In SERVER-114535 we removed the writeConflictRetry() loop from _findOrDeleteDocuments() in src/mongo/db/repl/storage_interface_impl.cpp when doing a find.

This means that a StorageUnavailableException could be thrown in case this function is called while a WT is opened and a write was previously done in the same transaction. Currently this should never happen, but if it ever did we would let the exception go up the stack and cancel the current user operation. This is fine, but we should add a comment explaining this, to make sure we don't have bad surprises in the future.