Common issues and how to resolve them.
Almost always kernel limits. Confirm vm.max_map_count=262144 and the nofile limits (Install hardening), then restart the ClickHouse service.
External checkers only run when the flow threshold or a Custom Rule needs their signal. Add the checker to a rule, or lower the threshold, to force it.
Callbacks aren't sent to private networks. Ensure the callback_url is public (HTTPS) and the merchant key matches, so the HMAC verifies.
The API runs database migrations on start (alembic upgrade head), so it's briefly unavailable while they apply — especially on the first boot of a new version. Give it a minute; if it persists, check the api logs.
Reclaim space from old Docker layers with docker system prune, and cap analytics growth with a retention window in Settings. If ClickHouse is the culprit, it's the analytics volume — trimming old events frees it without touching leads (those live in Postgres).
Grab the relevant service logs and send them to us; we don't need access to your box. Then reach the team on Telegram or email.
docker compose logs --tail=200 api worker