Skip to content

Troubleshooting

  • Common Failures

    Timeouts, DB connectivity, schema mismatches.

  • Validation Errors

    Pydantic constraints fail fast with precise messages.

  • Recovery

    Clear caches, reindex, restart services.

Get started Configuration API

Read the Error

Pydantic tells you exactly which field failed validation and why. Fix the config, regenerate types if needed, and retry.

Logs

Use /api/docker/{container}/logs and application logs to pinpoint failures. For DB errors, also inspect Postgres and Neo4j logs.

Data Loss Risk

Avoid deleting DB volumes unless you intend a full reset. Back up before destructive actions.

Symptom → Action

Symptom Likely Cause Action
500 on /api/search DB unavailable Check /api/ready, restart DB containers
No results from graph Neo4j empty or disconnected Rebuild graph, check credentials
Validation error on /api/config Field constraints violated Adjust values to allowed ranges
Slow queries High max_hops, large top_k Reduce hops, tune indexes 
flowchart TB
    Error["Error"] --> Check["/api/ready"]
    Check -->|"ok"| Investigate["Inspect Logs"]
    Check -->|"fail"| Restart["Restart Services"]
    Investigate --> Fix["Config Tune"]
import httpx
print(httpx.get("http://127.0.0.1:8012/api/ready").json())  # readiness

curl -sS http://127.0.0.1:8012/api/ready | jq .
curl -sS http://127.0.0.1:8012/api/docker/status | jq .

await fetch('/api/ready').then(r => r.ok || Promise.reject('Not ready'))
  • Verify readiness
  • Inspect logs
  • Reduce search/fusion parameters
  • Reindex corpus
Cache Issues

If you suspect stale caches, clear retrieval caches (if enabled) and restart the API to invalidate in-memory state.