Index creation - extreme low performance

I just intalled percona on our productive server to see if it will benefit from it.
So far I am not overwhelmed.

It’s a a 30k IOPS AWS EBS IO1 storage with (tested) 600mb/sec throughput at any time.
The server is 32 cores, totally idle.
IO is totally idle.

I did the ADD index command and monitored the server:
avg-cpu: %user %nice %system %iowait %steal %idle
8.36 0.00 0.76 1.91 0.00 88.97

Device: tps kB_read/s kB_wrtn/s kB_read kB_wrtn
xvda 0.60 0.00 3.20 0 16
xvdh 1850.80 16113.60 10475.20 80568 52376
xvdg 0.00 0.00 0.00 0 0
xvdf 0.00 0.00 0.00 0 0
nvme0n1 0.00 0.00 0.00 0 0
nvme1n1 0.00 0.00 0.00 0 0
nvme2n1 0.00 0.00 0.00 0 0
nvme3n1 0.00 0.00 0.00 0 0

As you can see, it’s not using CPU or IO

What is the reason that percona 8.0.13 does not utilize the server’s resources ?

Adding an index on one table is not a very challenging test. Remember one operation, even a long running one, will utilize only one CPU core as well as usually a fraction of total IO bandwidth.
I’d suggest to run some multi threaded test, like sysbench OLTP with at least as many threads as CPU cores you have.
Then, check utilization with both

mpstat -P ALL 1

and

iostat -dx 1

for example.