This incident has been resolved.

A fix has been implemented and we are monitoring the results.

AWS instance was in frozen state, automatic re-provisioning system wasn't able to act because of that, me manually updated config files to exclude the broken instance

We are currently investigating this issue.