Hello,

We have computers where just installing docker set the default bridge network to 172.17.0.0/16 (the more of them) and on some others computers this network is set to 172.18.0.0/24.
Docker don't seems consistent on this (this was the exactly same installation process, using Vagrant deployment).
No problem, i think this is because of some internal mechanism.

The problem is we wanted to set 172.17.0.0/24 for all, so we have set in daemon.json :

"bip": "172.17.0.0/24"

Docker cannot restart with this "bip": "172.17.0.0/24" setting.
But if we set "bip": "172.17.0.1/24", it works.

For information, we tested on several computers having already the native 172.17.0.0/24 and they can't either, this is not a problem of adresse availability on the computer. The only other network on all computers is eth0 with a 10.xx.xx.xx network. So there is no network/ip overlap.

Question 1 please, the bug:
Is it a bug that we cannot set "bip": "172.17.0.0/24" but only "bip": "172.17.0.1/24" ?
As result we cannot have the 172.17.0.0 IP address if we use bip option (so only starting from 172.17.0.1). But native configuration have it.

Question 2 please, impact :
Is it a problem for docker or all will work ok with 172.17.0.1/24 instead 172.17.0.0/24 ?

Thanks a lot.