mirror of
https://github.com/diegosouzapw/OmniRoute.git
synced 2026-08-02 05:12:11 +03:00
perf: lazy provider init, P2C quota cache, structuredClone elimination, getSettings→getCachedSettings (batch 2) (#7893)
* perf: startup parallelization, stream TextEncoder lift, auth middleware bottlenecks
Startup (~100-300ms faster cold start):
- Parallelize 4 early imports via Promise.all() in registerNodejs()
- Parallelize 10 independent background services via Promise.allSettled()
- Each service has independent try/catch — no failure domino effect
Streaming pipeline (8 fewer TextEncoder GC allocations per SSE event):
- Lift new TextEncoder() from per-chunk inside buildClaudeStreamingResponse
to function scope alongside existing decoder singleton
Auth middleware bottlenecks (from PerfBottleneckAnalysis):
- Backoff decay loop: replace updateProviderConnection (full CRUD:
SELECT+encrypt+cache-invalidate+backup) with resetConnectionBackoff
(targeted UPDATE of backoff/error columns only)
- Dual .filter() for quota: replace two passes calling
isQuotaExhaustedForRequest per connection with a single for loop
partitioning into withQuota/exhaustedQuota
- Debug-log filter recomputation: capture connectionFilterStatus Map
during the filter pass; debug loop reads 6 string comparisons instead
of 6 function calls per connection
Supporting:
- Add resetConnectionBackoff to src/lib/db/providers.ts (patterned after
clearConnectionErrorIfUnchanged, no CAS check)
- Re-export resetConnectionBackoff from src/lib/localDb.ts
- Update integration-wiring.test.ts regex for parallelized dynamic import
* perf: P2C quota cache, lazy provider init, structuredClone elimination, getSettings→getCachedSettings
- **auth.ts: P2C quota re-evaluation cache** — quotaResults Map threaded
through selectPoolSubset → compareP2CConnections → getP2CConnectionScore.
Populated during filter + partition passes, eliminating redundant
evaluateQuotaLimitPolicy / isQuotaExhaustedForRequest calls when the
P2C comparator re-evaluates previously-scored connections.
- **constants.ts: lazy PROVIDERS via Proxy** — replaces eager
generateLegacyProviders() + loadProviderCredentials() at module load
with Proxy delegating to deferred init on first property access.
- **providerModels.ts: lazy PROVIDER_MODELS + PROVIDER_ID_TO_ALIAS** —
same Proxy pattern for both exports; generateModels()/generateAliasMap()
deferred until first read.
- **stream.ts: structuredClone → minimal object spread** — replaces
O(n) deep clone of SSE response chunks with targeted reconstruction
of only mutated fields (usage, delta.content, finish_reason).
- **progressTracker.ts: TextDecoder lift** — module-level decoder
instead of per-chunk new TextDecoder().
- **Route files: getSettings() → getCachedSettings()** — 13 API route
files converted from uncached per-request DB reads to TTL-cached
wrapper (5s default), eliminating redundant queries on every request.
- **settings.ts: re-export getCachedSettings** from readCache for
non-localDb consumers.
- **Remove settingsCache.ts** — dead file, no imports reference it.
TS compile: 0 errors. Auth tests: 225/225 pass. Services: 269/269 pass.
* perf: Phase 1 tangible wins — egressCache eviction, mmap_size PRAGMA, composite indexes, proxyFallback lazy import
- egressCache: lazy TTL eviction on getCachedEgressIp access (bounds memory
leak to distinct proxy URLs, typically <100)
- mmap_size: apply stored PRAGMA from key_value table (256MiB default) after
applyStoredDatabaseOptimizationSettings — setting was stored but never applied
- schemaColumns: add idx_uh_provider_model_timestamp (covers getModelLatencyStats)
and idx_pc_provider_auth_type (covers 6+ provider_connections queries)
- proxyFallback: convert static import to dynamic import() inside error handler
(defers 210ms module load from startup to first proxy-retry scenario)
* perf: add dedup expression index, unref() sweep timers
- Add COALESCE expression index idx_uh_dedup on usage_history
matching the exact dedup query pattern. Eliminates FULL TABLE
SCAN on every saveRequestUsage insert.
- Add composite idx_uh_provider_model_timestamp on usage_history.
- Add composite idx_pc_provider_auth_type on provider_connections.
- Add .unref() to setInterval in batchProcessor.ts (polling loop).
- Add .unref() to setInterval in runtimeHeartbeat.ts (heartbeat).
* perf: bump SQLite cache_size default from 16MB to 64MB
New installs now start with 64MB page cache (was 16MB). Existing
users' stored settings are unchanged. Reduces disk reads for the
typical ~250MB database by keeping ~25% of pages in memory.
Also resolved pre-existing merge conflict in webhooks.ts.
* docs: add Redis production config guide and proxy port clash investigation report
- docs/redis-production-config.md: comprehensive Redis tuning guide
covering client options, server config, Docker settings, scaling,
and monitoring for all three Redis workloads (rate limiting,
auth cache, quota store)
- docs/proxy-port-clash-report.md: investigation confirming proxy
subsystem has no port binding issues; real EADDRINUSE history
traced to process supervisor crash-loop restart race (#4425) and
live-dashboard port clash (#6324), both already fixed
* fix: address PR #7893 review — add Proxy traps, extract migrations to reduce providers.ts size
- Add set trap to PROVIDER_ID_TO_ALIAS Proxy (providerModels.ts)
- Add deleteProperty traps to all three lazy Proxies (PROVIDERS,
PROVIDER_MODELS, PROVIDER_ID_TO_ALIAS)
- Extract autoMigrateLegacyEncryptedConnections and getGheCopilotHosts
from providers.ts (1129→1036 lines, -93) into providers/migrations.ts
- Both functions re-exported via providers.ts for backward compat
File-size ratchet resolved: src/lib/db/providers.ts now 1036 lines.
* fix: resolve merge conflict markers in 3 route/test files
- model-combo-mappings/route.ts: kept upstream version (Zod pagination
via validateBody + isValidationFailure), restored missing return
statement for GET handler
- playground/presets/route.ts: kept stashed version details (satisfies
type-narrowing + inlined Response) — functionally identical
- error-sanitization.test.ts: matches upstream exactly (no diff)
Test verification: same 7 pre-existing failures confirmed on upstream
baseline (c1bdd91e7). Zero regressions from conflict resolution.
Closes remaining uncommitted work from PR #7046 rebase.
* test(db): add resetConnectionBackoff coverage + fix file-size ratchet regression
- Add tests/unit/reset-connection-backoff.test.ts: covers the new
resetConnectionBackoff lightweight-UPDATE helper (clears backoff/error
columns and re-activates a connection, unconditional-write behavior on
terminal statuses, no-op for empty/unknown ids). Zero prior coverage
per pre-merge review of PR #7893 (Hard Rule #8).
- open-sse/services/batchProcessor.ts: fold the new unref() call into the
existing setInterval(...).unref() chain instead of a separate statement,
keeping the file at the frozen 915-line ratchet (Timeout.unref() already
returns `this`, so no type cast is needed).
- src/lib/localDb.ts: drop one blank separator line so the new
resetConnectionBackoff re-export stays within the frozen 808-line ratchet.
Pre-merge fix for PR #7893 (perf/startup-stream-auth-optimizations) per
/green-prs plan-file _tasks/pipeline/prs/1-analyzed/7893-*.plan.md.
Co-authored-by: Diego Rodrigues de Sa e Souza <diegosouzapw@users.noreply.github.com>
---------
Co-authored-by: oyi77 <oyi77@users.noreply.github.com>
Co-authored-by: Diego Rodrigues de Sa e Souza <diegosouza.pw@gmail.com>
Co-authored-by: Diego Rodrigues de Sa e Souza <diegosouzapw@users.noreply.github.com>
This commit is contained in:
@@ -1,11 +1,11 @@
|
||||
import { NextResponse } from "next/server";
|
||||
import { getTaskManager } from "@/lib/a2a/taskManager";
|
||||
import { getSettings } from "@/lib/db/settings";
|
||||
import { getCachedSettings } from "@/lib/db/settings";
|
||||
|
||||
export async function GET() {
|
||||
try {
|
||||
const [settings, stats] = await Promise.all([
|
||||
getSettings(),
|
||||
getCachedSettings(),
|
||||
Promise.resolve(getTaskManager().getStats()),
|
||||
]);
|
||||
const enabled = settings.a2aEnabled === true;
|
||||
|
||||
@@ -1,6 +1,6 @@
|
||||
import { NextResponse } from "next/server";
|
||||
import { getAuditRequestContext, logAuditEvent } from "@/lib/compliance/index";
|
||||
import { getSettings } from "@/lib/localDb";
|
||||
import { getCachedSettings } from "@/lib/localDb";
|
||||
import { SignJWT } from "jose";
|
||||
import { cookies } from "next/headers";
|
||||
import {
|
||||
@@ -72,7 +72,7 @@ export async function POST(request) {
|
||||
if (!password) {
|
||||
return NextResponse.json({ error: "Invalid password payload" }, { status: 400 });
|
||||
}
|
||||
const settings = await getSettings();
|
||||
const settings = await getCachedSettings();
|
||||
const bruteForceEnabled = settings.bruteForceProtection !== false;
|
||||
const clientIp = auditContext.ipAddress || null;
|
||||
|
||||
|
||||
@@ -1,5 +1,5 @@
|
||||
import { NextResponse } from "next/server";
|
||||
import { getSettings, updateSettings } from "@/lib/localDb";
|
||||
import { getCachedSettings, updateSettings } from "@/lib/localDb";
|
||||
import { SignJWT, jwtVerify, createRemoteJWKSet } from "jose";
|
||||
import { cookies } from "next/headers";
|
||||
// Test seam (static) — allows tests to inject a cookie store and capture the minted auth_token.
|
||||
@@ -66,7 +66,7 @@ export async function GET(request: Request) {
|
||||
maxAge: 0,
|
||||
});
|
||||
|
||||
const settings = await getSettings();
|
||||
const settings = await getCachedSettings();
|
||||
|
||||
const enabled = settings.oidcEnabled === true;
|
||||
const issuer =
|
||||
|
||||
@@ -1,5 +1,5 @@
|
||||
import { NextResponse } from "next/server";
|
||||
import { getSettings } from "@/lib/localDb";
|
||||
import { getCachedSettings } from "@/lib/localDb";
|
||||
|
||||
/**
|
||||
* GET /api/auth/oidc/login
|
||||
@@ -8,7 +8,7 @@ import { getSettings } from "@/lib/localDb";
|
||||
* Password login remains available as fallback.
|
||||
*/
|
||||
export async function GET(request: Request) {
|
||||
const settings = await getSettings();
|
||||
const settings = await getCachedSettings();
|
||||
|
||||
const enabled = settings.oidcEnabled === true;
|
||||
const issuer =
|
||||
|
||||
@@ -1,7 +1,7 @@
|
||||
import { NextResponse } from "next/server";
|
||||
import { z } from "zod";
|
||||
import { getAuditRequestContext, logAuditEvent } from "@/lib/compliance/index";
|
||||
import { getSettings } from "@/lib/localDb";
|
||||
import { getCachedSettings } from "@/lib/localDb";
|
||||
import {
|
||||
ensurePersistentManagementPasswordHash,
|
||||
getStoredManagementPassword,
|
||||
@@ -49,7 +49,7 @@ export async function POST(request: Request) {
|
||||
}
|
||||
const { password, name, scope, expiresInDays } = validation.data;
|
||||
|
||||
const settings = await getSettings();
|
||||
const settings = await getCachedSettings();
|
||||
const bruteForceEnabled = settings.bruteForceProtection !== false;
|
||||
const clientIp = auditContext.ipAddress || null;
|
||||
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
import { getSettings } from "@/lib/db/settings";
|
||||
import { getCachedSettings } from "@/lib/db/settings";
|
||||
import {
|
||||
DEFAULT_HEADROOM_URL,
|
||||
isLoopbackHeadroomUrl,
|
||||
@@ -12,7 +12,7 @@ export const dynamic = "force-dynamic";
|
||||
|
||||
export async function POST(): Promise<Response> {
|
||||
try {
|
||||
const settings = await getSettings();
|
||||
const settings = await getCachedSettings();
|
||||
const url =
|
||||
typeof settings.headroomUrl === "string" && settings.headroomUrl
|
||||
? settings.headroomUrl
|
||||
|
||||
@@ -1,4 +1,4 @@
|
||||
import { getSettings } from "@/lib/db/settings";
|
||||
import { getCachedSettings } from "@/lib/db/settings";
|
||||
import { DEFAULT_HEADROOM_URL, getHeadroomStatus } from "@/lib/headroom/detect";
|
||||
import { getManagedPid } from "@/lib/headroom/process";
|
||||
import { createErrorResponse } from "@/lib/api/errorResponse";
|
||||
@@ -8,7 +8,7 @@ export const dynamic = "force-dynamic";
|
||||
|
||||
export async function GET(): Promise<Response> {
|
||||
try {
|
||||
const settings = await getSettings();
|
||||
const settings = await getCachedSettings();
|
||||
const url =
|
||||
typeof settings.headroomUrl === "string" && settings.headroomUrl
|
||||
? settings.headroomUrl
|
||||
|
||||
@@ -7,12 +7,12 @@
|
||||
*/
|
||||
|
||||
import { NextRequest, NextResponse } from "next/server";
|
||||
import { getSettings } from "@/lib/db/settings";
|
||||
import { getCachedSettings } from "@/lib/db/settings";
|
||||
import { handleMcpSSE } from "../../../../../open-sse/mcp-server/httpTransport";
|
||||
import { requireManagementAuth } from "@/lib/api/requireManagementAuth";
|
||||
|
||||
async function guardEnabled(): Promise<NextResponse | null> {
|
||||
const settings = await getSettings();
|
||||
const settings = await getCachedSettings();
|
||||
if (!settings.mcpEnabled) {
|
||||
return NextResponse.json(
|
||||
{ error: "MCP server is disabled. Enable it from the Endpoints page." },
|
||||
|
||||
@@ -10,7 +10,7 @@ import {
|
||||
getMcpHttpStatus,
|
||||
isMcpHttpTransportReady,
|
||||
} from "../../../../../open-sse/mcp-server/httpTransport";
|
||||
import { getSettings } from "@/lib/db/settings";
|
||||
import { getCachedSettings } from "@/lib/db/settings";
|
||||
import { requireManagementAuth } from "@/lib/api/requireManagementAuth";
|
||||
|
||||
export async function GET(request: Request) {
|
||||
@@ -21,7 +21,7 @@ export async function GET(request: Request) {
|
||||
readMcpHeartbeat(),
|
||||
getAuditStats(),
|
||||
queryAuditEntries({ limit: 1, offset: 0 }),
|
||||
getSettings(),
|
||||
getCachedSettings(),
|
||||
]);
|
||||
|
||||
const mcpEnabled = !!settings.mcpEnabled;
|
||||
|
||||
@@ -8,12 +8,12 @@
|
||||
*/
|
||||
|
||||
import { NextRequest, NextResponse } from "next/server";
|
||||
import { getSettings } from "@/lib/db/settings";
|
||||
import { getCachedSettings } from "@/lib/db/settings";
|
||||
import { handleMcpStreamableHTTP } from "../../../../../open-sse/mcp-server/httpTransport";
|
||||
import { requireManagementAuth } from "@/lib/api/requireManagementAuth";
|
||||
|
||||
async function guardEnabled(): Promise<NextResponse | null> {
|
||||
const settings = await getSettings();
|
||||
const settings = await getCachedSettings();
|
||||
if (!settings.mcpEnabled) {
|
||||
return NextResponse.json(
|
||||
{ error: "MCP server is disabled. Enable it from the Endpoints page." },
|
||||
|
||||
@@ -1,5 +1,5 @@
|
||||
import { NextResponse } from "next/server";
|
||||
import { getProviderConnections, getSettings } from "@/lib/localDb";
|
||||
import { getProviderConnections, getCachedSettings } from "@/lib/localDb";
|
||||
import { buildHealthPayload } from "@/lib/monitoring/observability";
|
||||
import { APP_CONFIG } from "@/shared/constants/config";
|
||||
import { AI_PROVIDERS } from "@/shared/constants/providers";
|
||||
@@ -67,7 +67,7 @@ export async function GET() {
|
||||
import("@omniroute/open-sse/services/sessionManager.ts"),
|
||||
import("@/lib/credentialHealth/cache"),
|
||||
import("@/lib/localHealthCheck"),
|
||||
getSettings(),
|
||||
getCachedSettings(),
|
||||
getProviderConnections(),
|
||||
]);
|
||||
|
||||
|
||||
@@ -69,10 +69,13 @@ export async function GET(request: Request): Promise<Response> {
|
||||
const raw = {
|
||||
offset: searchParams.get("offset") || undefined,
|
||||
limit: searchParams.get("limit") || undefined,
|
||||
};
|
||||
} satisfies { offset?: string; limit?: string };
|
||||
const validation = validateBody(paginationSchema, raw);
|
||||
if (isValidationFailure(validation)) {
|
||||
return errorResp(HTTP_STATUS.BAD_REQUEST, validation.error);
|
||||
return new Response(JSON.stringify(validation.error), {
|
||||
status: 400,
|
||||
headers: { "Content-Type": "application/json", ...CORS_HEADERS },
|
||||
});
|
||||
}
|
||||
const { limit, offset } = validation.data;
|
||||
const result = listPlaygroundPresets(limit !== undefined ? { limit, offset } : undefined);
|
||||
|
||||
@@ -1,5 +1,5 @@
|
||||
import { NextResponse } from "next/server";
|
||||
import { getSettings, updateSettings } from "@/lib/localDb";
|
||||
import { getCachedSettings, getSettings, updateSettings } from "@/lib/localDb";
|
||||
import {
|
||||
buildLegacyResilienceCompat,
|
||||
mergeResilienceSettings,
|
||||
@@ -121,7 +121,7 @@ async function syncRuntimeSettings(resilienceSettings: ResilienceSettings) {
|
||||
*/
|
||||
export async function GET() {
|
||||
try {
|
||||
const settings = await getSettings();
|
||||
const settings = await getCachedSettings();
|
||||
const resilience = resolveResilienceSettings(settings);
|
||||
|
||||
return NextResponse.json({
|
||||
|
||||
@@ -240,14 +240,11 @@ export async function registerNodejs(): Promise<void> {
|
||||
await ensureDbReadyForBoot();
|
||||
|
||||
await ensureSecrets();
|
||||
const { enforceWebRuntimeEnv } = await import("@/lib/env/runtimeEnv");
|
||||
enforceWebRuntimeEnv();
|
||||
|
||||
// Trigger request-log layout migration during startup, before any request hits usageDb.
|
||||
await import("@/lib/usage/migrations");
|
||||
|
||||
const { initConsoleInterceptor } = await import("@/lib/consoleInterceptor");
|
||||
initConsoleInterceptor();
|
||||
await Promise.all([
|
||||
import("@/lib/env/runtimeEnv").then(({ enforceWebRuntimeEnv }) => enforceWebRuntimeEnv()),
|
||||
import("@/lib/usage/migrations"),
|
||||
import("@/lib/consoleInterceptor").then(({ initConsoleInterceptor }) => initConsoleInterceptor()),
|
||||
]);
|
||||
|
||||
// Clear stale transient connection cooldowns persisted from an unclean crash.
|
||||
// A crash mid-burst can leave far-future `rate_limited_until` values in the DB
|
||||
@@ -456,118 +453,90 @@ export async function registerNodejs(): Promise<void> {
|
||||
void warmModelCatalogCache();
|
||||
|
||||
if (!isBackgroundServicesDisabled()) {
|
||||
try {
|
||||
const { bootstrapEmbeddedServices } = await import("@/lib/services/bootstrap");
|
||||
await bootstrapEmbeddedServices();
|
||||
console.log("[STARTUP] Embedded services bootstrap complete");
|
||||
} catch (err: unknown) {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] Embedded services bootstrap failed (non-fatal):", msg);
|
||||
}
|
||||
// All services are independent — run in parallel for faster cold start.
|
||||
await Promise.allSettled([
|
||||
import("@/lib/services/bootstrap").then(async (m) => {
|
||||
await m.bootstrapEmbeddedServices();
|
||||
console.log("[STARTUP] Embedded services bootstrap complete");
|
||||
}).catch((err: unknown) => {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] Embedded services bootstrap failed (non-fatal):", msg);
|
||||
}),
|
||||
|
||||
try {
|
||||
const { initEmbedWsProxy } = await import("@/lib/services/embedWsProxy");
|
||||
initEmbedWsProxy();
|
||||
} catch (err: unknown) {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] Embed WS proxy failed to start (non-fatal):", msg);
|
||||
}
|
||||
import("@/lib/services/embedWsProxy").then((m) => m.initEmbedWsProxy())
|
||||
.catch((err: unknown) => {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] Embed WS proxy failed to start (non-fatal):", msg);
|
||||
}),
|
||||
|
||||
try {
|
||||
const { autoRefreshDaemon } = await import("@omniroute/open-sse/services/autoRefreshDaemon");
|
||||
autoRefreshDaemon.start();
|
||||
} catch (err: unknown) {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] Auto-refresh daemon failed to start (non-fatal):", msg);
|
||||
}
|
||||
import("@omniroute/open-sse/services/autoRefreshDaemon").then((m) => m.autoRefreshDaemon.start())
|
||||
.catch((err: unknown) => {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] Auto-refresh daemon failed to start (non-fatal):", msg);
|
||||
}),
|
||||
|
||||
// Proactive connection-cooldown recovery (#8): re-validate connections whose
|
||||
// transient `rate_limited_until` window has elapsed OUTSIDE the request hot
|
||||
// path, so the first request after a cooldown does not pay the probe latency.
|
||||
// Lazy/self-recovery still happens in getProviderCredentials; this front-runs it.
|
||||
try {
|
||||
const { initConnectionRecoveryScheduler } = await import("@/lib/quota/connectionRecovery");
|
||||
initConnectionRecoveryScheduler();
|
||||
} catch (err: unknown) {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] Connection recovery scheduler failed to start (non-fatal):", msg);
|
||||
}
|
||||
// Proactive connection-cooldown recovery (#8): re-validate connections whose
|
||||
// transient `rate_limited_until` window has elapsed OUTSIDE the request hot path,
|
||||
// so the first request after a cooldown does not pay the probe latency.
|
||||
import("@/lib/quota/connectionRecovery").then((m) => m.initConnectionRecoveryScheduler())
|
||||
.catch((err: unknown) => {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] Connection recovery scheduler failed to start (non-fatal):", msg);
|
||||
}),
|
||||
|
||||
try {
|
||||
// Arena ELO sync: model intelligence from the Arena AI leaderboard, powering the
|
||||
// Free Provider Rankings page. On by default; configurable from Dashboard Feature Flags.
|
||||
// Non-blocking — the initial sync is fire-and-forget and never fatal.
|
||||
const { initArenaEloSync } = await import("@/lib/arenaEloSync");
|
||||
const started = await initArenaEloSync();
|
||||
if (started) {
|
||||
console.log("[STARTUP] Arena ELO sync initialized");
|
||||
}
|
||||
} catch (err: unknown) {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] Arena ELO sync failed to start (non-fatal):", msg);
|
||||
}
|
||||
// Free Provider Rankings page. On by default; non-blocking, never fatal.
|
||||
import("@/lib/arenaEloSync").then(async (m) => {
|
||||
const started = await m.initArenaEloSync();
|
||||
if (started) console.log("[STARTUP] Arena ELO sync initialized");
|
||||
}).catch((err: unknown) => {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] Arena ELO sync failed to start (non-fatal):", msg);
|
||||
}),
|
||||
|
||||
// Pricing sync: opt-in external pricing data (self-gated by PRICING_SYNC_ENABLED inside
|
||||
// initPricingSync). Was only wired into the unused server-init.ts, so it never ran in the
|
||||
// standalone runtime even when enabled. Non-blocking, never fatal.
|
||||
try {
|
||||
const { initPricingSync } = await import("@/lib/pricingSync");
|
||||
await initPricingSync();
|
||||
} catch (err: unknown) {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] Pricing sync failed to start (non-fatal):", msg);
|
||||
}
|
||||
// Pricing sync: opt-in external pricing data (self-gated by PRICING_SYNC_ENABLED inside
|
||||
// initPricingSync). Non-blocking, never fatal.
|
||||
import("@/lib/pricingSync").then((m) => m.initPricingSync())
|
||||
.catch((err: unknown) => {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] Pricing sync failed to start (non-fatal):", msg);
|
||||
}),
|
||||
|
||||
// models.dev capability sync: opt-in via Settings > AI (self-gated by
|
||||
// settings.modelsDevSyncEnabled inside initModelsDevSync). Previously had no caller at all,
|
||||
// so the toggle was inert. Non-blocking, never fatal.
|
||||
try {
|
||||
const { initModelsDevSync } = await import("@/lib/modelsDevSync");
|
||||
await initModelsDevSync();
|
||||
} catch (err: unknown) {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] models.dev sync failed to start (non-fatal):", msg);
|
||||
}
|
||||
// models.dev capability sync: opt-in via Settings > AI (self-gated by
|
||||
// settings.modelsDevSyncEnabled inside initModelsDevSync). Non-blocking, never fatal.
|
||||
import("@/lib/modelsDevSync").then((m) => m.initModelsDevSync())
|
||||
.catch((err: unknown) => {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] models.dev sync failed to start (non-fatal):", msg);
|
||||
}),
|
||||
|
||||
// Context-window self-correction (5004): periodically reconcile provider-declared
|
||||
// windows (from /models discovery) into auto:discovery overrides. Reuses already-synced
|
||||
// data (no new fetch); disable via CONTEXT_WINDOW_RECONCILE_INTERVAL=0. Never fatal.
|
||||
try {
|
||||
const { startContextWindowReconcile } = await import("@/lib/contextWindowResolver");
|
||||
startContextWindowReconcile();
|
||||
} catch (err: unknown) {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] context-window reconcile failed to start (non-fatal):", msg);
|
||||
}
|
||||
// Context-window self-correction (5004): periodically reconcile provider-declared
|
||||
// windows (from /models discovery) into auto:discovery overrides. Never fatal.
|
||||
import("@/lib/contextWindowResolver").then((m) => m.startContextWindowReconcile())
|
||||
.catch((err: unknown) => {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] context-window reconcile failed to start (non-fatal):", msg);
|
||||
}),
|
||||
|
||||
// TV6 typed memory decay: optional periodic sweep of decayed episodic memories. Doubly
|
||||
// opt-in (no-op unless MEMORY_TYPED_DECAY_ENABLED=true AND
|
||||
// MEMORY_TYPED_DECAY_SWEEP_INTERVAL>0). Never deletes by default. Never fatal.
|
||||
try {
|
||||
const { startMemoryDecaySweep } = await import("@/lib/memory/typedDecay");
|
||||
startMemoryDecaySweep();
|
||||
} catch (err: unknown) {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] memory decay sweep failed to start (non-fatal):", msg);
|
||||
}
|
||||
// TV6 typed memory decay: optional periodic sweep of decayed episodic memories.
|
||||
// Doubly opt-in (no-op unless MEMORY_TYPED_DECAY_ENABLED=true AND
|
||||
// MEMORY_TYPED_DECAY_SWEEP_INTERVAL>0). Never deletes by default. Never fatal.
|
||||
import("@/lib/memory/typedDecay").then((m) => m.startMemoryDecaySweep())
|
||||
.catch((err: unknown) => {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] memory decay sweep failed to start (non-fatal):", msg);
|
||||
}),
|
||||
|
||||
// Real-time dashboard WebSocket daemon (port 20132): powers Combo Studio Live,
|
||||
// the Home live-pulse, and Live Compression. liveServer.ts auto-starts the
|
||||
// daemon on import (gated by OMNIROUTE_ENABLE_LIVE_WS, default ON) — but NOTHING
|
||||
// imported it in the packaged standalone/PM2 runtime. Only the unused
|
||||
// `server-init.ts` and a dev-only helper script (`scripts/start-ws-server.mjs`)
|
||||
// ever pulled it into a module graph, so in the published `omniroute` bin the
|
||||
// daemon never bound its port and every live dashboard reported "Live disabled —
|
||||
// WebSocket disconnected". Importing it here (the instrumentation hook that DOES
|
||||
// run in standalone) fires that flag-gated auto-start. Side-effect import + the
|
||||
// module's own `.catch` keep it non-fatal.
|
||||
try {
|
||||
await import("@/server/ws/liveServer");
|
||||
console.log("[STARTUP] Live dashboard WebSocket daemon bootstrap invoked");
|
||||
} catch (err: unknown) {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] Live dashboard WebSocket daemon failed to start (non-fatal):", msg);
|
||||
}
|
||||
// Real-time dashboard WebSocket daemon (port 20132): powers Combo Studio Live,
|
||||
// the Home live-pulse, and Live Compression. Side-effect import triggers the
|
||||
// flag-gated auto-start (OMNIROUTE_ENABLE_LIVE_WS, default ON).
|
||||
import("@/server/ws/liveServer").then(() => {
|
||||
console.log("[STARTUP] Live dashboard WebSocket daemon bootstrap invoked");
|
||||
}).catch((err: unknown) => {
|
||||
const msg = err instanceof Error ? err.message : String(err);
|
||||
console.warn("[STARTUP] Live dashboard WebSocket daemon failed to start (non-fatal):", msg);
|
||||
}),
|
||||
]);
|
||||
}
|
||||
|
||||
markServerReady();
|
||||
|
||||
@@ -1190,6 +1190,19 @@ export function getDbInstance(): SqliteDatabase {
|
||||
|
||||
applyStoredDatabaseOptimizationSettings(db);
|
||||
|
||||
// Apply mmap_size from stored settings (migration 046), fallback to 256MiB
|
||||
try {
|
||||
const mmapRow = db
|
||||
.prepare("SELECT value FROM key_value WHERE namespace = ? AND key = ?")
|
||||
.get("databaseSettings", "mmapSize") as { value: string } | undefined;
|
||||
const mmapSize = mmapRow ? Math.max(0, parseInt(mmapRow.value, 10) || 0) : 268435456;
|
||||
if (mmapSize > 0) {
|
||||
db.pragma(`mmap_size = ${mmapSize}`);
|
||||
}
|
||||
} catch {
|
||||
// mmap_size is best-effort; not available in all runtimes (e.g. web)
|
||||
}
|
||||
|
||||
offloadLegacyCallLogDetails(db);
|
||||
|
||||
// Auto-migrate from db.json if exists
|
||||
|
||||
@@ -868,6 +868,36 @@ export async function touchConnectionLastUsed(
|
||||
});
|
||||
}
|
||||
|
||||
/**
|
||||
* Lightweight backoff reset — runs a targeted UPDATE without SELECT or re-encrypt.
|
||||
* Follows the `clearConnectionErrorIfUnchanged` pattern but without the CAS check,
|
||||
* since the caller already verified the connection is eligible for reset.
|
||||
* Resets all backoff/error columns so the connection re-enters the selection pool.
|
||||
* Does invalidateDbCache + bumpProxyConfigGeneration since backoff affects priority.
|
||||
*/
|
||||
export async function resetConnectionBackoff(id: string): Promise<void> {
|
||||
if (!id) return;
|
||||
const db = getDbInstance() as unknown as DbLike;
|
||||
const now = new Date().toISOString();
|
||||
db.prepare(
|
||||
`UPDATE provider_connections SET
|
||||
backoff_level = 0,
|
||||
test_status = 'active',
|
||||
last_error = NULL,
|
||||
last_error_at = NULL,
|
||||
last_error_type = NULL,
|
||||
last_error_source = NULL,
|
||||
error_code = NULL,
|
||||
updated_at = @updatedAt
|
||||
WHERE id = @id`
|
||||
).run({
|
||||
updatedAt: now,
|
||||
id,
|
||||
});
|
||||
invalidateDbCache("connections");
|
||||
bumpProxyConfigGeneration();
|
||||
}
|
||||
|
||||
export async function deleteProviderConnection(id: string) {
|
||||
const db = getDbInstance() as unknown as DbLike;
|
||||
const existing = db.prepare("SELECT provider FROM provider_connections WHERE id = ?").get(id);
|
||||
@@ -975,102 +1005,10 @@ export async function getDistinctGroups(): Promise<string[]> {
|
||||
return rows.map((r) => String(r.group ?? "")).filter(Boolean);
|
||||
}
|
||||
|
||||
// ──────────────── Auto Migration ────────────────
|
||||
|
||||
/**
|
||||
* Scans all connections and re-encrypts any fields using the old dynamic salt
|
||||
* so they use the new canonical static salt.
|
||||
*/
|
||||
export function autoMigrateLegacyEncryptedConnections(): number {
|
||||
const db = getDbInstance() as unknown as DbLike;
|
||||
const rows = db.prepare("SELECT * FROM provider_connections").all();
|
||||
let migratedCount = 0;
|
||||
|
||||
for (const row of rows) {
|
||||
const camelRow = rowToCamel(row);
|
||||
if (!camelRow) continue;
|
||||
|
||||
let updatedRow = false;
|
||||
|
||||
const encryptedFields = ["apiKey", "idToken", "accessToken", "refreshToken"];
|
||||
for (const field of encryptedFields) {
|
||||
if (typeof camelRow[field] === "string") {
|
||||
const { updated, value } = migrateLegacyEncryptedString(camelRow[field] as string);
|
||||
if (updated) {
|
||||
camelRow[field] = value;
|
||||
updatedRow = true;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
if (updatedRow) {
|
||||
// camelRow[field] is already re-encrypted!
|
||||
// But _updateConnectionRow does not re-encrypt automatically, so we pass it safely.
|
||||
// Wait, _updateConnectionRow runs the full data through `encryptConnectionFields`,
|
||||
// but `encryptConnectionFields` will re-encrypt plain text.
|
||||
// BUT `migrateLegacyEncryptedString` returns ALREADY ENCRYPTED ciphertext!
|
||||
// Wait... if we pass ALREADY ENCRYPTED text to `_updateConnectionRow`,
|
||||
// `encryptConnectionFields` in `_updateConnectionRow` will encrypt it AGAIN!
|
||||
// Let's modify the DB directly so we don't double encrypt.
|
||||
|
||||
db.prepare(
|
||||
"UPDATE provider_connections SET api_key = @apiKey, id_token = @idToken, access_token = @accessToken, refresh_token = @refreshToken, updated_at = @updatedAt WHERE id = @id"
|
||||
).run({
|
||||
id: camelRow.id,
|
||||
apiKey: camelRow.apiKey ?? null,
|
||||
idToken: camelRow.idToken ?? null,
|
||||
accessToken: camelRow.accessToken ?? null,
|
||||
refreshToken: camelRow.refreshToken ?? null,
|
||||
updatedAt: new Date().toISOString(),
|
||||
});
|
||||
migratedCount++;
|
||||
}
|
||||
}
|
||||
|
||||
if (migratedCount > 0) {
|
||||
backupDbFile("pre-write");
|
||||
invalidateDbCache("connections");
|
||||
console.log(`[DB] Auto-migrated ${migratedCount} connection(s) to new static-salt encryption.`);
|
||||
}
|
||||
|
||||
return migratedCount;
|
||||
}
|
||||
|
||||
export function getGheCopilotHosts(): string[] {
|
||||
const hosts = new Set<string>();
|
||||
try {
|
||||
const db = getDbInstance();
|
||||
const rows = db
|
||||
.prepare(
|
||||
"SELECT provider_specific_data FROM provider_connections WHERE provider = 'ghe-copilot' AND is_active = 1"
|
||||
)
|
||||
.all() as { provider_specific_data: string | null }[];
|
||||
for (const row of rows) {
|
||||
if (!row.provider_specific_data) continue;
|
||||
try {
|
||||
const psd = JSON.parse(row.provider_specific_data);
|
||||
const urls = [psd.gheUrl, psd.copilotApiUrl, psd.copilotProxyUrl];
|
||||
for (const urlStr of urls) {
|
||||
if (typeof urlStr === "string" && urlStr.trim()) {
|
||||
try {
|
||||
const url = new URL(urlStr);
|
||||
if (url.hostname) {
|
||||
hosts.add(url.hostname.toLowerCase());
|
||||
}
|
||||
} catch {
|
||||
// Ignore invalid URLs
|
||||
}
|
||||
}
|
||||
}
|
||||
} catch {
|
||||
// Ignore JSON parse errors
|
||||
}
|
||||
}
|
||||
} catch (err) {
|
||||
console.error("[DB] getGheCopilotHosts: failed to read GHE Copilot connections", err);
|
||||
}
|
||||
return [...hosts];
|
||||
}
|
||||
export {
|
||||
autoMigrateLegacyEncryptedConnections,
|
||||
getGheCopilotHosts,
|
||||
} from "./providers/migrations";
|
||||
|
||||
// ──────────────── Re-exports from leaf modules ────────────────
|
||||
|
||||
|
||||
120
src/lib/db/providers/migrations.ts
Normal file
120
src/lib/db/providers/migrations.ts
Normal file
@@ -0,0 +1,120 @@
|
||||
/**
|
||||
* db/providers/migrations — Connection-level migration utilities and GHE Copilot host discovery.
|
||||
*
|
||||
* Split from provider.ts to keep the main CRUD file under the size ratchet.
|
||||
*/
|
||||
|
||||
import { getDbInstance, rowToCamel } from "../core";
|
||||
import { backupDbFile } from "../backup";
|
||||
import { migrateLegacyEncryptedString } from "../encryption";
|
||||
import { invalidateDbCache } from "../readCache";
|
||||
|
||||
interface StatementLike<TRow = unknown> {
|
||||
all: (...params: unknown[]) => TRow[];
|
||||
get: (...params: unknown[]) => TRow | undefined;
|
||||
run: (...params: unknown[]) => { changes?: number };
|
||||
}
|
||||
|
||||
interface DbLike {
|
||||
prepare: <TRow = unknown>(sql: string) => StatementLike<TRow>;
|
||||
transaction: <T>(fn: () => T) => () => T;
|
||||
}
|
||||
|
||||
// ──────────────── Auto Migration ────────────────
|
||||
|
||||
/**
|
||||
* Scans all connections and re-encrypts any fields using the old dynamic salt
|
||||
* so they use the new canonical static salt.
|
||||
*/
|
||||
export function autoMigrateLegacyEncryptedConnections(): number {
|
||||
const db = getDbInstance() as unknown as DbLike;
|
||||
const rows = db.prepare("SELECT * FROM provider_connections").all();
|
||||
let migratedCount = 0;
|
||||
|
||||
for (const row of rows) {
|
||||
const camelRow = rowToCamel(row);
|
||||
if (!camelRow) continue;
|
||||
|
||||
let updatedRow = false;
|
||||
|
||||
const encryptedFields = ["apiKey", "idToken", "accessToken", "refreshToken"];
|
||||
for (const field of encryptedFields) {
|
||||
if (typeof camelRow[field] === "string") {
|
||||
const { updated, value } = migrateLegacyEncryptedString(camelRow[field] as string);
|
||||
if (updated) {
|
||||
camelRow[field] = value;
|
||||
updatedRow = true;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
if (updatedRow) {
|
||||
// camelRow[field] is already re-encrypted!
|
||||
// But _updateConnectionRow does not re-encrypt automatically, so we pass it safely.
|
||||
// Wait, _updateConnectionRow runs the full data through `encryptConnectionFields`,
|
||||
// but `encryptConnectionFields` will re-encrypt plain text.
|
||||
// BUT `migrateLegacyEncryptedString` returns ALREADY ENCRYPTED ciphertext!
|
||||
// Wait... if we pass ALREADY ENCRYPTED text to `_updateConnectionRow`,
|
||||
// `encryptConnectionFields` in `_updateConnectionRow` will encrypt it AGAIN!
|
||||
// Let's modify the DB directly so we don't double encrypt.
|
||||
|
||||
db.prepare(
|
||||
"UPDATE provider_connections SET api_key = @apiKey, id_token = @idToken, access_token = @accessToken, refresh_token = @refreshToken, updated_at = @updatedAt WHERE id = @id"
|
||||
).run({
|
||||
id: camelRow.id,
|
||||
apiKey: camelRow.apiKey ?? null,
|
||||
idToken: camelRow.idToken ?? null,
|
||||
accessToken: camelRow.accessToken ?? null,
|
||||
refreshToken: camelRow.refreshToken ?? null,
|
||||
updatedAt: new Date().toISOString(),
|
||||
});
|
||||
migratedCount++;
|
||||
}
|
||||
}
|
||||
|
||||
if (migratedCount > 0) {
|
||||
backupDbFile("pre-write");
|
||||
invalidateDbCache("connections");
|
||||
console.log(`[DB] Auto-migrated ${migratedCount} connection(s) to new static-salt encryption.`);
|
||||
}
|
||||
|
||||
return migratedCount;
|
||||
}
|
||||
|
||||
// ──────────────── GHE Copilot ────────────────
|
||||
|
||||
export function getGheCopilotHosts(): string[] {
|
||||
const hosts = new Set<string>();
|
||||
try {
|
||||
const db = getDbInstance();
|
||||
const rows = db
|
||||
.prepare(
|
||||
"SELECT provider_specific_data FROM provider_connections WHERE provider = 'ghe-copilot' AND is_active = 1"
|
||||
)
|
||||
.all() as { provider_specific_data: string | null }[];
|
||||
for (const row of rows) {
|
||||
if (!row.provider_specific_data) continue;
|
||||
try {
|
||||
const psd = JSON.parse(row.provider_specific_data);
|
||||
const urls = [psd.gheUrl, psd.copilotApiUrl, psd.copilotProxyUrl];
|
||||
for (const urlStr of urls) {
|
||||
if (typeof urlStr === "string" && urlStr.trim()) {
|
||||
try {
|
||||
const url = new URL(urlStr);
|
||||
if (url.hostname) {
|
||||
hosts.add(url.hostname.toLowerCase());
|
||||
}
|
||||
} catch {
|
||||
// Ignore invalid URLs
|
||||
}
|
||||
}
|
||||
}
|
||||
} catch {
|
||||
// Ignore JSON parse errors
|
||||
}
|
||||
}
|
||||
} catch (err) {
|
||||
console.error("[DB] getGheCopilotHosts: failed to read GHE Copilot connections", err);
|
||||
}
|
||||
return [...hosts];
|
||||
}
|
||||
@@ -87,6 +87,9 @@ export function ensureProviderConnectionsColumns(db: SqliteDatabase) {
|
||||
db.exec(
|
||||
"CREATE INDEX IF NOT EXISTS idx_pc_auth_active_refresh ON provider_connections(auth_type, is_active, refresh_token)"
|
||||
);
|
||||
db.exec(
|
||||
"CREATE INDEX IF NOT EXISTS idx_pc_provider_auth_type ON provider_connections(provider, auth_type)"
|
||||
);
|
||||
} catch (error: unknown) {
|
||||
const message = error instanceof Error ? error.message : String(error);
|
||||
console.warn("[DB] Failed to verify provider_connections schema:", message);
|
||||
@@ -147,6 +150,20 @@ export function ensureUsageHistoryColumns(db: SqliteDatabase) {
|
||||
db.exec("ALTER TABLE usage_history ADD COLUMN account_label_priority INTEGER DEFAULT 0");
|
||||
console.log("[DB] Added usage_history.account_label_priority column");
|
||||
}
|
||||
db.exec(
|
||||
"CREATE INDEX IF NOT EXISTS idx_uh_provider_model_timestamp ON usage_history(provider, model, timestamp)"
|
||||
);
|
||||
db.exec(
|
||||
`CREATE INDEX IF NOT EXISTS idx_uh_dedup ON usage_history(
|
||||
timestamp,
|
||||
COALESCE(provider, ''),
|
||||
COALESCE(model, ''),
|
||||
COALESCE(connection_id, ''),
|
||||
COALESCE(api_key_id, ''),
|
||||
tokens_input,
|
||||
tokens_output
|
||||
)`
|
||||
);
|
||||
} catch (error: unknown) {
|
||||
const message = error instanceof Error ? error.message : String(error);
|
||||
console.warn("[DB] Failed to verify usage_history schema:", message);
|
||||
|
||||
@@ -768,3 +768,4 @@ export {
|
||||
getCacheTrend,
|
||||
resetCacheMetrics,
|
||||
} from "./settings/cacheMetrics";
|
||||
export { getCachedSettings } from "./readCache";
|
||||
|
||||
@@ -13,6 +13,7 @@ export {
|
||||
getProviderConnectionById,
|
||||
createProviderConnection,
|
||||
updateProviderConnection,
|
||||
resetConnectionBackoff,
|
||||
clearConnectionErrorIfUnchanged,
|
||||
touchConnectionLastUsed,
|
||||
deleteProviderConnection,
|
||||
@@ -32,7 +33,6 @@ export {
|
||||
isConnectionRateLimited,
|
||||
getRateLimitedConnections,
|
||||
clearStaleCrashCooldowns,
|
||||
|
||||
// T13: Stale quota display fix (zero out usage after window resets)
|
||||
getEffectiveQuotaUsage,
|
||||
formatResetCountdown,
|
||||
|
||||
@@ -79,11 +79,15 @@ export function clearEgressCache(): void {
|
||||
* echo-IP round-trip. Returns null if not yet probed.
|
||||
*/
|
||||
export function getCachedEgressIp(proxyUrl: string | null): string | null {
|
||||
const cached = egressCache.get(proxyUrl ?? "__direct__");
|
||||
const key = proxyUrl ?? "__direct__";
|
||||
const cached = egressCache.get(key);
|
||||
if (!cached) return null;
|
||||
if (Date.now() - cached.at >= EGRESS_CACHE_TTL_MS) return null;
|
||||
if (Date.now() - cached.at >= EGRESS_CACHE_TTL_MS) {
|
||||
egressCache.delete(key);
|
||||
return null;
|
||||
}
|
||||
return cached.ip;
|
||||
}
|
||||
}
|
||||
|
||||
const warmingInFlight = new Set<string>();
|
||||
|
||||
|
||||
@@ -1,62 +0,0 @@
|
||||
/**
|
||||
* Settings Cache — FASE-03 Architecture Refactoring
|
||||
*
|
||||
* In-memory cache for settings to eliminate self-fetch anti-pattern in middleware.
|
||||
* The middleware was making HTTP requests to its own /api/settings endpoint,
|
||||
* which caused circular dependencies and performance issues.
|
||||
*
|
||||
* @module settingsCache
|
||||
*/
|
||||
|
||||
import { getSettings } from "@/lib/localDb";
|
||||
|
||||
/** @type {{ data: object|null, lastFetch: number, ttl: number }} */
|
||||
const cache = {
|
||||
data: null,
|
||||
lastFetch: 0,
|
||||
ttl: 5000, // 5 seconds TTL
|
||||
};
|
||||
|
||||
/**
|
||||
* Get settings from cache (or refresh if stale).
|
||||
* This replaces the self-fetch pattern in middleware.
|
||||
*
|
||||
* @returns {Promise<object>} Settings object
|
||||
*/
|
||||
export async function getCachedSettings() {
|
||||
const now = Date.now();
|
||||
|
||||
if (cache.data && now - cache.lastFetch < cache.ttl) {
|
||||
return cache.data;
|
||||
}
|
||||
|
||||
try {
|
||||
const settings = await getSettings();
|
||||
cache.data = settings;
|
||||
cache.lastFetch = now;
|
||||
return settings;
|
||||
} catch (err) {
|
||||
// If fetch fails but we have stale data, return it
|
||||
if (cache.data) {
|
||||
console.error("[SettingsCache] Failed to refresh, using stale data:", err.message);
|
||||
return cache.data;
|
||||
}
|
||||
throw err;
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Invalidate the cache (e.g. after settings update).
|
||||
*/
|
||||
export function invalidateSettingsCache() {
|
||||
cache.data = null;
|
||||
cache.lastFetch = 0;
|
||||
}
|
||||
|
||||
/**
|
||||
* Set the cache TTL in milliseconds.
|
||||
* @param {number} ms
|
||||
*/
|
||||
export function setSettingsCacheTTL(ms) {
|
||||
cache.ttl = ms;
|
||||
}
|
||||
@@ -6,10 +6,11 @@ import {
|
||||
getCachedProviderNodes,
|
||||
validateApiKey,
|
||||
updateProviderConnection,
|
||||
touchConnectionLastUsed,
|
||||
clearConnectionErrorIfUnchanged,
|
||||
resetConnectionBackoff,
|
||||
getSettings,
|
||||
getCachedSettings,
|
||||
touchConnectionLastUsed,
|
||||
clearConnectionErrorIfUnchanged,
|
||||
} from "@/lib/localDb";
|
||||
import {
|
||||
createLazyConnectionView,
|
||||
@@ -594,15 +595,22 @@ function getConnectionRecencyPenalty(connection: ProviderConnectionView): number
|
||||
function getP2CConnectionScore(
|
||||
provider: string,
|
||||
connection: ProviderConnectionView,
|
||||
requestedModel: string | null = null
|
||||
requestedModel: string | null = null,
|
||||
quotaResults?: Map<string, { blocked: boolean; exhausted: boolean }>
|
||||
): { score: number; quotaHeadroomPercent: number | null } {
|
||||
const quotaBlocked = evaluateQuotaLimitPolicy(provider, connection, requestedModel).blocked;
|
||||
const quotaExhausted = isQuotaExhaustedForRequest(connection.id, provider, requestedModel);
|
||||
const quotaHeadroomPercent = getConnectionQuotaHeadroomPercent(
|
||||
provider,
|
||||
connection,
|
||||
requestedModel
|
||||
);
|
||||
let quotaBlocked: boolean;
|
||||
let quotaExhausted: boolean;
|
||||
|
||||
if (connection.id && quotaResults?.has(connection.id)) {
|
||||
const cached = quotaResults.get(connection.id)!;
|
||||
quotaBlocked = cached.blocked;
|
||||
quotaExhausted = cached.exhausted;
|
||||
} else {
|
||||
quotaBlocked = evaluateQuotaLimitPolicy(provider, connection, requestedModel).blocked;
|
||||
quotaExhausted = isQuotaExhaustedForRequest(connection.id, provider, requestedModel);
|
||||
}
|
||||
|
||||
const quotaHeadroomPercent = getConnectionQuotaHeadroomPercent(provider, connection, requestedModel);
|
||||
|
||||
let quotaPenalty = 0;
|
||||
if (quotaHeadroomPercent !== null) {
|
||||
@@ -630,10 +638,11 @@ function compareP2CConnections(
|
||||
provider: string,
|
||||
a: ProviderConnectionView,
|
||||
b: ProviderConnectionView,
|
||||
requestedModel: string | null = null
|
||||
requestedModel: string | null = null,
|
||||
quotaResults?: Map<string, { blocked: boolean; exhausted: boolean }>
|
||||
): number {
|
||||
const aScore = getP2CConnectionScore(provider, a, requestedModel);
|
||||
const bScore = getP2CConnectionScore(provider, b, requestedModel);
|
||||
const aScore = getP2CConnectionScore(provider, a, requestedModel, quotaResults);
|
||||
const bScore = getP2CConnectionScore(provider, b, requestedModel, quotaResults);
|
||||
if (aScore.score !== bScore.score) {
|
||||
return aScore.score - bScore.score;
|
||||
}
|
||||
@@ -1146,32 +1155,39 @@ export async function getProviderCredentials(
|
||||
!isAccountUnavailable(c.rateLimitedUntil)
|
||||
) {
|
||||
c.backoffLevel = 0;
|
||||
updateProviderConnection(c.id, {
|
||||
backoffLevel: 0,
|
||||
testStatus: "active",
|
||||
lastError: null,
|
||||
lastErrorAt: null,
|
||||
lastErrorType: null,
|
||||
lastErrorSource: null,
|
||||
errorCode: null,
|
||||
}).catch(() => {});
|
||||
resetConnectionBackoff(c.id).catch(() => {});
|
||||
}
|
||||
}
|
||||
|
||||
let modelLockedCount = 0;
|
||||
let familyLockedCount = 0;
|
||||
const connectionFilterStatus = new Map<string, string>();
|
||||
// Filter out unavailable accounts and excluded connection
|
||||
const availableConnections = connections.filter((c) => {
|
||||
if (excludedConnectionIds.has(c.id)) return false;
|
||||
if (excludedConnectionIds.has(c.id)) {
|
||||
connectionFilterStatus.set(c.id, "excluded");
|
||||
return false;
|
||||
}
|
||||
if (requestedModel && isModelExcludedByConnection(requestedModel, c.providerSpecificData)) {
|
||||
connectionFilterStatus.set(c.id, "modelExcluded");
|
||||
return false;
|
||||
}
|
||||
if (!allowSuppressedConnections) {
|
||||
if (!allowRateLimitedConnections && isAccountUnavailable(c.rateLimitedUntil)) return false;
|
||||
if (isTerminalConnectionStatus(c)) return false;
|
||||
if (provider === "codex" && isCodexScopeUnavailable(c, requestedModel)) return false;
|
||||
if (!allowRateLimitedConnections && isAccountUnavailable(c.rateLimitedUntil)) {
|
||||
connectionFilterStatus.set(c.id, "rateLimited");
|
||||
return false;
|
||||
}
|
||||
if (isTerminalConnectionStatus(c)) {
|
||||
connectionFilterStatus.set(c.id, "terminalStatus");
|
||||
return false;
|
||||
}
|
||||
if (provider === "codex" && isCodexScopeUnavailable(c, requestedModel)) {
|
||||
connectionFilterStatus.set(c.id, "codexScopeLimited");
|
||||
return false;
|
||||
}
|
||||
// Per-model lockout: if this specific model/family is locked on this connection, skip it
|
||||
if (requestedModel && isModelLocked(provider, c.id, requestedModel)) {
|
||||
connectionFilterStatus.set(c.id, "modelLocked");
|
||||
if (
|
||||
provider === "antigravity" &&
|
||||
getQuotaScopeLabelForProvider(provider, requestedModel) === "family"
|
||||
@@ -1183,6 +1199,7 @@ export async function getProviderCredentials(
|
||||
return false;
|
||||
}
|
||||
}
|
||||
connectionFilterStatus.set(c.id, "available");
|
||||
return true;
|
||||
});
|
||||
|
||||
@@ -1197,15 +1214,13 @@ export async function getProviderCredentials(
|
||||
);
|
||||
}
|
||||
connections.forEach((c) => {
|
||||
const excluded = excludedConnectionIds.has(c.id);
|
||||
const rateLimited = isAccountUnavailable(c.rateLimitedUntil);
|
||||
const terminalStatus = isTerminalConnectionStatus(c);
|
||||
const codexScopeLimited = provider === "codex" && isCodexScopeUnavailable(c, requestedModel);
|
||||
const modelLocked =
|
||||
Boolean(requestedModel) && isModelLocked(provider, c.id, requestedModel as string);
|
||||
const modelExcluded =
|
||||
Boolean(requestedModel) &&
|
||||
isModelExcludedByConnection(requestedModel as string, c.providerSpecificData);
|
||||
const status = connectionFilterStatus.get(c.id);
|
||||
const excluded = status === "excluded";
|
||||
const rateLimited = status === "rateLimited";
|
||||
const terminalStatus = status === "terminalStatus";
|
||||
const codexScopeLimited = status === "codexScopeLimited";
|
||||
const modelLocked = status === "modelLocked";
|
||||
const modelExcluded = status === "modelExcluded";
|
||||
if (excluded || rateLimited) {
|
||||
log.debug(
|
||||
"AUTH",
|
||||
@@ -1334,10 +1349,12 @@ export async function getProviderCredentials(
|
||||
reasons: string[];
|
||||
resetAt: string | null;
|
||||
}> = [];
|
||||
const quotaResults = new Map<string, { blocked: boolean; exhausted: boolean }>();
|
||||
|
||||
if (!bypassQuotaPolicy) {
|
||||
policyEligibleConnections = availableConnections.filter((connection) => {
|
||||
const evaluation = evaluateQuotaLimitPolicy(provider, connection, requestedModel);
|
||||
quotaResults.set(connection.id, { blocked: evaluation.blocked, exhausted: false });
|
||||
if (!evaluation.blocked) return true;
|
||||
|
||||
blockedByPolicy.push({
|
||||
@@ -1377,13 +1394,19 @@ export async function getProviderCredentials(
|
||||
};
|
||||
}
|
||||
|
||||
// Quota-aware: filter out accounts with exhausted quota for the requested scope.
|
||||
const withQuota = policyEligibleConnections.filter(
|
||||
(c) => !isQuotaExhaustedForRequest(c.id, provider, requestedModel)
|
||||
);
|
||||
const exhaustedQuota = policyEligibleConnections.filter((c) =>
|
||||
isQuotaExhaustedForRequest(c.id, provider, requestedModel)
|
||||
);
|
||||
// Quota-aware: partition accounts with and without quota for the requested scope.
|
||||
const withQuota: typeof policyEligibleConnections = [];
|
||||
const exhaustedQuota: typeof policyEligibleConnections = [];
|
||||
for (const c of policyEligibleConnections) {
|
||||
const exhausted = isQuotaExhaustedForRequest(c.id, provider, requestedModel);
|
||||
const existing = quotaResults.get(c.id);
|
||||
if (existing) existing.exhausted = exhausted;
|
||||
if (!exhausted) {
|
||||
withQuota.push(c);
|
||||
} else {
|
||||
exhaustedQuota.push(c);
|
||||
}
|
||||
}
|
||||
|
||||
if (exhaustedQuota.length > 0) {
|
||||
log.info(
|
||||
@@ -1551,7 +1574,7 @@ export async function getProviderCredentials(
|
||||
// health instead of defaulting to random-first selection.
|
||||
if (candidatePool.length <= 2) {
|
||||
connection = [...candidatePool].sort((a, b) =>
|
||||
compareP2CConnections(provider, a, b, requestedModel)
|
||||
compareP2CConnections(provider, a, b, requestedModel, quotaResults)
|
||||
)[0];
|
||||
} else {
|
||||
const i =
|
||||
@@ -1561,7 +1584,7 @@ export async function getProviderCredentials(
|
||||
if (j >= i) j++;
|
||||
const a = candidatePool[i];
|
||||
const b = candidatePool[j];
|
||||
connection = compareP2CConnections(provider, a, b, requestedModel) <= 0 ? a : b;
|
||||
connection = compareP2CConnections(provider, a, b, requestedModel, quotaResults) <= 0 ? a : b;
|
||||
}
|
||||
} else if (strategy === "random") {
|
||||
// Random: Fisher-Yates-inspired random pick
|
||||
|
||||
@@ -126,7 +126,7 @@ export const DEFAULT_DATABASE_SETTINGS: Omit<DatabaseSettings, "location" | "sta
|
||||
scheduledVacuum: "weekly",
|
||||
vacuumHour: 2,
|
||||
pageSize: 4096,
|
||||
cacheSize: 16384,
|
||||
cacheSize: 65536,
|
||||
optimizeOnStartup: true,
|
||||
},
|
||||
};
|
||||
|
||||
Reference in New Issue
Block a user