Troubleshooting
Everything fails with key_service_unavailable
The key service is sealed, usually after a restart. Check with ./rv keys status, and have three key holders run ./rv keys unseal.
A service shows “Not running here”
Optional services (logs, AI, proxy, adapters, files) only run when their Compose profile is on. Start it with docker compose --profile <name> up -d. A service that has never answered isn’t reported as down.
A service shows as down
Check its logs: docker compose logs <service>. Services report down after two missed checks (about a minute). If the control API itself is down, nothing is reported: check /healthz from outside.
Logs aren’t reaching my vendor
- Is the pipeline key right? A wrong key is refused.
- With “buffer and retry”, the processor answers 503 while the vault is unreachable; your collector should be retrying. Check the key service.
- Check the destination URL on the connection.
Someone can’t sign in with single sign-on
- Is their email in the configured domains?
- Do they exist in Onshore, or is account creation on, with a group that maps to a role?
- Did they leave every mapped group? Then they’re refused.
- Is the groups claim configured at your provider? See Provider guides.
“Too many attempts”
Eight failed sign-ins for one email or address lock it for 15 minutes.
A data type won’t publish
The message says why: samples failed (fix the pattern or the samples), it matches more than 5% of benchmark lines (narrow it), or it’s too slow (simplify the pattern).
Finding one request
Copy the request_id from the error, and search for it in Service logs.
Getting help
Run ./rv support-bundle and send the file. It holds versions, configuration without secrets, and redacted logs.