We want to move our monitoring system to be more focused on service availability rather than host health status. To this end we need to implement new alerts based on API Status, example:
GET /api/v1/measurements is not returning 200 for more than 5 minutes
POST /reports/<report_id> is timing out or taking more than a threshold
GET /health on the clickhouse proxy or fastpath is returning an error
We can use something similar to what is already done in the "API HTTPS Probes" dashboard but more focused on service availability
See ooni/backend#1071