dohertj2/lmxopcua
1
0
Fork 0
Code Issues 32 Pull Requests Actions Packages Projects Releases Wiki Activity

refactor(fleet): #542 delete the dead fleet-status chain (broadcaster, SignalR bridge, hub, messages)

Browse Source 
#542 filed FleetStatusBroadcaster as blind to site nodes since the Phase 6 per-cluster mesh
split: it derives "the fleet" from central's own Akka gossip ring, which after the split is
just central's own pair. Scoping the cross-mesh rewrite found the deeper truth — the whole
chain is dead end to end, so this deletes it rather than reviving it.

  - Zero consumers. FleetStatusBroadcaster published FleetStatusChanged on the DPS
    "fleet-status" topic; FleetStatusSignalRBridge forwarded it to FleetStatusHub
    ("/hubs/fleet-status"); nothing anywhere connected. The AdminUI bans self-connecting
    HubConnections, no browser JS subscribes, and /fleet (Fleet.razor) reads
    NodeDeploymentState from the ConfigDb plus IClusterRoleInfo on a 5 s poll instead.
  - Its second input never existed. DriverHostStatusHeartbeat had no production sender —
    only the broadcaster's own test told it — so CurrentRevision was always null and the
    30 s staleness -> Degraded flip could never fire.
  - Post-Phase-6 DPS is pair-local, so even the central-only view it did produce was an
    accident of topology rather than a design.

The fleet visibility that actually works is unchanged and already cross-mesh-correct:
/fleet over NodeDeploymentState, and the four Phase 5 per-node telemetry dials.

Precedent for deleting inert machinery rather than activating it: #522 (driver stability
tiers), #507 (discovered-node injection), #525 (GetMemoryFootprint).

Deleted: FleetStatusBroadcaster (+ its nested DriverHostStatusHeartbeat), its admin
singleton registration and FleetStatusBroadcasterKey, FleetStatusSignalRBridge and its
key, FleetStatusHub and its route mapping, and the FleetStatusChanged / FleetNodeStatus /
FleetNodeHealth messages. Both Fleet/ directories survive — ControlPlane keeps
ClusterNodeAddressReconciler, Commons keeps GetDiagnostics. Docs swept where they claimed
a fleet-status channel exists (Telemetry.md's Phase 5 non-migrated channels,
Configuration.md, CLAUDE.md); completed-program plan docs under docs/plans and docs/v2 are
left as the historical record they are.

Build 0 errors, 868 warnings — byte-identical to master's rebuild baseline.
ControlPlane.Tests 231/231, AdminUI.Tests 1109/1109.

Claude-Session: https://claude.ai/code/session_014WNM4vjoVksyyBraTXSZE1
This commit is contained in:
Joseph Doherty
2026-08-06 17:21:30 -04:00
parent 97854a074e
commit f1c188114b
15 changed files with 38 additions and 357 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/Server/ZB.MOM.WW.OtOpcUa.ControlPlane/Fleet/FleetStatusBroadcaster.cs
-168
View File
@@ -1,168 +0,0 @@
using Akka.Actor;
using Akka.Cluster;
using Akka.Cluster.Tools.PublishSubscribe;
using Akka.Event;
using ZB.MOM.WW.OtOpcUa.Commons.Messages.Fleet;
using ZB.MOM.WW.OtOpcUa.Commons.Types;
namespace ZB.MOM.WW.OtOpcUa.ControlPlane.Fleet;
/// <summary>
/// Admin-role cluster singleton that maintains an in-memory <see cref="FleetStatusChanged"/>
/// snapshot from cluster membership + reachability events plus per-node heartbeats, and pushes
/// diffs on the <c>fleet-status</c> DistributedPubSub topic. The SignalR hub layer
/// subscribes to that topic and forwards to browser clients.
///
/// Heartbeat staleness: a node we haven't heard from in <see cref="HeartbeatTimeout"/> flips
/// from <see cref="FleetNodeHealth.Healthy"/> to <see cref="FleetNodeHealth.Degraded"/>. A
/// cluster <c>Unreachable</c> event flips it straight to <see cref="FleetNodeHealth.Unreachable"/>.
/// </summary>
public sealed class FleetStatusBroadcaster : ReceiveActor, IWithTimers
{
public const string Topic = "fleet-status";
public static readonly TimeSpan BroadcastInterval = TimeSpan.FromSeconds(5);
public static readonly TimeSpan HeartbeatTimeout = TimeSpan.FromSeconds(30);
private readonly Akka.Cluster.Cluster _cluster;
private readonly Action<object> _broadcast;
private readonly ILoggingAdapter _log = Context.GetLogger();
private readonly Dictionary<NodeId, NodeRecord> _nodes = new();
/// <summary>Gets or sets the timer scheduler for this actor.</summary>
public ITimerScheduler Timers { get; set; } = null!;
/// <summary>
/// Per-node heartbeat carrying the node's currently-applied revision hash, sent by
/// <c>DriverHostActor</c> periodically.
/// </summary>
public sealed record DriverHostStatusHeartbeat(NodeId NodeId, RevisionHash? CurrentRevision);
/// <summary>Creates actor props for the fleet status broadcaster.</summary>
/// <param name="broadcast">Optional custom broadcast delegate; defaults to distributed pub-sub mediator.</param>
/// <returns>Props configured to create a FleetStatusBroadcaster.</returns>
public static Props Props(Action<object>? broadcast = null) =>
Akka.Actor.Props.Create(() => new FleetStatusBroadcaster(broadcast));
/// <summary>Initializes a new instance of the FleetStatusBroadcaster actor.</summary>
/// <param name="broadcast">Optional custom broadcast delegate; if null, uses distributed pub-sub mediator.</param>
public FleetStatusBroadcaster(Action<object>? broadcast = null)
{
_cluster = Akka.Cluster.Cluster.Get(Context.System);
_broadcast = broadcast ?? (msg =>
DistributedPubSub.Get(Context.System).Mediator.Tell(new Publish(Topic, msg)));
Receive<ClusterEvent.MemberUp>(e => OnMemberUp(e.Member));
Receive<ClusterEvent.MemberRemoved>(e => OnMemberRemoved(e.Member));
Receive<ClusterEvent.UnreachableMember>(e => OnUnreachable(e.Member));
Receive<ClusterEvent.ReachableMember>(e => OnReachable(e.Member));
Receive<ClusterEvent.LeaderChanged>(_ => PublishSnapshot());
Receive<ClusterEvent.RoleLeaderChanged>(_ => PublishSnapshot());
Receive<DriverHostStatusHeartbeat>(OnHeartbeat);
Receive<Tick>(_ => OnTick());
}
/// <inheritdoc />
protected override void PreStart()
{
_cluster.Subscribe(
Self,
ClusterEvent.InitialStateAsEvents,
typeof(ClusterEvent.IMemberEvent),
typeof(ClusterEvent.LeaderChanged),
typeof(ClusterEvent.RoleLeaderChanged),
typeof(ClusterEvent.ReachabilityEvent));
Timers.StartPeriodicTimer("tick", Tick.Instance, BroadcastInterval);
}
/// <inheritdoc />
protected override void PostStop() => _cluster.Unsubscribe(Self);
private void OnMemberUp(Member m)
{
if (!TryNode(m, out var nodeId)) return;
_nodes[nodeId] = _nodes.TryGetValue(nodeId, out var prev)
? prev with { Health = FleetNodeHealth.Healthy, LastSeenUtc = DateTime.UtcNow }
: new NodeRecord(nodeId, FleetNodeHealth.Healthy, CurrentRevision: null, DateTime.UtcNow);
PublishSnapshot();
}
private void OnMemberRemoved(Member m)
{
if (!TryNode(m, out var nodeId)) return;
_nodes.Remove(nodeId);
PublishSnapshot();
}
private void OnUnreachable(Member m)
{
if (!TryNode(m, out var nodeId) || !_nodes.TryGetValue(nodeId, out var rec)) return;
_nodes[nodeId] = rec with { Health = FleetNodeHealth.Unreachable };
PublishSnapshot();
}
private void OnReachable(Member m)
{
if (!TryNode(m, out var nodeId) || !_nodes.TryGetValue(nodeId, out var rec)) return;
_nodes[nodeId] = rec with { Health = FleetNodeHealth.Healthy, LastSeenUtc = DateTime.UtcNow };
PublishSnapshot();
}
private void OnHeartbeat(DriverHostStatusHeartbeat hb)
{
var rec = _nodes.TryGetValue(hb.NodeId, out var prev)
? prev with
{
CurrentRevision = hb.CurrentRevision,
LastSeenUtc = DateTime.UtcNow,
Health = prev.Health == FleetNodeHealth.Unreachable ? prev.Health : FleetNodeHealth.Healthy,
}
: new NodeRecord(hb.NodeId, FleetNodeHealth.Healthy, hb.CurrentRevision, DateTime.UtcNow);
_nodes[hb.NodeId] = rec;
}
/// <summary>
/// Periodic tick — flips heartbeat-stale nodes to <see cref="FleetNodeHealth.Degraded"/> and
/// rebroadcasts the snapshot regardless so SignalR subscribers always have a recent baseline.
/// </summary>
private void OnTick()
{
var stale = DateTime.UtcNow - HeartbeatTimeout;
var flippedCount = 0;
foreach (var kv in _nodes.ToList())
{
if (kv.Value.LastSeenUtc < stale && kv.Value.Health == FleetNodeHealth.Healthy)
{
_nodes[kv.Key] = kv.Value with { Health = FleetNodeHealth.Degraded };
flippedCount++;
}
}
if (flippedCount > 0) _log.Debug("FleetStatusBroadcaster flipped {Count} nodes to Degraded", flippedCount);
PublishSnapshot();
}
private void PublishSnapshot()
{
var statuses = _nodes.Values
.Select(r => new FleetNodeStatus(r.NodeId, r.Health, r.CurrentRevision, r.LastSeenUtc))
.ToList();
var msg = new FleetStatusChanged(statuses, CurrentDeployment: null, DateTime.UtcNow, CorrelationId.NewId());
_broadcast(msg);
}
private static bool TryNode(Member m, out NodeId nodeId)
{
nodeId = default;
var host = m.Address.Host;
if (string.IsNullOrWhiteSpace(host)) return false;
nodeId = NodeId.Parse(host);
return true;
}
private sealed record NodeRecord(NodeId NodeId, FleetNodeHealth Health, RevisionHash? CurrentRevision, DateTime LastSeenUtc);
public sealed class Tick
{
public static readonly Tick Instance = new();
private Tick() { }
}
}
src/Server/ZB.MOM.WW.OtOpcUa.ControlPlane/ServiceCollectionExtensions.cs
+6 -7
View File
@@ -26,12 +26,11 @@ public static class ServiceCollectionExtensions
public const string ConfigPublishSingletonName = "config-publish";
public const string AdminOperationsSingletonName = "admin-operations";
public const string AuditWriterSingletonName = "audit-writer";
public const string FleetStatusSingletonName = "fleet-status";
public const string RedundancyStateSingletonName = "redundancy-state";
public const string ClusterNodeAddressReconcilerSingletonName = "cluster-node-address-reconciler";
/// <summary>
/// Registers the admin-role cluster singletons on the AkkaConfigurationBuilder — five with
/// Registers the admin-role cluster singletons on the AkkaConfigurationBuilder — three with
/// registry proxies, plus the address reconciler, which is proxy-less because nothing addresses it.
/// Must be called against the same builder used by <c>AkkaHostedService</c> so the singletons
/// share the host's ActorSystem.
@@ -139,10 +138,11 @@ public static class ServiceCollectionExtensions
},
singletonOptions);
builder.WithSingleton<FleetStatusBroadcasterKey>(
FleetStatusSingletonName,
(system, registry, resolver) => FleetStatusBroadcaster.Props(),
singletonOptions);
// Gitea #542 (2026-08-06): the fleet-status singleton is NO LONGER registered here — the
// whole FleetStatusBroadcaster → DPS → SignalR bridge → FleetStatusHub chain was DELETED.
// It had zero clients end-to-end (no browser JS, no HubConnection — /fleet reads
// NodeDeploymentState from the ConfigDb on a poll), its DriverHostStatusHeartbeat input had
// no production sender, and after mesh Phase 6 its "fleet" was only central's own pair.
// Per-cluster mesh Phase 6: the redundancy-state singleton is NO LONGER registered here.
// It was an admin-scoped, fleet-wide singleton — one Primary elected across the whole mesh.
@@ -253,7 +253,6 @@ public static class ServiceCollectionExtensions
public sealed class ConfigPublishCoordinatorKey { }
public sealed class AdminOperationsActorKey { }
public sealed class AuditWriterActorKey { }
public sealed class FleetStatusBroadcasterKey { }
public sealed class RedundancyStateActorKey { }
public sealed class ClusterNodeAddressReconcilerKey { }
public sealed class CentralCommunicationActorKey { }