Debian 10 Buster can't renew DHCP as client, when IPv6 is disabled

Question

After I upgraded Debian 9 to 10 Buster server lost network connection after 24 hours since reboot, I think problem is it can't renew IP address from DHCP server.

In /var/lib/dhcp I see the file dhclient.eth0.leases with expired date inside.

This service was ok with Debian 9, with Debian 10 I see this error:

root@debian:~# systemctl status networking.service
● networking.service - Raise network interfaces
   Loaded: loaded (/lib/systemd/system/networking.service; enabled;
vendor preset: enabled)
   Active: failed (Result: exit-code) since Thu 2019-07-25 15:18:02
CEST; 2h 39min ago
     Docs: man:interfaces(5)
  Process: 380 ExecStart=/sbin/ifup -a --read-environment
(code=exited, status=1/FAILURE)
 Main PID: 380 (code=exited, status=1/FAILURE)

Jul 25 15:18:01 debian-2gb-nbg1-2 dhclient[402]: DHCPREQUEST for
116.203.xx.xx on eth0 to 255.255.255.255 port 67
Jul 25 15:18:01 debian-2gb-nbg1-2 dhclient[402]: DHCPACK of
116.203.xx.xx from 172.31.1.1
Jul 25 15:18:01 debian-2gb-nbg1-2 ifup[380]: DHCPACK of 116.203.xx.xx
from 172.31.1.1
Jul 25 15:18:01 debian-2gb-nbg1-2 dhclient[402]: bound to
116.203.xx.xx -- renewal in 34516 seconds.
Jul 25 15:18:01 debian-2gb-nbg1-2 ifup[380]: bound to 116.203.xx.xx
-- renewal in 34516 seconds.
Jul 25 15:18:02 debian-2gb-nbg1-2 ifup[380]: RTNETLINK answers:
Permission denied
Jul 25 15:18:02 debian-2gb-nbg1-2 ifup[380]: ifup: failed to bring up eth0:0
Jul 25 15:18:02 debian-2gb-nbg1-2 systemd[1]: networking.service: Main
process exited, code=exited, status=1/FAILURE
Jul 25 15:18:02 debian-2gb-nbg1-2 systemd[1]: networking.service:
Failed with result 'exit-code'.
Jul 25 15:18:02 debian-2gb-nbg1-2 systemd[1]: Failed to start Raise
network interfaces.

Thanks for any help.

Edit: problem solved, looks like Debian 10 doesn't like disabling of IPv6 by the same way as v9 do.

If anyone knows how to disable IPv6 in Buster, let us know.

# disabling of IPv6 in Debian 9, not working in Debian 10
/etc/sysctl.conf
    net.ipv6.conf.all.disable_ipv6=1
    net.ipv6.conf.default.disable_ipv6=1
    net.ipv6.conf.lo.disable_ipv6=1

this is a VPS server so log from dhcp server is not available, but I found what is wrong, now just how to solve it? — Roman_G, Jul 25 '19 at 16:43
Looks like "RTNETLINK answers: Permission denied" might be related to ipv6. Has its enabled/disabled status changed because of the upgrade? — Mark Wagner, Jul 25 '19 at 17:44
`cat /etc/network/interfaces # interfaces(5) file used by ifup(8) and ifdown(8) # loopback network interface auto lo iface lo inet loopback # Include files from /etc/network/interfaces.d: source /etc/network/interfaces.d/*.cfg` — Roman_G, Jul 25 '19 at 23:03
`cat /etc/network/interfaces.d/50-cloud-init.cfg # This file is generated from information provided by # the datasource. Changes to it will not persist across an instance. auto lo iface lo inet loopback auto eth0 iface eth0 inet dhcp dns-nameservers 213.133.xx.xx auto eth0:0 iface eth0:0 inet6 static address 2a01:4f8:c2c:xxxx::1/64 gateway fe80::1 post-up route add -A inet6 default gw fe80::1%eth0 || true pre-down route del -A inet6 default gw fe80::1%eth0 || true` — Roman_G, Jul 25 '19 at 23:07
I personally kill ipv6 by recompiling my kernel without ipv6 support. It is the only way to be sure. Some crap will surely happen what re-enables it... — peterh, Jul 29 '19 at 12:31
I see the same on aws and virtualbox vm's without disabling ipv6 — Matt, Mar 01 '21 at 03:32
@Roman_G Did you found any solution with disabled ipv6? I disable ipv6 without sysctl using grub: /etc/default/grub GRUB_CMDLINE_LINUX="ipv6.disable=1" update-grub And I wait to see if it works. — Christos Chatzaras, Oct 04 '21 at 09:28
Finally there is no way to make it work with ipv6 disabled. I had to replace dhcp with static ip configuration. — Christos Chatzaras, Oct 20 '21 at 16:25

Debian 10 Buster can't renew DHCP as client, when IPv6 is disabled

0 Answers0