Post

Shrinking the problem instead of moving 56TB is the clever bit. Do the old runs just age out on Aurora under retention, so the migration eventually finishes itself with no cutover weekend?