Freqtrade Logging and Monitoring: What to Watch
A practical monitoring checklist for self-hosted Freqtrade bots: service health, trade state, API failures, disk growth and evidence preservation.
Monitor the system, not only PnL
A profitable strategy can still fail operationally. Service state, stale data, API errors, disk pressure and restart behavior deserve explicit monitoring.
Logs need boundaries
Logs should preserve enough context to reconstruct failures without exposing secrets or growing until the server runs out of disk. Rotation and retention are operational controls.
State matters after restart
A restart test should confirm that persistent configuration, trade state and databases survive and that the process does not create unintended duplicate behavior.
Our deployment principle
The bot runs as a managed service on the trading host. Public reporting is separate, so monitoring or website failure does not need to become a trading-control failure.
Before trusting the result
Confirm that data timing is valid, configuration is reproducible, state survives restart, costs are represented and the exact strategy version is preserved. Operational correctness comes before interpreting PnL.