We have a periodic job @ BinaryStoreErrorNotificationJob that sends binary provider related errors every 30 seconds if there are any. Right now the email we send is not very helpful since the content only specifies which binary provider had an error??and what its type is.
We should attach a complete stack trace to those emails, otherwise they are not very helpful. It is possible to try and trace back those errors on the log but it is cumbersome and may not be very straightforward on a multi-node setup.