Most of the connectivity issues we were seeing were with instances in one of the us-east-1 AZs but we were seeing issues in other AZs in us-east-1 as well. Not sure why AWS is acting like this issue in only affecting one AZ.