Noted backups weren’t running again because the nightly uploads to Amazon S3 didn’t appear on the Bandwidth graph:
So in the case of this Discourse / Digital Ocean instance, confirmed failure notifications didn’t occur and backups were failing due to space issues.