Issues with API availability
Beginning on August 10 around 07:08 UTC until 10:30 UTC and again from 19:03 UTC to 19:18 UTC, the Heroku platform experienced issues reading and writing from/to Amazon S3. This is a service that we use for various functionality on the platform including storing slugs from new builds, launching new dynos, and storing and retrieving database backups. These services were in various states of degradation throughout the incident. We sincerely apologize for the problems this incident caused.