Files
OmniRoute/open-sse/services/combo/comboCooldownRetry.ts
ViFigueiredo 577bbf3e47 [defer] feat(combo): universal cooldown-aware retry & auto-strategy combo-ref guard (#7301)
* feat(combo): universal cooldown-aware retry & auto-strategy combo-ref guard

Two changes:

1. Universal cooldown-aware retry (combo.ts):
   - Remove strategy==="quota-share" gate from comboCooldownWaitEnabled
   - Enables all 18 combo strategies (priority, weighted, round-robin, etc.)
     to wait out a short transient cooldown and retry the full set
   - Non-quota-share strategies use shouldWaitForComboCooldown directly
     with earliestRetryAfter and reason="rate_limit" (no per-model lockout)
   - quota-share retains its existing per-target model lockout logic

2. Auto-strategy combo-ref guard (autoStrategy.ts):
   - expandAutoComboCandidatePool now detects kind==="combo-ref" entries
   - When present, returns eligibleTargets without expanding to ALL providers
   - Fixes scenario where an "auto" combo with combo-ref delegates to a
     sub-combo but pulls in every model from every active provider

* feat(combo): global comboTimeoutMs + aggregated error diagnostics

Adds two features to improve combo resilience and debuggability:

1. Global combo timeout (comboTimeoutMs)
   - Configurable per-combo via DEFAULT_COMBO_CONFIG (default 0 = disabled)
   - After each target completes, checks if total elapsed time exceeds limit
   - When exceeded, stops trying further targets and returns 504 immediately
   - Backward-compatible: 0 preserves legacy unlimited-iteration behavior

2. Aggregated error diagnostics
   - comboErrors array accumulates per-model failure details (model, status, error)
   - On combo timeout or all-models-exhausted, returns a message listing the
     first (up to 5) model-level errors with their HTTP status codes
   - Enables operators to see WHICH models failed and WHY without digging
     through individual server logs

* test(combo): cover universal cooldown-aware retry, comboTimeoutMs, and combo-ref guard

Adds/updates automated coverage for this PR's production changes (PR Test
Policy requires tests alongside src/open-sse/electron/bin changes):

- Update the "preserves the first failure status" expectation in
  combo-routing-engine.test.ts: the aggregated per-model error-diagnostics
  suffix is new intended output, not a regression.
- Add two new tests for the global comboTimeoutMs feature: the combo stops
  dispatching further targets and returns 504/COMBO_TIMEOUT with aggregated
  diagnostics once the ceiling trips, and comboTimeoutMs=0 (default) never
  trips it.
- Rewrite the non-quota-share (priority) cooldown-wait scenario in
  combo-quota-share-cooldown-wait-timing.test.ts: comboCooldownWait is no
  longer gated on strategy === "quota-share", so a priority combo now waits
  out a short 429 and re-dispatches too (via shouldWaitForComboCooldown with
  reason "rate_limit"), instead of propagating immediately as before. Also
  adds the disabled-flag counterpart for parity with the quota-share suite.
- Add coverage for the #COMBO-REF guard in expandAutoComboCandidatePool:
  a combo whose models array contains a kind:"combo-ref" entry must not be
  expanded to every model of every active provider.

Co-authored-by: diegosouzapw <diegosouza.pw@gmail.com>

* test(combo): keep the new comboTimeoutMs tests free of no-explicit-any

The two new tests initially copied the neighbouring tests' `any`-typed
handleSingleModel params / json() casts. Those neighbours are pre-existing
violations frozen in config/quality/eslint-suppressions.json at a count of 261
for this file, so the 7 new occurrences pushed it to 268 and broke `npm run
lint` (no-explicit-any is an error in tests/ since #6218; new violations must
be fixed, not re-frozen).

Type the new tests properly instead: `unknown`/`string` params, a
ComboErrorPayload interface for the parsed body, and drop the unused
`relayOptions: null as any` (the sibling combo cooldown suites already omit
it). Back to exactly 261 — the suppression file is untouched.

Co-authored-by: diegosouzapw <diegosouza.pw@gmail.com>

* fix(combo): gate the universal cooldown retry on the REAL lock reason, not a hardcoded "rate_limit"

The universal cooldown-aware retry kept the quota-share path on
resolveComboCooldownWaitDecision but gave every OTHER strategy a shortcut that
hardcoded `reason: "rate_limit"` and fed shouldWaitForComboCooldown the
earliestRetryAfter directly.

comboCooldownRetry.ts documents TWO deliberate barriers ("SECURITY —
quota_exhausted must be excluded"): (1) the reason allow-list, and (2) the
maxWaitMs ceiling, explicitly called the SECOND barrier. Hardcoding the reason
removed barrier 1 for 17 of the 18 strategies and left only the ceiling — which
does NOT cover a quota_exhausted lock whose wait lands under maxWaitMs. In that
case the combo waits, redispatches against a model that is locked until the
quota resets, and burns the retry budget for nothing.

The shortcut's premise ("non-quota-share combos have no per-connection model
lockout tracking") is also false: recordModelLockoutFailure in the target loop
is not gated on quota-share, so every strategy records model lockouts and the
real reason is always available.

Fix: one decision path for every strategy, always through
resolveComboCooldownWaitDecision, so the reason always crosses the allow-list.
The lock lookup is now keyed on each TARGET's own model (via a new third
`target` arg on lookupLock) — quota-share combos are single-model/multi-account
so this is identical to the previous orderedTargets[0] behavior, but
heterogeneous combos (priority, weighted, round-robin, …) carry a different
model per target and would otherwise miss every lock but the first.

Regression guard (tests/unit/serial/combo-quota-share-cooldown-wait-timing.test.ts):
a priority combo where modelLockout.errorCodes=[403] leaves the 403's
quota_exhausted lock as the only one in play while a 429 crystallizes the
status, and the resulting wait is short enough that the ceiling lets it through
— so only the allow-list can stop it. Verified failing-then-passing: with the
hardcoded reason the combo makes 6 dispatches (wait+redispatch x2); with the
real reason it makes 2 and propagates the 429.

Co-authored-by: diegosouzapw <diegosouza.pw@gmail.com>

* chore(quality): rebaseline file-size for PR #7301 own growth (combo +91, combo-routing-engine test +68)

---------

Co-authored-by: diegosouzapw <diegosouza.pw@gmail.com>
2026-07-21 16:30:37 -03:00

259 lines
9.8 KiB
TypeScript

/**
* comboCooldownRetry.ts — pure decision helper for the quota-share combo
* cooldown-aware retry (Variante A).
*
* Problem it solves
* -----------------
* In a single-connection quota-share combo (`qtSd/…`), when the upstream
* returns a SHORT transient 429 (reset in a few seconds), the combo loop
* treats the only target as locked-by-resilience, skips it, and crystallizes a
* 429 `model_cooldown` to the client immediately — even though waiting a couple
* of seconds would let the same connection succeed. This helper decides whether
* `handleComboChat` should WAIT for that short cooldown and re-dispatch instead
* of propagating the 429.
*
* Why a pure helper
* -----------------
* The actual wait (`waitForCooldownAwareRetry`) and the lock lookup
* (`getModelLockoutInfo`) live in `handleComboChat`; only the *gating policy*
* lives here. Keeping the policy pure (no I/O, no clock) means every branch is
* unit-testable deterministically and the orchestration in combo.ts stays thin.
*
* SECURITY — `quota_exhausted` must be excluded
* ---------------------------------------------
* `isRetryableModelLockoutReason` (src/sse/services/auth.ts:533) considers
* `quota_exhausted` RETRYABLE — its NON_RETRYABLE set only holds
* `not_found`/`not_found_local`. But `recordModelLockoutFailure`
* (accountFallback.ts) locks a quota-exhausted model UNTIL MIDNIGHT. If this
* helper trusted that classifier, a combo would wait until midnight. So this
* helper uses its OWN allow-list semantics: a reason qualifies for a wait only
* when it is a known short/transient reason AND is not in the explicit
* non-retryable set below. The small `maxWaitMs` ceiling is the second barrier.
*/
/**
* Reasons that must NEVER trigger a wait, regardless of the upstream-provided
* retry hint. `quota_exhausted` is the critical one (locked until midnight);
* the auth/not-found reasons can never be cured by waiting a few seconds.
*/
export const COMBO_COOLDOWN_NON_RETRYABLE_REASONS: ReadonlySet<string> = new Set([
"quota_exhausted",
"auth_error",
"not_found",
"not_found_local",
]);
/**
* Reasons recognised as short/transient and therefore eligible for a wait. We
* use an allow-list (rather than "anything not in the deny-list") so an unknown
* or empty reason fails closed — only an explicit transient reason qualifies.
* These match the model-lockout reasons produced by classifyLockoutReason /
* recordModelLockoutFailure for rate-limit-class failures.
*/
export const COMBO_COOLDOWN_RETRYABLE_REASONS: ReadonlySet<string> = new Set([
"rate_limit",
"rate_limited",
"transient",
"overloaded",
"server_error",
]);
export interface ComboCooldownWaitSettings {
/** Master switch — when false the helper always returns wait=false. */
enabled: boolean;
/** Hard ceiling (ms) on a single wait. Inclusive upper bound. */
maxWaitMs: number;
/** Maximum number of wait+redispatch attempts for one request. */
maxAttempts: number;
/** Total wait budget (ms) for the request — combo.ts decrements it. */
budgetMs: number;
}
export interface ShouldWaitForComboCooldownInput {
/** Lock reason from getModelLockoutInfo (may be null/unknown). */
reason: unknown;
/** Computed wait derived from the upstream retry-after hint (ms). */
waitMs: unknown;
/** Zero-based count of waits already performed for this request. */
attempt: number;
/** Remaining wait budget (ms) for this request. */
budgetLeftMs: number;
settings: ComboCooldownWaitSettings;
}
export interface ShouldWaitForComboCooldownResult {
wait: boolean;
/**
* The wait duration the caller should pass to waitForCooldownAwareRetry. It
* is surfaced even when `wait` is false (clamped to a finite >= 0 value) so
* the caller can log "would have waited Xms" without re-deriving it.
*/
waitMs: number;
}
function toFiniteWaitMs(value: unknown): number {
if (typeof value === "number" && Number.isFinite(value) && value > 0) {
return value;
}
return 0;
}
function isRetryableComboCooldownReason(reason: unknown): boolean {
if (typeof reason !== "string" || reason.length === 0) return false;
if (COMBO_COOLDOWN_NON_RETRYABLE_REASONS.has(reason)) return false;
return COMBO_COOLDOWN_RETRYABLE_REASONS.has(reason);
}
/**
* Decide whether a quota-share combo should wait for a short cooldown and
* re-dispatch. Pure — no I/O, no clock. All inputs are explicit.
*
* Returns wait=true only when ALL hold:
* - settings.enabled
* - reason is a recognised transient reason AND not in the non-retryable set
* (quota_exhausted / auth_error / not_found / not_found_local)
* - waitMs is finite and 0 < waitMs <= settings.maxWaitMs
* - attempt < settings.maxAttempts
* - budgetLeftMs >= waitMs
*/
export function shouldWaitForComboCooldown(
input: ShouldWaitForComboCooldownInput
): ShouldWaitForComboCooldownResult {
const { reason, attempt, budgetLeftMs, settings } = input;
const waitMs = toFiniteWaitMs(input.waitMs);
const wait =
settings.enabled === true &&
isRetryableComboCooldownReason(reason) &&
waitMs > 0 &&
waitMs <= settings.maxWaitMs &&
attempt < settings.maxAttempts &&
budgetLeftMs >= waitMs;
return { wait, waitMs };
}
/** Minimal lock shape resolved from getModelLockoutInfo. */
export interface ComboCooldownLockInfo {
reason: unknown;
remainingMs: number;
}
/** Minimal combo-target shape this helper inspects. */
export interface ComboCooldownTarget {
provider?: string | null;
connectionId?: string | null;
/**
* The target's own model. Heterogeneous combos (priority, weighted,
* round-robin, …) hold a DIFFERENT model per target, so the lock lookup must
* be keyed on each target's own model — keying every lookup on the first
* target's model would miss every other target's lock and silently degrade
* the reason allow-list to "no lock found".
*/
modelStr?: string | null;
}
export interface ResolveComboCooldownDecisionInput {
/** Combo targets to inspect for an active short lock. */
targets: ReadonlyArray<ComboCooldownTarget>;
/** Earliest retry-after the loop crystallized (string | number | Date | null). */
earliestRetryAfter: unknown;
attempt: number;
budgetLeftMs: number;
settings: ComboCooldownWaitSettings;
/**
* Per-target lock lookup (getModelLockoutInfo). Receives the target itself so
* the caller can key the lookup on that target's own model.
*/
lookupLock: (
provider: string,
connectionId: string,
target: ComboCooldownTarget
) => ComboCooldownLockInfo | null;
/** Derives the wait (ms) from the retry-after hint (computeClosestRetryAfter). */
computeWaitMs: (retryAfter: unknown) => number | null;
}
export interface ResolveComboCooldownDecisionResult extends ShouldWaitForComboCooldownResult {
/** Reason that drove the decision (for logging); null when none resolved. */
reason: string | null;
}
/**
* Small safety margin (ms) added on top of the wait so the model lock is
* reliably EXPIRED when the loop re-checks `isModelLocked` after the wait (real
* clocks drift a few ms between locking and re-evaluation; under-waiting would
* make the 2nd pass skip the target again and crystallize a 503/429).
*/
export const COMBO_COOLDOWN_WAIT_MARGIN_MS = 50;
/**
* Resolve, from the combo targets and the crystallized retry-after, whether the
* quota-share combo should wait and re-dispatch. Picks the locked target with
* the SMALLEST positive remaining cooldown (the soonest to recover) and uses its
* reason. The wait must last long enough to actually clear that lock, so it is
* `max(lock remainingMs, upstream retry-after hint) + margin` — the lock's own
* remaining time is authoritative (it already folds in a longer upstream hint
* via selectLockoutCooldownMs), and we honor an even longer hint if present. The
* ceiling check in shouldWaitForComboCooldown then gates on that real duration,
* so only genuinely SHORT cooldowns are waited. Pure given the injected
* `lookupLock`/`computeWaitMs` — no I/O, no clock — so it is unit-testable.
*/
export function resolveComboCooldownWaitDecision(
input: ResolveComboCooldownDecisionInput
): ResolveComboCooldownDecisionResult {
const {
targets,
earliestRetryAfter,
attempt,
budgetLeftMs,
settings,
lookupLock,
computeWaitMs,
} = input;
// Short-circuit before any lookup when the feature is off.
if (!settings.enabled) return { wait: false, waitMs: 0, reason: null };
let best: { reason: unknown; remainingMs: number } | null = null;
for (const target of targets) {
const provider = typeof target.provider === "string" ? target.provider : "";
if (!provider) continue;
const connectionId = typeof target.connectionId === "string" ? target.connectionId : "";
const info = lookupLock(provider, connectionId, target);
if (!info) continue;
const remainingMs =
typeof info.remainingMs === "number" && Number.isFinite(info.remainingMs)
? info.remainingMs
: 0;
if (remainingMs <= 0) continue;
if (!best || remainingMs < best.remainingMs) {
best = { reason: info.reason, remainingMs };
}
}
if (!best) return { wait: false, waitMs: 0, reason: null };
// Wait long enough to actually clear the lock: the larger of the lock's own
// remaining time and the upstream retry-after hint, plus a small margin.
const rawHintWaitMs = computeWaitMs(earliestRetryAfter);
const hintWaitMs =
typeof rawHintWaitMs === "number" && Number.isFinite(rawHintWaitMs) && rawHintWaitMs > 0
? rawHintWaitMs
: 0;
const waitMs = Math.max(best.remainingMs, hintWaitMs) + COMBO_COOLDOWN_WAIT_MARGIN_MS;
const decision = shouldWaitForComboCooldown({
reason: best.reason,
waitMs,
attempt,
budgetLeftMs,
settings,
});
return {
...decision,
reason: typeof best.reason === "string" ? best.reason : null,
};
}