On January 10, 2019, from 6:40 PM EST until 7:06 PM EST, most endpoints in the Platform API were unavailable, with clients receiving immediate error responses. During this same time, a few specific features within the Customer Portal were also unavailable, with users receiving errors when trying to access those features. No customer data was impacted.
Routine maintenance of a shared configuration storage system unexpectedly caused a full outage of that system. By design, most systems continued to operate normally during the outage, but a few systems became unavailable. We will be correcting those systems to also be robust against temporary outages of the shared configuration storage system. We will also be reviewing our maintenance procedures for the configuration system.