Fix second-pass review findings: subscription leak on rebuild, metrics accuracy, and MxAccess startup recovery

- Preserve and replay subscription ref counts across address space rebuilds to prevent MXAccess subscription leaks
- Mark read timeouts and write failures as unsuccessful in PerformanceMetrics for accurate health reporting
- Add deferred MxAccess reconnect path when initial connection fails at startup
- Update code review document with verified completions and new findings
- Add covering tests for all fixes

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
Joseph Doherty
2026-03-25 09:41:12 -04:00
parent 71254e005e
commit 09ed15bdda
12 changed files with 307 additions and 51 deletions

View File

@@ -29,6 +29,7 @@ namespace ZB.MOM.WW.LmxOpcUa.Tests.Helpers
public int UnregisterCallCount { get; private set; }
public bool ShouldFailRegister { get; set; }
public bool ShouldFailWrite { get; set; }
public bool SkipWriteCompleteCallback { get; set; }
public int WriteCompleteStatus { get; set; } = 0; // 0 = success
public int Register(string clientName)
@@ -87,7 +88,8 @@ namespace ZB.MOM.WW.LmxOpcUa.Tests.Helpers
status[0].success = 0;
status[0].detail = (short)WriteCompleteStatus;
}
OnWriteComplete?.Invoke(_connectionHandle, itemHandle, ref status);
if (!SkipWriteCompleteCallback)
OnWriteComplete?.Invoke(_connectionHandle, itemHandle, ref status);
}
/// <summary>

View File

@@ -259,5 +259,32 @@ namespace ZB.MOM.WW.LmxOpcUa.Tests.Integration
await fixture.DisposeAsync();
}
}
[Fact]
public async Task Rebuild_PreservesSubscriptionBookkeeping_ForSurvivingNodes()
{
var fixture = OpcUaServerFixture.WithFakeMxAccessClient();
await fixture.InitializeAsync();
try
{
var nodeManager = fixture.Service.NodeManagerInstance!;
var mxClient = fixture.MxAccessClient!;
nodeManager.SubscribeTag("TestMachine_001.MachineID");
mxClient.ActiveSubscriptionCount.ShouldBe(1);
fixture.Service.TriggerRebuild();
await Task.Delay(200);
mxClient.ActiveSubscriptionCount.ShouldBe(1);
nodeManager.UnsubscribeTag("TestMachine_001.MachineID");
mxClient.ActiveSubscriptionCount.ShouldBe(0);
}
finally
{
await fixture.DisposeAsync();
}
}
}
}

View File

@@ -69,6 +69,20 @@ namespace ZB.MOM.WW.LmxOpcUa.Tests.MxAccess
result.Quality.ShouldBe(Quality.BadCommFailure);
}
[Fact]
public async Task Read_Timeout_RecordsFailedMetrics()
{
await _client.ConnectAsync();
var result = await _client.ReadAsync("TestTag.Attr");
result.Quality.ShouldBe(Quality.BadCommFailure);
var stats = _metrics.GetStatistics();
stats.ShouldContainKey("Read");
stats["Read"].TotalCount.ShouldBe(1);
stats["Read"].SuccessCount.ShouldBe(0);
}
[Fact]
public async Task Write_NotConnected_ReturnsFalse()
{
@@ -97,6 +111,21 @@ namespace ZB.MOM.WW.LmxOpcUa.Tests.MxAccess
result.ShouldBe(false);
}
[Fact]
public async Task Write_Timeout_ReturnsFalse_AndRecordsFailedMetrics()
{
await _client.ConnectAsync();
_proxy.SkipWriteCompleteCallback = true;
var result = await _client.WriteAsync("TestTag.Attr", 42);
result.ShouldBe(false);
var stats = _metrics.GetStatistics();
stats.ShouldContainKey("Write");
stats["Write"].TotalCount.ShouldBe(1);
stats["Write"].SuccessCount.ShouldBe(0);
}
[Fact]
public async Task Read_RecordsMetrics()
{

View File

@@ -102,6 +102,42 @@ namespace ZB.MOM.WW.LmxOpcUa.Tests.MxAccess
callbackInvoked.ShouldBe(true);
}
[Fact]
public async Task OneShotRead_DoesNotRemovePersistentSubscription_OnReconnect()
{
await _client.ConnectAsync();
var callbackInvoked = false;
await _client.SubscribeAsync("TestTag.Attr", (_, _) => callbackInvoked = true);
var readTask = _client.ReadAsync("TestTag.Attr");
await Task.Delay(50);
_proxy.SimulateDataChangeByAddress("TestTag.Attr", 42, 192);
(await readTask).Value.ShouldBe(42);
callbackInvoked = false;
await _client.ReconnectAsync();
_proxy.SimulateDataChangeByAddress("TestTag.Attr", "after_reconnect", 192);
callbackInvoked.ShouldBe(true);
_client.ActiveSubscriptionCount.ShouldBe(1);
}
[Fact]
public async Task OneShotWrite_DoesNotBreakPersistentUnsubscribe()
{
await _client.ConnectAsync();
await _client.SubscribeAsync("TestTag.Attr", (_, _) => { });
_proxy.Items.Values.ShouldContain("TestTag.Attr");
var writeResult = await _client.WriteAsync("TestTag.Attr", 7);
writeResult.ShouldBe(true);
await _client.UnsubscribeAsync("TestTag.Attr");
_client.ActiveSubscriptionCount.ShouldBe(0);
_proxy.Items.Values.ShouldNotContain("TestTag.Attr");
}
[Fact]
public async Task ProbeTag_SubscribedOnConnect()
{

View File

@@ -1,4 +1,5 @@
using System.Collections.Generic;
using System.Threading.Tasks;
using Shouldly;
using Xunit;
using ZB.MOM.WW.LmxOpcUa.Host;
@@ -69,5 +70,60 @@ namespace ZB.MOM.WW.LmxOpcUa.Tests.Wiring
service.Stop();
}
}
[Fact]
public async Task Start_WhenMxAccessIsInitiallyDown_MonitorReconnectsInBackground()
{
var config = new AppConfiguration
{
OpcUa = new OpcUaConfiguration
{
Port = 14841,
GalaxyName = "TestGalaxy",
EndpointPath = "/LmxOpcUa"
},
MxAccess = new MxAccessConfiguration
{
ClientName = "Test",
MonitorIntervalSeconds = 1,
AutoReconnect = true
},
GalaxyRepository = new GalaxyRepositoryConfiguration(),
Dashboard = new DashboardConfiguration { Enabled = false }
};
var proxy = new FakeMxProxy { ShouldFailRegister = true };
var repo = new FakeGalaxyRepository
{
Hierarchy = new List<GalaxyObjectInfo>
{
new GalaxyObjectInfo { GobjectId = 1, TagName = "TestObj", BrowseName = "TestObj", ParentGobjectId = 0, IsArea = false }
},
Attributes = new List<GalaxyAttributeInfo>
{
new GalaxyAttributeInfo { GobjectId = 1, TagName = "TestObj", AttributeName = "TestAttr", FullTagReference = "TestObj.TestAttr", MxDataType = 5, IsArray = false }
}
};
var service = new OpcUaService(config, proxy, repo);
service.Start();
try
{
service.ServerHost.ShouldNotBeNull();
service.MxClient.ShouldNotBeNull();
service.MxClient!.State.ShouldBe(ConnectionState.Error);
proxy.ShouldFailRegister = false;
await Task.Delay(2500);
service.MxClient.State.ShouldBe(ConnectionState.Connected);
proxy.RegisterCallCount.ShouldBeGreaterThan(1);
}
finally
{
service.Stop();
}
}
}
}