Amazon Web Services (AWS) had an issue with their Route 53 service that started a little after 11 AM (PDT). While they were having issues, it caused changes in the service to backup while waiting to be queued. This was not an issue for vCreative because we had already scaled up to our accommodate mid day traffic earlier in the morning.
In the afternoon however, as our processes took over and our website traffic decreased, we started releasing servers. When this happened, the servers went away, but Route 53 could not propagate the DNS changes out so that other servers knew they were no longer there, this caused many requests to fail simply trying to talk to servers that were no longer there. When amazon fixed their interrupted service, and processed through the backlog, vCreative service returned to normal.
We apologize for any inconvenience and want to thank our users for their patience.