LXC second network isn't getting assigned static ipv6 upon reboot

Kris von Mach mach at swishmail.com
Sun Aug 23 13:11:06 CEST 2020


LXC second network isn't getting assigned static ipv6 upon reboot:

I've added second network in the VE web interface.

It shows up correctly in networks and is live:

root at test:/home/mach# cat /etc/network/interfaces
auto lo
iface lo inet loopback

auto eth0
iface eth0 inet static
         address 208.72.56.240/22
         gateway 208.72.56.1

iface eth0 inet6 static
         address 2602:ffb8::208:72:56:240/64
         gateway 2602:ffb8::1

auto net1
iface net1 inet static
         address 208.72.58.11/32
         gateway 208.72.56.1

iface net1 inet6 static
         address 2602:ffb8::3:2:2/64
         gateway 2602:ffb8::1

test:/home/mach# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN 
group default qlen 1000
     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
     inet 127.0.0.1/8 scope host lo
        valid_lft forever preferred_lft forever
     inet6 ::1/128 scope host
        valid_lft forever preferred_lft forever
2: eth0 at if220: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue 
state UP group default qlen 1000
     link/ether 2e:df:f8:2a:4b:af brd ff:ff:ff:ff:ff:ff link-netnsid 0
     inet 208.72.56.240/22 brd 208.72.59.255 scope global eth0
        valid_lft forever preferred_lft forever
     inet6 2602:ffb8::208:72:56:240/64 scope global
        valid_lft forever preferred_lft forever
     inet6 fe80::2cdf:f8ff:fe2a:4baf/64 scope link
        valid_lft forever preferred_lft forever
3: net1 at if224: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue 
state UP group default qlen 1000
     link/ether 8a:83:f7:1e:6a:c6 brd ff:ff:ff:ff:ff:ff link-netnsid 0
     inet 208.72.58.11/32 scope global net1
        valid_lft forever preferred_lft forever
     inet6 2602:ffb8::3:2:2/64 scope global
        valid_lft forever preferred_lft forever
     inet6 2602:ffb8::8883:f7ff:fe1e:6ac6/64 scope global dynamic mngtmpaddr
        valid_lft 2591888sec preferred_lft 604688sec
     inet6 fe80::8883:f7ff:fe1e:6ac6/64 scope link
        valid_lft forever preferred_lft forever

upon reboot net1 static 2602:ffb8::3:2:2/64 ipv6 is missing:

mach at test:~$ ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN 
group default qlen 1000
     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
     inet 127.0.0.1/8 scope host lo
        valid_lft forever preferred_lft forever
     inet6 ::1/128 scope host
        valid_lft forever preferred_lft forever
2: eth0 at if228: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue 
state UP group default qlen 1000
     link/ether 2e:df:f8:2a:4b:af brd ff:ff:ff:ff:ff:ff link-netnsid 0
     inet 208.72.56.240/22 brd 208.72.59.255 scope global eth0
        valid_lft forever preferred_lft forever
     inet6 2602:ffb8::208:72:56:240/64 scope global
        valid_lft forever preferred_lft forever
     inet6 fe80::2cdf:f8ff:fe2a:4baf/64 scope link
        valid_lft forever preferred_lft forever
3: net1 at if232: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue 
state UP group default qlen 1000
     link/ether 8a:83:f7:1e:6a:c6 brd ff:ff:ff:ff:ff:ff link-netnsid 0
     inet 208.72.58.11/32 scope global net1
        valid_lft forever preferred_lft forever
     inet6 2602:ffb8::8883:f7ff:fe1e:6ac6/64 scope global dynamic mngtmpaddr
        valid_lft 2591921sec preferred_lft 604721sec
     inet6 fe80::8883:f7ff:fe1e:6ac6/64 scope link
        valid_lft forever preferred_lft forever

If I use the web interface and edit net1 ipv6 and change it to a new 
address, it will work correctly again till reboot.

I also tried net.ipv6.conf.all.forwarding=1 which just got rid of the 
SLAAC address. The assigned static ip was still missing up reboot.

Am I doing something incorrectly, or is this a bug?


More information about the pve-user mailing list