ominroute server is not booting

[paragrudani1]

I also tried previous versions but didn't work.

Thanks in advance!

[diegosouzapw]

Hey @paragrudani1! Thanks for the screenshot - that actually tells me the most important part: the server hangs at "Starting server..." with no error.

That silence is expected behavior, not a clue in itself: by default OmniRoute runs the server under a supervisor that doesn't print the child process's stdout - it only dumps logs if the process crashes. If boot hangs (instead of crashing), you see nothing. So the first step is to surface the real error:

omniroute serve --log

That streams the live server output and will show exactly where it stops. Please paste that here.

While you're at it, the usual culprits for a boot hang on a fresh macOS install:

1. better-sqlite3 native module - rebuild it: cd ~/.omniroute && npm rebuild better-sqlite3 (if build tools are missing: xcode-select --install)
2. Port 20128 already in use by a stale process - check with lsof -i :20128 and kill it, or start on another port: omniroute serve --port 30128
3. Corrupted/locked database - temporarily move it aside to test a clean boot: mv ~/.omniroute/storage.sqlite ~/.omniroute/storage.sqlite.bak then start again (your settings reset, but it isolates a DB issue)
4. Node version - OmniRoute needs Node >=20.20.2 (and not v21/v23). Check with node -v.

Run omniroute serve --log, paste the output, and we'll pin it down fast.

[333apollo]

Next.js 16.2.6

• Local: http://localhost:20128
• Network: http://0.0.0.0:20128
  ✓ Ready in 0ms
  [CREDENTIALS] No external credentials file found, using defaults.
  [DB] Preserved critical DB state before potential recreation: key_value(55), provider_connections(34), api_keys(3)
  [DB] SQLite database ready: C:\Users\murder.omniroute\storage.sqlite
  [STARTUP] Global fetch proxy patch initialized
  [HealthCheck] Starting proactive token health-check (tick every 60s)
  [Shutdown] Graceful shutdown handlers registered.
  [STARTUP] Spend batch writer started
  [STARTUP] Guardrail registry initialized
  [STARTUP] Builtin skill handlers registered
  [STARTUP] Quota cache background refresh started
  [ProviderLimitsSync] Scheduler started — interval: 70m
  [STARTUP] Provider limits sync scheduler started
  [ModelSync] Scheduler started — interval: 24h
  [STARTUP] Cloud/model sync background bootstrap initialized
  [BATCH] Initializing batch processor polling...
  [STARTUP] Batch processor started
  [STARTUP] Runtime settings hydrated: payloadRules, modelAliases, backgroundDegradation, cliCompatProviders, cacheControl, usageTracking, healthCheckLogs, thoughtSignature, modelsDevSync, corsOrigins, ccBridgeTransforms, systemTransforms
  [STARTUP] Model alias seed: applied=0, skipped=9, failed=0
  [HOT_RELOAD] Runtime config hot-reload started (poll=5000ms, fsWatch=on)
  [COMPLIANCE] Audit log table initialized
  [proxyLogger] Loaded 55 proxy logs from SQLite
  [LocalHealthCheck] Starting local provider health check (initial delay 15s)
  [ModelSync] No connections with autoSync enabled — skipping cycle

[diegosouzapw]

@333apollo thanks for the log dump. Looking at it: the server actually started successfully -- last visible line is [LocalHealthCheck] Starting l... and that is the health-check kicking off, not a boot failure. So the boot finished, the issue must be something else (port not reachable, dashboard 502, etc.).

@paragrudani1 -- a couple of follow-ups to narrow it down:

1. After omniroute serve --log finishes printing, does it open the dashboard at http://localhost:20128 ? Try curling: curl -I http://localhost:20128. Expecting HTTP/1.1 200 or 307.
2. If curl works but the browser does not, an antivirus/firewall is most likely blocking the loopback connection.
3. If curl fails too, paste the FULL --log output (looks like 333apollo's was truncated mid-line).

Boot itself looks healthy from the log -- next step is figuring out which downstream call is failing, not the startup.

[diegosouzapw]

@paragrudani1 -- still hitting the boot failure on the latest version (v3.8.16)? A lot has changed in the build/startup path since this thread. If it is resolved I will close it out; otherwise reply with the current startup log and I will trace it. Will close in ~7 days if there is no response.

[diegosouzapw]

Closing for inactivity -- feel free to reopen if you still hit this, or open a fresh issue with a current log. Thanks!