Some service data was reset due to data corruption due to an instability with the non-existent API with the monitoring service.

This data is archived, and I'll see if I can eventually fix it upstream.