fix(outbound): preserve retries for budget-deferred deliveries (#91241)

849261680 · clawsweeper[bot] · Takhoffman · web-flow · commit 75c1790b504c · 2026-06-08T01:43:02.000Z
Summary: - The branch removes the budget-deferred `failDelivery` path from outbound recovery and updates the `maxRecoveryMs` regression expectation so unattempted deliveries keep retry counts at zero. - PR surface: Source -11, Tests -1. Total -12 across 2 files. - Reproducibility: yes. at source level: current main reaches `failDelivery` from the exhausted recovery-budge ... in this read-only review, but the PR body also supplies terminal output showing the after-fix queue state. Automerge notes: - PR branch already contained follow-up commit before automerge: fix(outbound): preserve retries for budget-deferred deliveries Validation: - ClawSweeper review passed for head aff2b9d. - Required merge gates passed before the squash merge. Prepared head SHA: aff2b9d Review: #91241 (comment) Co-authored-by: 宇宙熊Yzx <53250620+849261680@users.noreply.github.com> Co-authored-by: clawsweeper <274271284+clawsweeper[bot]@users.noreply.github.com> Co-authored-by: clawsweeper[bot] <274271284+clawsweeper[bot]@users.noreply.github.com> Approved-by: takhoffman Co-authored-by: takhoffman <781889+takhoffman@users.noreply.github.com>
diff --git a/src/infra/outbound/delivery-queue-recovery.ts b/src/infra/outbound/delivery-queue-recovery.ts
@@ -326,17 +326,6 @@ async function moveEntryToFailedWithLogging(
   }
 }
 
-async function deferRemainingEntriesForBudget(
-  entries: readonly QueuedDelivery[],
-  stateDir: string | undefined,
-): Promise<void> {
-  // Increment retryCount so entries that are repeatedly deferred by the
-  // recovery budget eventually hit MAX_RETRIES and get pruned.
-  await Promise.allSettled(
-    entries.map((entry) => failDelivery(entry.id, "recovery time budget exceeded", stateDir)),
-  );
-}
-
 /** Compute the backoff delay in ms for a given retry count. */
 export function computeBackoffMs(retryCount: number): number {
   if (retryCount <= 0) {
@@ -624,12 +613,12 @@ export async function recoverPendingDeliveries(opts: {
   const deadline = resolveRecoveryDeadlineMs(opts.maxRecoveryMs);
   const summary = createEmptyRecoverySummary();
 
-  for (let i = 0; i < pending.length; i++) {
-    const entry = pending[i];
+  for (const entry of pending) {
     const now = Date.now();
     if (now >= deadline) {
       opts.log.warn(`Recovery time budget exceeded — remaining entries deferred to next startup`);
-      await deferRemainingEntriesForBudget(pending.slice(i), opts.stateDir);
+      // Budget deferral is not a delivery attempt. Keep entries pending without
+      // consuming retry budget; attempted failures still flow through failDelivery.
       break;
     }
 
diff --git a/src/infra/outbound/delivery-queue.recovery.test.ts b/src/infra/outbound/delivery-queue.recovery.test.ts
@@ -615,7 +615,7 @@ describe("delivery-queue recovery", () => {
     });
   });
 
-  it("respects maxRecoveryMs time budget and bumps deferred retries", async () => {
+  it("respects maxRecoveryMs time budget without bumping deferred retries", async () => {
     await enqueueCrashRecoveryEntries();
     await enqueueDelivery(
       { channel: "demo-channel-c", to: "#c", payloads: [{ text: "c" }] },
@@ -638,8 +638,7 @@ describe("delivery-queue recovery", () => {
 
     const remaining = await loadPendingDeliveries(tmpDir());
     expect(remaining).toHaveLength(3);
-    const entriesWithUnexpectedRetryCount = remaining.filter((entry) => entry.retryCount !== 1);
-    expect(entriesWithUnexpectedRetryCount).toStrictEqual([]);
+    expect(remaining.map((entry) => entry.retryCount)).toStrictEqual([0, 0, 0]);
     expectMockMessageContaining(log.warn, "deferred to next startup");
   });
 
