replied()
{
var sink = new FaultingSink(throwOnEnsureFolder: true);
var applier = new AddressSpaceApplier(sink, NullLogger<AddressSpaceApplier>.Instance);
var actor = Sys.ActorOf(OpcUaPublishActor.PropsForTests(sink: sink, applier: applier));
actor.Tell(new OpcUaPublishActor.RebuildAddressSpace(
CorrelationId.NewId(), DeploymentId.NewId(), BuildEquipmentArtifact("eq-1")));
// Positive control first: a follow-up REQUESTED outcome must arrive, which proves the first
// message was fully processed before the absence assertion is read.
actor.Tell(new OpcUaPublishActor.RebuildAddressSpace(
CorrelationId.NewId(), DeploymentId.NewId(), BuildEquipmentArtifact("eq-2"), ReportOutcome: true));
var only = ExpectMsg<OpcUaPublishActor.RebuildCompleted>(PresenceBudget);
only.Degraded.ShouldBeTrue();
ExpectNoMsg(TimeSpan.FromMilliseconds(200));
}
private static byte[] BuildEquipmentArtifact(string equipmentId) =>
JsonSerializer.SerializeToUtf8Bytes(new
{
UnsAreas = new[] { new { UnsAreaId = "area-1", ClusterId = "c1", Name = "Area" } },
UnsLines = new[] { new { UnsLineId = "line-1", UnsAreaId = "area-1", Name = "Line" } },
Equipment = new[]
{
new
{
EquipmentId = equipmentId,
DriverInstanceId = (string?)null,
UnsLineId = "line-1",
Name = equipmentId,
MachineCode = equipmentId.ToUpperInvariant(),
},
},
DriverInstances = Array.Empty<object>(),
Namespaces = Array.Empty<object>(),
Tags = Array.Empty<object>(),
ScriptedAlarms = Array.Empty<object>(),
});
/// <summary>A sink that optionally throws from <c>EnsureFolder</c> (a swallowed per-node fault).</summary>
private sealed class FaultingSink(bool throwOnEnsureFolder) : IOpcUaAddressSpaceSink
{
public void WriteValue(string nodeId, object? value, uint statusCode, DateTime sourceTimestampUtc, AddressSpaceRealm realm = AddressSpaceRealm.Uns) { }
public void WriteAlarmQuality(string alarmNodeId, uint statusCode, DateTime sourceTimestampUtc, AddressSpaceRealm realm) { }
public void WriteAlarmCondition(string alarmNodeId, AlarmConditionSnapshot state, DateTime occurredUtc, AddressSpaceRealm realm = AddressSpaceRealm.Uns) { }
public void MaterialiseAlarmCondition(string alarmNodeId, string equipmentNodeId, string displayName, string alarmType, int severity, AddressSpaceRealm realm, bool isNative = false) { }
public void EnsureFolder(string folderNodeId, string? parentNodeId, string displayName, AddressSpaceRealm realm = AddressSpaceRealm.Uns)
{
if (throwOnEnsureFolder) throw new InvalidOperationException("simulated sink fault");
}
public void EnsureVariable(string variableNodeId, string? parentFolderNodeId, string displayName, string dataType, bool writable, AddressSpaceRealm realm, string? historianTagname = null, bool isArray = false, uint? arrayLength = null, bool refuseArrayOnScalar = false) { }
public void RebuildAddressSpace() { }
public void RaiseNodesAddedModelChange(string affectedNodeId, AddressSpaceRealm realm = AddressSpaceRealm.Uns) { }
public void WireAlarmNotifiers(string alarmNodeId, AddressSpaceRealm alarmRealm, IReadOnlyList<string> notifierFolderNodeIds, AddressSpaceRealm notifierFolderRealm) { }
public void AddReference(string sourceNodeId, AddressSpaceRealm sourceRealm, string targetNodeId, AddressSpaceRealm targetRealm, string referenceType = "Organizes") { }
}
}