14
Monitoring and alerts
What we watch, and what a partner sees of it.
- In the sandboxlogs, metrics and alerts
- Not in the sandbox yetrail alerts
In plain words
Monitoring shows when something needs attention: a legal deadline getting close, a credential about to expire, the same error repeating, a route failing. The service writes logs with no invoice content, counts per route and records alerts. A partner sees the result through the care list and webhooks.
We watch four things: the legal clocks on Romania and Poland, the expiry of stored credentials, repeated errors for one client, and the health of each route. The service records what it finds as alerts. Logs and alerts hold ids, codes and counts and never an invoice body. In the sandbox an alert goes to the log; where alerts go in production is not decided.