There are thousands of them, they're expected, and they just pollute
output and make it hard to visually inspect results from earlier stages
of the VM image creation scripts.