fix(localdb): phase-2 live gate — 4 production defects found and fixed
Gate record: docs/plans/2026-07-20-localdb-phase2-live-gate.md. Checks 1, 2, 5, 6 pass. Checks 3 and 4 are NOT satisfied — the defect they were meant to confirm turned out to be the opposite of what the plan assumed. Three defects crash-looped every driver node before check 1 could even run: 1. An empty ServerHistorian:ApiKey kills the host. ServerHistorianOptions- Validator exists to turn exactly that class of failure into a named OptionsValidationException, but its documented fail tier explicitly excluded ApiKey on the reasoning that a keyless client "degrades — the gateway rejects calls". It does not: the client validates its own options at construction, so the process dies during Akka startup and never makes a call. 2. UseTls disagreeing with the endpoint scheme kills the host too, in both directions (both messages confirmed in the shipped client assembly). Moving an endpoint from https to http without clearing UseTls is an ordinary migration slip. 3. Plaintext h2c was UNREACHABLE. HistorianGatewayClientAdapter forwarded the TLS-only options unconditionally, and AllowUntrustedServerCertificate defaults to false, so it always sent RequireCertificateValidation=true — which the client rejects outright when UseTls=false. Every http:// deployment crashed, though the scheme is documented as the supported way to select h2c, and the only workaround was to assert a certificate posture for a connection that has no certificate. The fourth was the blocker, and it is Phase 2's own: 4. The drain gate deferred to a Primary that cannot deliver. Redundancy roles are elected CLUSTER-WIDE; the alarm queue is PAIR-LOCAL. On the rig the elected driver Primary is central-1 — it carries the driver Akka role, replicates nobody's LocalDb and does not even run the alarm historian — so every driver node logged "Historian drain suspended", including the two site-b nodes that have no peer at all. Nothing drained anywhere, where before Phase 2 it drained fine. The cost is not a duplicate; it is the buffer growing to the capacity wall and evicting the audit trail it exists to protect. Fixed in three layers: a separate ShouldDrainAlarmHistory policy (unknown role drains; the two gates now deliberately disagree, and a test pins that); peer- host matching in DriverHostActor so a node stands down only for a Primary holding its rows; and AddAlarmHistorian short-circuiting the gate when replication is unconfigured — testing BOTH Replication:PeerAddress and SyncListenPort, since only the dialing half sets the former while both halves share the queue. Every one of these follows from the asymmetry: a false allow costs a duplicate row, which at-least-once delivery already accepts and payload-hash ids collapse; a false deny loses data silently. A third vacuous test, caught by the same delete-the-guard discipline: the role-view tests stayed green with the guard removed, because AwaitAssert polls until an assertion passes and the assertion was "reads open" — which is the SEEDED value, satisfied at the first poll before the actor processed anything. They now assert the sequence of published values through a recording view; the control then goes red for exactly the cases that matter. Migration evidence: 11 legacy rows across two deliberately overlapping files converged to exactly 9 identical rows on both nodes, proving D-6's payload-hash identity on real nodes rather than in a fixture. Open design fork, recorded in the gate doc rather than decided here: a pair cannot currently identify its own Primary, so both halves drain. Safe in every topology — nothing loses data — but the gate's de-duplication benefit is unrealised until roles are scoped per pair. Claude-Session: https://claude.ai/code/session_01GASWkNEi68FSCtvr6rLoEW
This commit is contained in:
Joseph Doherty
@@ -107,6 +107,34 @@ public static class ServiceCollectionExtensions
|
||||
// posture for a deployment that runs no redundancy at all.
|
||||
services.TryAddSingleton<IRedundancyRoleView, RedundancyRoleView>();
|
||||
|
||||
// THE GATE ONLY APPLIES TO A REPLICATED QUEUE. Standing down is only ever safe because some
|
||||
// other node holds the same rows and will send them instead — and the only node that does is
|
||||
// this node's LocalDb replication peer. Without replication configured, these rows exist here
|
||||
// and nowhere else, so deferring to anyone means they are never delivered by anyone.
|
||||
//
|
||||
// This is not hypothetical. The redundancy role is a CLUSTER-WIDE election
|
||||
// (RedundancyStateActor keys on Akka's RoleLeader("driver")), while the queue is PAIR-LOCAL.
|
||||
// On the docker-dev rig the elected driver Primary is a central node — which carries the
|
||||
// driver Akka role, replicates nobody's LocalDb, and does not even run the alarm historian —
|
||||
// so every site node dutifully suspended its drain in favour of a node that could not
|
||||
// possibly deliver its events. Scoping the gate to "is my queue actually shared?" is what
|
||||
// keeps the two scopes from disagreeing.
|
||||
// BOTH halves of a pair share the queue, but only one of them dials: the initiator sets
|
||||
// Replication:PeerAddress, its partner only sets SyncListenPort and waits. Testing the dial
|
||||
// side alone would leave the listening half permanently ungated — one drainer per pair by
|
||||
// accident rather than by role, and the wrong one whenever the roles swap.
|
||||
var replicated =
|
||||
!string.IsNullOrWhiteSpace(configuration["LocalDb:Replication:PeerAddress"])
|
||||
|| !string.IsNullOrWhiteSpace(configuration["LocalDb:SyncListenPort"]);
|
||||
|
||||
if (!replicated)
|
||||
{
|
||||
Serilog.Log.Logger.ForContext<LocalDbStoreAndForwardSink>().Information(
|
||||
"Alarm historian: LocalDb replication is not configured, so this node's queue is not "
|
||||
+ "shared with any peer and the Primary drain gate does not apply — this node always "
|
||||
+ "drains its own alarm queue.");
|
||||
}
|
||||
|
||||
services.AddSingleton<IAlarmHistorianSink>(sp =>
|
||||
{
|
||||
// LocalDbStoreAndForwardSink takes a Serilog ILogger (not Microsoft.Extensions.Logging).
|
||||
@@ -121,7 +149,7 @@ public static class ServiceCollectionExtensions
|
||||
capacity: opts.Capacity,
|
||||
deadLetterRetention: TimeSpan.FromDays(opts.DeadLetterRetentionDays),
|
||||
maxAttempts: opts.MaxAttempts,
|
||||
drainGate: () => roleView.ShouldServiceAsPrimary);
|
||||
drainGate: () => !replicated || roleView.ShouldDrainAlarmHistory);
|
||||
sink.StartDrainLoop(TimeSpan.FromSeconds(opts.DrainIntervalSeconds));
|
||||
return sink;
|
||||
});
|
||||
@@ -252,6 +280,16 @@ public static class ServiceCollectionExtensions
|
||||
// Registered by AddAlarmHistorian; absent when no durable sink is configured, in which
|
||||
// case there is nothing downstream to inform.
|
||||
var redundancyRoleView = resolver.GetService<IRedundancyRoleView>();
|
||||
// Host of this node's LocalDb replication partner: the ONLY node that holds a copy of this
|
||||
// node's alarm queue, and so the only node it may stand down in favour of. Null when this
|
||||
// node dials nobody, which correctly means "never stand down".
|
||||
var replicationPeerHost =
|
||||
Uri.TryCreate(
|
||||
resolver.GetService<IConfiguration>()?["LocalDb:Replication:PeerAddress"],
|
||||
UriKind.Absolute,
|
||||
out var peerUri)
|
||||
? peerUri.Host
|
||||
: null;
|
||||
// Root script logger backs the ScriptedAlarm host's engine + script logging. Registered in
|
||||
// Host DI inside the hasDriver block; may be absent in some role configs / test harnesses,
|
||||
// in which case the DriverHostActor gracefully skips spawning the ScriptedAlarm host.
|
||||
@@ -371,7 +409,8 @@ public static class ServiceCollectionExtensions
|
||||
scriptRootLogger: scriptRootLogger,
|
||||
invokerFactory: invokerFactory,
|
||||
deploymentArtifactCache: deploymentArtifactCache,
|
||||
redundancyRoleView: redundancyRoleView),
|
||||
redundancyRoleView: redundancyRoleView,
|
||||
replicationPeerHost: replicationPeerHost),
|
||||
DriverHostActorName);
|
||||
registry.Register<DriverHostActorKey>(driverHost);
|
||||
|
||||
|
||||
Reference in New Issue
Block a user