We're going to be taking the Chef Cluster (Where am I?) offline while we rebuild the indices from its database. This should take care of the "Duplicate entry xxxxx-xxx for key 1" errors a lot of people have received today.
I'm disabling comments on this post for the time being.
Update: The main database is running ok now, but we need to run maintenance on the other two machines in the cluster.