They haven't been since March 8 or 9. I don't have time to investigate right now, but I'll get to it if I can.
Not sure if related, we ran out of disk space during tonight's backup. I deleted the one that broke and the oldest one, but we probably need to do some maintenance and/or get more space provisioned, or this will happen again starting in 2 days. The backup script could also be made more efficient (make tgz file in 1 step (if our tar supports that), or at least delete sql files between tar and gzip), but it looks like it was generated from something, so I'm a little wary of editing it.