Previous incidents

We have identified the root cause (an infrastructure migration script which overused database resources) and fixed it. All APIs are available again.