This is incorrect. EBS is part of the issue, as well as intermittent connectivity failures for EC2 instances.

We're also seeing connectivity issues with our elastic load balancers.