fix(store-and-forward): resolve S&F delivery + replication wiring (3 Critical findings)

Resolves StoreAndForward-001, ExternalSystemGateway-001, NotificationService-001
— one systemic gap where buffered messages were persisted but never delivered,
and the active node never replicated its buffer to the standby.

Delivery handlers (ExternalSystemGateway-001 / NotificationService-001):
- AkkaHostedService registers delivery handlers for the ExternalSystem,
  CachedDbWrite and Notification categories after StoreAndForwardService starts;
  each resolves its scoped consumer in a fresh DI scope.
- ExternalSystemClient, DatabaseGateway and NotificationDeliveryService each
  gain a DeliverBufferedAsync method: re-resolve the target and re-attempt
  delivery, returning true/false/throwing per the transient-vs-permanent contract.
- EnqueueAsync gains an attemptImmediateDelivery flag; CachedCallAsync and
  NotificationDeliveryService.SendAsync pass false (they already attempted
  delivery themselves) so registering a handler does not dispatch twice.

Replication (StoreAndForward-001):
- ReplicationService is injected into StoreAndForwardService; a new BufferAsync
  helper replicates every enqueue, and successful-retry removes and parks are
  replicated too. Fire-and-forget, no-op when replication is disabled.

Tests: StoreAndForwardReplicationTests (Add/Remove/Park observed),
attemptImmediateDelivery behaviour, and DeliverBufferedAsync paths for each
consumer. Full solution builds; StoreAndForward/ExternalSystemGateway/
NotificationService suites green.
This commit is contained in:
Joseph Doherty
2026-05-16 18:58:11 -04:00
parent a9bd7ee37c
commit 61253e3269
15 changed files with 538 additions and 37 deletions

View File

@@ -8,7 +8,7 @@
| Last reviewed | 2026-05-16 |
| Reviewer | claude-agent |
| Commit reviewed | `9c60592` |
| Open findings | 14 |
| Open findings | 13 |
## Summary
@@ -53,7 +53,7 @@ requirements (timeout, retry settings) that are declared but not implemented.
|--|--|
| Severity | Critical |
| Category | Error handling & resilience |
| Status | Open |
| Status | Resolved |
| Location | `src/ScadaLink.ExternalSystemGateway/ExternalSystemClient.cs:109`, `src/ScadaLink.ExternalSystemGateway/DatabaseGateway.cs:81` |
**Description**
@@ -89,7 +89,19 @@ verifies it is delivered by a retry sweep.
**Resolution**
_Unresolved._
Resolved 2026-05-16. Delivery handlers for `StoreAndForwardCategory.ExternalSystem` and
`CachedDbWrite` are now registered at site startup in `AkkaHostedService`, after
`StoreAndForwardService.StartAsync()`. Each handler resolves its consumer in a fresh DI
scope and calls a new `DeliverBufferedAsync`: `ExternalSystemClient.DeliverBufferedAsync`
re-resolves the system/method and re-invokes `InvokeHttpAsync`, and
`DatabaseGateway.DeliverBufferedAsync` executes the buffered SQL — each returning `true`
on success, `false` (park) when the target no longer exists or fails permanently, and
throwing on transient failure so the engine retries. `EnqueueAsync` gained an
`attemptImmediateDelivery` parameter; `CachedCallAsync` passes `false` so registering the
handler does not dispatch the request twice (the double-dispatch noted in
`ExternalSystemGateway-003`). Regression tests cover the success, target-removed and
transient-retry paths. Fixed by the commit whose message references
`ExternalSystemGateway-001`.
### ExternalSystemGateway-002 — Per-system call timeout is never applied to HTTP requests