Ok, we’ve worked out the chain of events that caused this to happen. Unfortunately it was due to a change in our configuration that interacted badly with some of our old servers. We’ve resolved the issue, and have fixed it so things should all be really running fine now again.