Approximately 17:30 UTC on 28 February, 2017, CloudSight detected issues with receiving blank images to our system and not being able to provide tags to images already in our system.
It was determined that issues upstream with Amazon Web Services (AWS) Simple Storage Service (S3) had experienced a significant outage, which was preventing images submitted to the CloudSight API and by CloudSight apps (CamFind and TapTapSee) from being received and stored on AWS S3, and returned to our main application for tagging said images.
We worked closely with AWS' support to monitor the situation, and created additional instances of our application so that when services became available, approximately at 22:00 UTC on 28 February, 2017, CloudSight was able to handle the queued volume of images to be tagged.
CloudSight experienced a high volume of images once AWS service was restored, but was back to normal operational capacity at approximately 23:00 UTC on 28 February, 2017.
We have continued to monitor this situation and monitor AWS services used by CloudSight, and will do so through 2 March, 2017.