Skip to main content

Troubleshooting & FAQ

Use this page to triage the most common problems before opening a ticket.

Quick checks​

  • Verify the server is reachable: redis-cli -h <host> -p <port> PING.
  • Confirm authentication works: AUTH <password> or your client’s auth call.
  • Inspect server logs for rejections, timeouts, or OOM messages.
  • Run a simple read-only query: MATCH (n) RETURN n LIMIT 1.

Connection issues​

  • Connection refused / timed out: ensure ports are open only to allowed sources; check container or pod health. In Kubernetes, confirm the Service type and NetworkPolicy allow traffic.
  • TLS failures: if using a TLS terminator, test plain TCP locally, then from the terminator; validate certificates and SNI.

Authentication and ACLs​

  • NOAUTH or NOPERM errors: the user lacks rights. Revisit ACL rules and ensure the client uses the correct username and password.
  • Rotate credentials and re-deploy clients that cache connections to avoid using stale secrets.

Query errors​

  • SyntaxError: run GRAPH.EXPLAIN on the query to spot syntax and planning issues early.
  • Parameter errors: ensure parameter names match exactly; prefer parameterized queries in clients to avoid quoting mistakes.
  • Timeouts: raise TIMEOUT_MAX cautiously or simplify the query. Use LIMIT while iterating on query shapes.

Performance and memory​

Data and indexing​

  • Missing results: confirm labels and relationship types match your queries; indexes only apply when the label/property matches the index definition.
  • Index creation slow: build indexes during off-peak hours and monitor resource usage; tune vector index options (M, efConstruction, efRuntime) per vector indexing.

When to ask for help​

  • Reproduce with the smallest possible dataset and query.
  • Include server version, client language and version, config overrides, and a sample query when filing an issue.