Our apologies for the staging API being down for so long. We have a team performing a major hardware migration and reconfiguration in the datacenter (DC) that hosts our staging infrastructure. That work did not complete in time, and we failed to update the status to reflect that. We are fixing it now.
Unlike our production infrastructure, which is spread across multiple DCs for redundancy, our staging infrastructure only exists in one DC. If we take that DC offline, as we did yesterday, staging goes down with it.
All production API traffic was moved to a second DC so it should all be working fine. It should not have been affected by this work at any point.
We will complete the work within the next eight hours, hopefully much less.