
Our automated processes to detect and repair the instance were not successful since there was a hardware failure on one of the load balancers. The script normally handles this type of issue, but there was a unique occurrence we had not seen before. Needless to say, we have updated the script and can now handle this type of error.
Here is a log of the successful transactions so you can see that most went thru. The ramp up after 8 is normal since that is when more users are awake and beginning to sign up for races: