We have been experiencing some problem with our backend deployment environment, which results in frequent short downtimes throughout the day. We managed to fix it at around 1:00 am on 10th after migrating our services to a better provider.