Spike arrest throws an Error every time it engages. In a production system this can happen quite often. Creating an Error object is inefficient because the JS VM has to walk the stack to populate Error.stack. Volos should try to avoid creating Error objects so frequently.