It appears that an AWS resource changed the network interface, however in doing so it prevented the /run tmpfs partion to write its update. After a series of failures, the network interface stopped working.
We are working with AWS to prevent this issue from happening in the future.