fix(health): rows 80+73 — unconfigured replication reports null, reconciler presence surfaced

ROW 80 (observability correctness). An unreplicated site published
LocalDbReplicationConnected=false / LocalDbOplogBacklog=0, contradicting the
nullable semantics every doc and the proto already state ("null means no
reading, NOT disconnected with an empty backlog").

Root cause: the replication engine is registered unconditionally and idles
without a peer, so ISyncStatus answers false/0 on a node that was never meant to
replicate, and LocalDbReplicationStatusReporter passed that straight through.
Two consequences: an operator read connected:false on a healthy site as a fault,
and central's null branches — row 33's conditional KPI emission and
sustained-badge anchor exclusion — never received the null they exist for, so
rig site-b/site-c emitted a permanently misleading
localDbReplicationConnected=0 series.

Fix mirrors the SiteEventLogReplicationCheck precedent: the Host evaluates
SiteLocalDbSetup.ReplicationIsConfigured ONCE and passes it to
AddLocalDbReplicationHealthBridge as a third provider; unconfigured
short-circuits the probe to (null, null). HealthMonitoring never reads LocalDb
config itself, and the same captured value now also feeds
SiteEventLogReplicationCheck so the two surfaces cannot disagree.
ISiteHealthCollector.SetLocalDbReplicationStatus widens to bool? — the
non-nullable bool was the type-level root cause (null was inexpressible).
false now means exactly one thing: configured, peer session down = a real fault.

Wire/central/UI needed no change: the proto already carries BoolValue/Int64Value
wrappers and the aggregator + KPI source already branch on null. This makes code
match doc, so no doc claim was weakened.

Watch the trap that caused this: configured-ness is PeerAddress OR ApiKey (only
the dialling half sets a peer address). The pre-existing SiteLocalDbWiringTests
fixture sets ApiKey and IS a replicating node, so its false/0 assertion is
correct and is kept, with a comment saying so; the new
UnreplicatedSiteLocalDbHealthTests fixture is the genuinely-unreplicated one.

ROW 73 (silent loss of self-heal). AkkaHostedService skips
SiteReconciliationActor when no IDeploymentConfigFetcher is resolvable, and that
skip was a lone boot Warning — a node with no self-heal path at all went on
reporting itself perfectly healthy forever.

Adds tri-state SiteHealthReport.StartupReconcilerActive (null = not reported,
i.e. the central self-report or a pre-row-73 site build; true = running;
false = booted without it), latched once at boot by the Host in both branches.
Additive proto field 29 (BoolValue, vendored C# regenerated via
docker/regen-proto.sh centralcontrol; csproj restored, verified empty diff).
The skip branch also raises its log to Error — no configuration makes the
fetcher legitimately absent on a site node — and writes a best-effort
deployment/Error site event so the evidence outlives the boot log.

Central treatment: a "No self-heal" WARNING badge on the site tile, not a
failure. Config still reaches the node via LocalDb replication while it is up;
what it lost is recovery from a deploy that landed while it was down — degraded,
not offline. Only the explicit false is badged (badging null would light up the
whole fleet the moment central is upgraded ahead of the sites), and offline
sites are skipped, matching the replication-backlog badge.

Fail-before evidence: with the row-80 gate behaviourally reverted,
UnreplicatedSite_HealthReport_ReportsNullForBothLocalDbFields fails
"Expected: null / Actual: False" and 3 of the 8 new reporter unit tests fail.

Tests: HealthMonitoring 128, Host 518, CentralUI 1051, Communication 717,
Commons 715, SiteEventLogging 81 — all green.

src/ZB.MOM.WW.ScadaBridge.HealthMonitoring/ISiteHealthCollector.cs

@@ -162,14 +162,38 @@ public interface ISiteHealthCollector
     /// service. Point-in-time: values are NOT reset on <see cref="CollectReport"/>.
     /// </summary>
     /// <param name="connected">
-    /// Whether a sync session is currently running. <see langword="false"/> is the
-    /// normal state on a node with no peer configured — replication ships default-OFF.
+    /// Whether a sync session is currently running, or <see langword="null"/> when there
+    /// is no reading to give — which is the state of a node with replication UNCONFIGURED
+    /// (register row 80). <see langword="false"/> means "replication is configured on this
+    /// node and the peer session is currently down", i.e. a genuine fault worth a badge;
+    /// reporting that on a deliberately-unreplicated site reads as a fault that isn't one.
     /// </param>
     /// <param name="oplogBacklog">
-    /// Unacked oplog entries, or <see langword="null"/> when unknown. Pass null through
-    /// unchanged: a failed poll rendered as 0 would report a broken pair as healthy.
+    /// Unacked oplog entries, or <see langword="null"/> when unknown (failed poll) or not
+    /// applicable (replication unconfigured). Pass null through unchanged: a failed poll
+    /// rendered as 0 would report a broken pair as healthy.
     /// </param>
-    void SetLocalDbReplicationStatus(bool connected, long? oplogBacklog)
+    void SetLocalDbReplicationStatus(bool? connected, long? oplogBacklog)
+    {
+        // Default no-op so test fakes do not need to be updated.
+    }
+
+    /// <summary>
+    /// Record whether this node spawned its <c>SiteReconciliationActor</c> at boot, for the
+    /// <c>StartupReconcilerActive</c> field of every subsequent <see cref="CollectReport"/>
+    /// (deferred-work register row 73).
+    /// </summary>
+    /// <remarks>
+    /// Called ONCE at boot by the Host, which owns the spawn decision — this is a latched
+    /// boot fact, not a polled gauge, so there is no reporter hosted service behind it.
+    /// A collector that is never told (central's self-report, or any test host) leaves the
+    /// field null, which reads as "not reported" rather than "reconciler absent".
+    /// </remarks>
+    /// <param name="active">
+    /// <see langword="true"/> when the reconciler was created; <see langword="false"/> when
+    /// it was skipped because no <c>IDeploymentConfigFetcher</c> was available.
+    /// </param>
+    void SetStartupReconcilerActive(bool active)
     {
         // Default no-op so test fakes do not need to be updated.
     }

src/ZB.MOM.WW.ScadaBridge.HealthMonitoring/LocalDbReplicationStatusReporter.cs

@@ -22,6 +22,23 @@ namespace ZB.MOM.WW.ScadaBridge.HealthMonitoring;
 /// report a replication pair that cannot read its own oplog as perfectly healthy.
 /// </para>
 /// <para>
+/// <b>An UNCONFIGURED node reports null for both fields</b> (register row 80). The
+/// replication engine is registered unconditionally and sits inert without a peer, so
+/// <c>ISyncStatus</c> answers <c>Connected=false</c> / <c>OplogBacklog=0</c> on a site
+/// that was never meant to replicate at all. Publishing that pair is actively misleading:
+/// a dashboard reader sees "connected: false" on a healthy unreplicated site and reads a
+/// fault, and the central null-branch machinery (conditional KPI emission, sustained-badge
+/// anchor exclusion) never receives the null it was built for. The
+/// <c>replicationConfiguredProvider</c> gate therefore short-circuits the probe and pushes
+/// <c>(null, null)</c> — "no reading", per the documented nullable semantics.
+/// </para>
+/// <para>
+/// <b>The gate is supplied by the Host, not read from configuration here</b>, mirroring
+/// <c>SiteEventLogReplicationCheck</c>: the Host owns
+/// <c>SiteLocalDbSetup.ReplicationIsConfigured</c> (the PeerAddress-OR-ApiKey rule) and
+/// HealthMonitoring must not learn to read LocalDb configuration itself.
+/// </para>
+/// <para>
 /// <b>Cadence.</b> 30 s, matching <see cref="SiteEventLogFailureCountReporter"/> and
 /// <c>SiteAuditBacklogReporter</c>. Any exception during a probe is logged and swallowed;
 /// the next tick retries.
@@ -34,6 +51,7 @@ public sealed class LocalDbReplicationStatusReporter : IHostedService, IDisposab
 
     private readonly Func<bool> _connectedProvider;
     private readonly Func<long?> _oplogBacklogProvider;
+    private readonly Func<bool> _replicationConfiguredProvider;
     private readonly ISiteHealthCollector _collector;
     private readonly ILogger<LocalDbReplicationStatusReporter> _logger;
     private readonly TimeSpan _refreshInterval;
@@ -43,18 +61,26 @@ public sealed class LocalDbReplicationStatusReporter : IHostedService, IDisposab
     /// <summary>Initializes a new instance of <see cref="LocalDbReplicationStatusReporter"/>.</summary>
     /// <param name="connectedProvider">Reads whether a peer sync session is currently running.</param>
     /// <param name="oplogBacklogProvider">Reads the unacked oplog backlog, or null when unknown.</param>
+    /// <param name="replicationConfiguredProvider">
+    /// Reads whether replication is configured on this node at all (the Host's
+    /// <c>SiteLocalDbSetup.ReplicationIsConfigured</c>). When it answers false, both fields
+    /// are published as null rather than as the engine's inert false/0 defaults.
+    /// </param>
     /// <param name="collector">The site health collector receiving the snapshot.</param>
     /// <param name="logger">Logger instance.</param>
     /// <param name="refreshInterval">Poll interval override; defaults to 30 s.</param>
     public LocalDbReplicationStatusReporter(
         Func<bool> connectedProvider,
         Func<long?> oplogBacklogProvider,
+        Func<bool> replicationConfiguredProvider,
         ISiteHealthCollector collector,
         ILogger<LocalDbReplicationStatusReporter> logger,
         TimeSpan? refreshInterval = null)
     {
         _connectedProvider = connectedProvider ?? throw new ArgumentNullException(nameof(connectedProvider));
         _oplogBacklogProvider = oplogBacklogProvider ?? throw new ArgumentNullException(nameof(oplogBacklogProvider));
+        _replicationConfiguredProvider = replicationConfiguredProvider
+            ?? throw new ArgumentNullException(nameof(replicationConfiguredProvider));
         _collector = collector ?? throw new ArgumentNullException(nameof(collector));
         _logger = logger ?? throw new ArgumentNullException(nameof(logger));
         _refreshInterval = refreshInterval ?? DefaultRefreshInterval;
@@ -114,6 +140,14 @@ public sealed class LocalDbReplicationStatusReporter : IHostedService, IDisposab
     {
         try
         {
+            if (!_replicationConfiguredProvider())
+            {
+                // Replication unconfigured: "no reading", NOT "disconnected with an empty
+                // backlog". Publishing the inert engine's false/0 here is register row 80.
+                _collector.SetLocalDbReplicationStatus(null, null);
+                return;
+            }
+
             _collector.SetLocalDbReplicationStatus(
                 _connectedProvider(), _oplogBacklogProvider());
         }

src/ZB.MOM.WW.ScadaBridge.HealthMonitoring/ServiceCollectionExtensions.cs

@@ -170,19 +170,33 @@ public static class ServiceCollectionExtensions
     /// a reference on the replication library, matching
     /// <see cref="AddSiteEventLogHealthMetricsBridge"/>. Pass the backlog through as
     /// nullable — coercing null to 0 would report an unreadable oplog as a healthy one.
+    /// <para>
+    /// <paramref name="replicationConfiguredProvider"/> is the register-row-80 gate: the
+    /// replication engine is registered unconditionally and answers <c>false</c>/<c>0</c>
+    /// while inert, so without it a deliberately-unreplicated site publishes
+    /// "disconnected, empty backlog" instead of the documented "no reading" null. The
+    /// caller supplies it — the Host owns <c>SiteLocalDbSetup.ReplicationIsConfigured</c>,
+    /// exactly as it supplies <c>SiteEventLogReplicationCheck</c> to SiteEventLogging.
+    /// </para>
     /// </remarks>
     /// <param name="services">The service collection to register into.</param>
     /// <param name="connectedProvider">Given the root provider, returns a reader for "a sync session is running".</param>
     /// <param name="oplogBacklogProvider">Given the root provider, returns a reader for the unacked backlog (null = unknown).</param>
+    /// <param name="replicationConfiguredProvider">
+    /// Given the root provider, returns a reader for "replication is configured on this node".
+    /// False makes the reporter publish null for BOTH fields.
+    /// </param>
     /// <returns>The same <see cref="IServiceCollection"/> for chaining.</returns>
     public static IServiceCollection AddLocalDbReplicationHealthBridge(
         this IServiceCollection services,
         Func<IServiceProvider, Func<bool>> connectedProvider,
-        Func<IServiceProvider, Func<long?>> oplogBacklogProvider)
+        Func<IServiceProvider, Func<long?>> oplogBacklogProvider,
+        Func<IServiceProvider, Func<bool>> replicationConfiguredProvider)
     {
         ArgumentNullException.ThrowIfNull(services);
         ArgumentNullException.ThrowIfNull(connectedProvider);
         ArgumentNullException.ThrowIfNull(oplogBacklogProvider);
+        ArgumentNullException.ThrowIfNull(replicationConfiguredProvider);
 
         if (services.Any(d => d.ServiceType == typeof(LocalDbReplicationHealthBridgeMarker)))
         {
@@ -193,6 +207,7 @@ public static class ServiceCollectionExtensions
         services.AddHostedService(sp => new LocalDbReplicationStatusReporter(
             connectedProvider(sp),
             oplogBacklogProvider(sp),
+            replicationConfiguredProvider(sp),
             sp.GetRequiredService<ISiteHealthCollector>(),
             sp.GetRequiredService<ILogger<LocalDbReplicationStatusReporter>>()));
 

src/ZB.MOM.WW.ScadaBridge.HealthMonitoring/SiteHealthCollector.cs

@@ -20,8 +20,18 @@ public class SiteHealthCollector : ISiteHealthCollector
     private long _siteEventLogWriteFailures;
     // One volatile tuple rather than two independent fields: a torn read that paired a
     // fresh Connected with a stale backlog would be indistinguishable from a real state.
-    // Null = the reporter has not yet run (or replication is not wired).
-    private volatile Tuple<bool, long?>? _localDbReplicationStatus;
+    // Null = the reporter has not yet run. Once it HAS run, an unreplicated node latches
+    // (null, null) — "no reading" — rather than the inert engine's (false, 0); see
+    // LocalDbReplicationStatusReporter and register row 80.
+    private volatile Tuple<bool?, long?>? _localDbReplicationStatus;
+    // Latched once at boot by the Host (register row 73). A tri-state int rather than a
+    // bool? because `volatile` cannot be applied to Nullable<bool> (a struct) and this is
+    // written on the startup thread but read on the report-sender thread:
+    //   0 = never told  -> report null ("not reported": central self-report, or a test host)
+    //   1 = reconciler spawned
+    //  -1 = reconciler SKIPPED (no IDeploymentConfigFetcher)
+    // Null must stay distinct from -1: "we don't know" is not "self-heal is missing".
+    private volatile int _startupReconcilerActive;
     private readonly ConcurrentDictionary<string, ConnectionHealth> _connectionStatuses = new();
     private readonly ConcurrentDictionary<string, TagResolutionStatus> _tagResolutionCounts = new();
     private readonly ConcurrentDictionary<string, string> _connectionEndpoints = new();
@@ -109,11 +119,17 @@ public class SiteHealthCollector : ISiteHealthCollector
     }
 
     /// <inheritdoc />
-    public void SetLocalDbReplicationStatus(bool connected, long? oplogBacklog)
+    public void SetLocalDbReplicationStatus(bool? connected, long? oplogBacklog)
     {
         _localDbReplicationStatus = Tuple.Create(connected, oplogBacklog);
     }
 
+    /// <inheritdoc />
+    public void SetStartupReconcilerActive(bool active)
+    {
+        _startupReconcilerActive = active ? 1 : -1;
+    }
+
     /// <inheritdoc />
     public void UpdateConnectionHealth(string connectionName, ConnectionHealth health)
     {
@@ -256,6 +272,7 @@ public class SiteHealthCollector : ISiteHealthCollector
         // Single read of the volatile tuple — two reads could straddle a reporter tick and
         // pair a fresh Connected with a stale backlog.
         var localDbReplication = _localDbReplicationStatus;
+        var startupReconciler = _startupReconcilerActive;
 
         // Snapshot current connection and tag resolution state
         var connectionStatuses = new Dictionary<string, ConnectionHealth>(_connectionStatuses);
@@ -307,7 +324,10 @@ public class SiteHealthCollector : ISiteHealthCollector
             // with an empty backlog".
             LocalDbReplicationConnected = localDbReplication?.Item1,
             LocalDbOplogBacklog = localDbReplication?.Item2,
-            SiteStreamAlarmDropCount = Interlocked.Read(ref _siteStreamAlarmDropCount)
+            SiteStreamAlarmDropCount = Interlocked.Read(ref _siteStreamAlarmDropCount),
+            // Tri-state: 0 stays null ("not reported"), which must never collapse into
+            // false ("self-heal is missing"). Register row 73.
+            StartupReconcilerActive = startupReconciler == 0 ? null : startupReconciler > 0
         };
     }
 }