(because we’re already mirroring their dataset :exploding_head: - at least on with 16TB)