Starting at 2:58pm PDT, a subset of requests to our west coast facilities were unable to reach Amazon S3. This would have caused client connections to timeout. The issue was resolved at 3:15pm PDT.