This incident occurred when an influx of usage in the eventScribe.net server pool created a CPU spike and the eventScribe.net servers scaled up more quickly than planned. Typically, a new server is put into service from the standby “warm pool,” which contains servers ready for use. In this case, CPU levels increased across the cluster of servers faster than could be supported by servers in the warm pool. As a result, servers not ready were put into service, resulting in errors. Once the servers became ready the issue was resolved. We are currently working to implement changes to prevent this from happening.