Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.debian.user > #256012 > unrolled thread

Debian 12 IPv6 client?

Back to article view | Back to linux.debian.user

Contents

  Debian 12 IPv6 client? jeremy ardley <jeremy@ardley.org> - 2023-03-14 05:30 +0100
    Re: Debian 12 IPv6 client? Jeremy Ardley <jeremy@ardley.org> - 2023-03-14 06:10 +0100
    Re: Debian 12 IPv6 client? David Wright <deblis@lionunicorn.co.uk> - 2023-03-14 06:20 +0100
      Re: Debian 12 IPv6 client? Jeremy Ardley <jeremy@ardley.org> - 2023-03-14 08:10 +0100
        Re: Debian 12 IPv6 client? jeremy ardley <jeremy@ardley.org> - 2023-03-14 10:10 +0100
          Re: Debian 12 IPv6 client? David Wright <deblis@lionunicorn.co.uk> - 2023-03-14 19:00 +0100
            Re: Debian 12 IPv6 client? Jeremy Ardley <jeremy@ardley.org> - 2023-03-14 23:00 +0100
        Re: Debian 12 IPv6 client? David Wright <deblis@lionunicorn.co.uk> - 2023-03-14 19:00 +0100
    Re: Debian 12 IPv6 client? Charles Curley <charlescurley@charlescurley.com> - 2023-03-14 14:30 +0100

#256012 — Debian 12 IPv6 client?

I am testing the alpha 2 release of Debian 12
(I'm quite annoyed they have done away with /var/log/syslog)

My system is "pure" debian 12 and was net installed a few minutes before 
my checks. Nothing was changed from the original install.

My problem today is identifying what bit of the system is getting the 
IPv6 address. I can find nothing in journalctl

The networking  is pure systemd networkd.service controlled

root@debian12:/etc/network# systemctl status networking
● networking.service - Raise network interfaces
      Loaded: loaded (/lib/systemd/system/networking.service; enabled; 
preset: enabled)
      Active: active (exited) since Tue 2023-03-14 12:04:46 AWST; 23min ago
        Docs: man:interfaces(5)
     Process: 299 ExecStart=/sbin/ifup -a --read-environment 
(code=exited, status=0/SUCCESS)
     Process: 374 ExecStart=/bin/sh -c if [ -f 
/run/network/restart-hotplug ]; then /sbin/ifup -a --read-environment 
--allow=hotplug; fi (code=exited, status=0/SUCCESS)
    Main PID: 374 (code=exited, status=0/SUCCESS)
         CPU: 15ms

Mar 14 12:04:46 debian12 systemd[1]: Starting networking.service - Raise 
network interfaces...
Mar 14 12:04:46 debian12 systemd[1]: Finished networking.service - Raise 
network interfaces.


cat /etc.network/interfaces
# This file describes the network interfaces available on your system
# and how to activate them. For more information, see interfaces(5).

source /etc/network/interfaces.d/*

# The loopback network interface
auto lo
iface lo inet loopback

# The primary network interface
allow-hotplug enp0s3
iface enp0s3 inet dhcp
# This is an autoconfigured IPv6 interface
iface enp0s3 inet6 auto

Checking dhclient I see

ps ax | grep dhcl
     354 ?        Ss     0:00 dhclient -4 -v -i -pf 
/run/dhclient.enp0s3.pid -lf /var/lib/dhcp/dhclient.enp0s3.leases -I -df 
/var/lib/dhcp/dhclient6.enp0s3.leases enp0s3
     579 pts/0    S+     0:00 grep dhcl

So no dhclient -6 running.

root@debian12:/etc/network# journalctl | grep -i ipv6
Mar 14 11:48:25 debian12 kernel: Segment Routing with IPv6
Mar 14 11:48:25 debian12 kernel: In-situ OAM (IOAM) with IPv6
Mar 14 11:48:25 debian12 kernel: mip6: Mobile IPv6
Mar 14 11:48:26 debian12 kernel: IPv6: ADDRCONF(NETDEV_CHANGE): enp0s3: 
link becomes ready
Mar 14 12:04:46 debian12 kernel: Segment Routing with IPv6
Mar 14 12:04:46 debian12 kernel: In-situ OAM (IOAM) with IPv6
Mar 14 12:04:46 debian12 kernel: mip6: Mobile IPv6
Mar 14 12:04:48 debian12 kernel: IPv6: ADDRCONF(NETDEV_CHANGE): enp0s3: 
link becomes ready


My question remains. What is getting the IPv6 address and how can I make 
configuration changes if required?




Jeremy

[toc] | [next] | [standalone]

#256013

On 14/3/23 12:29, jeremy ardley wrote:
> I am testing the alpha 2 release of Debian 12
> (I'm quite annoyed they have done away with /var/log/syslog)
>
> My system is "pure" debian 12 and was net installed a few minutes 
> before my checks. Nothing was changed from the original install.
>
> My problem today is identifying what bit of the system is getting the 
> IPv6 address. I can find nothing in journalctl
>
> The networking  is pure systemd networkd.service controlled
>
Correction. I mistyped above. My system uses the systemd networking.service

root@debian12:/etc/network# systemctl status networking
● networking.service - Raise network interfaces
      Loaded: loaded (/lib/systemd/system/networking.service; enabled; 
preset: enabled)
      Active: active (exited) since Tue 2023-03-14 12:04:46 AWST; 23min ago
        Docs: man:interfaces(5)
     Process: 299 ExecStart=/sbin/ifup -a --read-environment 
(code=exited, status=0/SUCCESS)
     Process: 374 ExecStart=/bin/sh -c if [ -f 
/run/network/restart-hotplug ]; then /sbin/ifup -a --read-environment 
--allow=hotplug; fi (code=exited, status=0/SUCCESS)
    Main PID: 374 (code=exited, status=0/SUCCESS)
         CPU: 15ms

Mar 14 12:04:46 debian12 systemd[1]: Starting networking.service - Raise 
network interfaces...
Mar 14 12:04:46 debian12 systemd[1]: Finished networking.service - Raise 
network interfaces.

-- 
Jeremy
(Lists)

[toc] | [prev] | [next] | [standalone]

#256014

On Tue 14 Mar 2023 at 12:29:22 (+0800), jeremy ardley wrote:
> I am testing the alpha 2 release of Debian 12
> (I'm quite annoyed they have done away with /var/log/syslog)
> 
> My system is "pure" debian 12 and was net installed a few minutes
> before my checks. Nothing was changed from the original install.
> 
> My problem today is identifying what bit of the system is getting the
> IPv6 address. I can find nothing in journalctl

[ removed the typo ]

> cat /etc.network/interfaces
> # This file describes the network interfaces available on your system
> # and how to activate them. For more information, see interfaces(5).
> 
> source /etc/network/interfaces.d/*
> 
> # The loopback network interface
> auto lo
> iface lo inet loopback
> 
> # The primary network interface
> allow-hotplug enp0s3
> iface enp0s3 inet dhcp
> # This is an autoconfigured IPv6 interface
> iface enp0s3 inet6 auto
               ↑↑↑↑↑↑↑↑↑↑

I'm not very familiar with interfaces nowadays, but that looks
as if it's asking for enp0s3 to be autoconfigured with an IPv6
address. I thought that one got an IPv6 link address autoconfigured
anyway—I certainly do.

Did the debian-installer write that line (and comment)?

> Checking dhclient I see
> 
> ps ax | grep dhcl
>     354 ?        Ss     0:00 dhclient -4 -v -i -pf
> /run/dhclient.enp0s3.pid -lf /var/lib/dhcp/dhclient.enp0s3.leases -I
> -df /var/lib/dhcp/dhclient6.enp0s3.leases enp0s3
>     579 pts/0    S+     0:00 grep dhcl
> 
> So no dhclient -6 running.
> 
> root@debian12:/etc/network# journalctl | grep -i ipv6
> Mar 14 11:48:25 debian12 kernel: Segment Routing with IPv6
> Mar 14 11:48:25 debian12 kernel: In-situ OAM (IOAM) with IPv6
> Mar 14 11:48:25 debian12 kernel: mip6: Mobile IPv6
> Mar 14 11:48:26 debian12 kernel: IPv6: ADDRCONF(NETDEV_CHANGE):
> enp0s3: link becomes ready
> Mar 14 12:04:46 debian12 kernel: Segment Routing with IPv6
> Mar 14 12:04:46 debian12 kernel: In-situ OAM (IOAM) with IPv6
> Mar 14 12:04:46 debian12 kernel: mip6: Mobile IPv6
> Mar 14 12:04:48 debian12 kernel: IPv6: ADDRCONF(NETDEV_CHANGE):
> enp0s3: link becomes ready
> 
> 
> My question remains. What is getting the IPv6 address and how can I
> make configuration changes if required?

What's the output from:

  $ ip a

Mine (skipping lo):

  2: enp3s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
      link/ether a4:01:23:45:67:89 brd ff:ff:ff:ff:ff:ff
      inet 192.168.1.14/24 brd 192.168.1.255 scope global dynamic enp3s0
         valid_lft 74247sec preferred_lft 74247sec
      inet6 fe80::a601:23ff:fe45:6789/64 scope link      ← this here
         valid_lft forever preferred_lft forever

Cheers,
David.

[toc] | [prev] | [next] | [standalone]

#256017

On 14/3/23 13:15, David Wright wrote:
>
>> cat /etc.network/interfaces
>> # This file describes the network interfaces available on your system
>> # and how to activate them. For more information, see interfaces(5).
>>
>> source /etc/network/interfaces.d/*
>>
>> # The loopback network interface
>> auto lo
>> iface lo inet loopback
>>
>> # The primary network interface
>> allow-hotplug enp0s3
>> iface enp0s3 inet dhcp
>> # This is an autoconfigured IPv6 interface
>> iface enp0s3 inet6 auto
>                 ↑↑↑↑↑↑↑↑↑↑
>
> I'm not very familiar with interfaces nowadays, but that looks
> as if it's asking for enp0s3 to be autoconfigured with an IPv6
> address. I thought that one got an IPv6 link address autoconfigured
> anyway—I certainly do.
>
> Did the debian-installer write that line (and comment)?

Yes. It was a result of the installation process identifying the 
interface and generating the necessary /etc/network/interfaces

> What's the output from:
>
>    $ ip a
>
> Mine (skipping lo):
>
>    2: enp3s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
>        link/ether a4:01:23:45:67:89 brd ff:ff:ff:ff:ff:ff
>        inet 192.168.1.14/24 brd 192.168.1.255 scope global dynamic enp3s0
>           valid_lft 74247sec preferred_lft 74247sec
>        inet6 fe80::a601:23ff:fe45:6789/64 scope link      ← this here
>           valid_lft forever preferred_lft forever
>
root@debian12:/etc/network# 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: enp0s3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel 
state UP group default qlen 1000
     link/ether 08:00:27:18:b6:ac brd ff:ff:ff:ff:ff:ff
     inet 10.31.40.166/24 brd 10.31.40.255 scope global dynamic enp0s3
        valid_lft 76395sec preferred_lft 76395sec
     inet6 2403:5800:c101:b700:a00:27ff:fe18:b6ac/64 scope global 
dynamic mngtmpaddr
        valid_lft 6277sec preferred_lft 2676sec
     inet6 fe80::a00:27ff:fe18:b6ac/64 scope link
        valid_lft forever preferred_lft forever

The difference to yours is  I have a routable IPv6 address in addition 
to the local link.

Your  one has  an autogenerated address

inet6 fe80::a601:23ff:fe45:6789/64

This is made as a function of your ethernet interface MAC address

I have an autogenerated address

inet6 fe80::a00:27ff:fe18:b6ac/64

and an address that must have been created as  a result of an RA or DHCP 
message as it is prefixed with my IPv6 address range

inet6 2403:5800:c101:b700:a00:27ff:fe18:b6ac/64

I note my two address  have the same tails so I guess my routable IPv6 
address is generated from the interface MAC address

I conclude there is no IPv6 DHCP involved but there must be something 
that listens to RA announcements and generates a MAC derived address 
from that.

I still need to know how to control that as there are options relating 
to privacy that can do things like generate an expendable address in a 
different way and later generate another and deprecate earlier addresses.

-- 
Jeremy
(Lists)

[toc] | [prev] | [next] | [standalone]

#256021

On 14/3/23 16:21, Tim Woodall wrote:
> On Tue, 14 Mar 2023, Jeremy Ardley wrote:
>> I conclude there is no IPv6 DHCP involved but there must be something 
>> that listens to RA announcements and generates a MAC derived address 
>> from that.
>>
>> I still need to know how to control that as there are options 
>> relating to privacy that can do things like generate an expendable 
>> address in a different way and later generate another and deprecate 
>> earlier addresses.
>>
>>
>
> Just add them to the ipv6 section.
>
> I have
>
> iface xenbr0-active inet6 auto
>         up echo 64 
> >/proc/sys/net/ipv6/conf/$IFACE/accept_ra_rt_info_max_plen
>         up ip token set ::128/64 dev $IFACE
>         bridge_ports usb0
>         bridge_stp off       # disable Spanning Tree Protocol
>         bridge_waitport 0    # no delay before a port becomes available
>         bridge_fd 0          # no forwarding delay
>

I had hoped (not very hopefully) the solution wouldn't involve bashing 
kernel parameters.

 From previous journeys in this area systemd-networkd does a much neater 
job and has huge numbers of options.

This is my worstation config

root@client:/etc/systemd/network# cat wired_lan.network

[Match]
Name=enp7s0

[Network]

DHCP=yes

Address=10.31.40.68/24
Gateway=10.31.40.1
DNS=10.31.40.4 10.31.40.5

Address=2403:5800:c101:b700:beef::44/64
IPv6PrivacyExtensions=yes
IPv6AcceptRA=yes

Domains=lan example.com

[IPv6AcceptRA]


-- 


Jeremy

[toc] | [prev] | [next] | [standalone]

#256035

On Tue 14 Mar 2023 at 17:06:19 (+0800), jeremy ardley wrote:
> On 14/3/23 16:21, Tim Woodall wrote:
> > On Tue, 14 Mar 2023, Jeremy Ardley wrote:
> > > I conclude there is no IPv6 DHCP involved but there must be
> > > something that listens to RA announcements and generates a MAC
> > > derived address from that.
> > > 
> > > I still need to know how to control that as there are options
> > > relating to privacy that can do things like generate an
> > > expendable address in a different way and later generate
> > > another and deprecate earlier addresses.
> > 
> > Just add them to the ipv6 section.
> > 
> > I have
> > 
> > iface xenbr0-active inet6 auto
> >         up echo 64
> > >/proc/sys/net/ipv6/conf/$IFACE/accept_ra_rt_info_max_plen
> >         up ip token set ::128/64 dev $IFACE
> >         bridge_ports usb0
> >         bridge_stp off       # disable Spanning Tree Protocol
> >         bridge_waitport 0    # no delay before a port becomes available
> >         bridge_fd 0          # no forwarding delay

I didn't receive that post, and perhaps I wouldn't understand it
anyway until I'm running bookworm.

> I had hoped (not very hopefully) the solution wouldn't involve bashing
> kernel parameters.
> 
> From previous journeys in this area systemd-networkd does a much
> neater job and has huge numbers of options.
> 
> This is my worstation config

I hope it gets better :)

> root@client:/etc/systemd/network# cat wired_lan.network
> 
> [Match]
> Name=enp7s0
> 
> [Network]
> 
> DHCP=yes
> 
> Address=10.31.40.68/24
> Gateway=10.31.40.1
> DNS=10.31.40.4 10.31.40.5
> 
> Address=2403:5800:c101:b700:beef::44/64
> IPv6PrivacyExtensions=yes
> IPv6AcceptRA=yes
> 
> Domains=lan example.com
> 
> [IPv6AcceptRA]

Any reason for running ifupdown in preference to systemd-networkd
on the enp0s3 machine? All mine run the latter, hence being rusty
on the contents of /e/n/i.

Cheers,
David.

[toc] | [prev] | [next] | [standalone]

#256045

On 15/3/23 01:53, David Wright wrote:
>
>> I had hoped (not very hopefully) the solution wouldn't involve bashing
>> kernel parameters.
>>
>>  From previous journeys in this area systemd-networkd does a much
>> neater job and has huge numbers of options.
>>
>> This is my worstation config
> I hope it gets better :)
>
>> root@client:/etc/systemd/network# cat wired_lan.network
>>
>> [Match]
>> Name=enp7s0
>>
>> [Network]
>>
>> DHCP=yes
>>
>> Address=10.31.40.68/24
>> Gateway=10.31.40.1
>> DNS=10.31.40.4 10.31.40.5
>>
>> Address=2403:5800:c101:b700:beef::44/64
>> IPv6PrivacyExtensions=yes
>> IPv6AcceptRA=yes
>>
>> Domains=lan example.com
>>
>> [IPv6AcceptRA]
> Any reason for running ifupdown in preference to systemd-networkd
> on the enp0s3 machine? All mine run the latter, hence being rusty
> on the contents of /e/n/i.
>
systemd-networkd  is definitely the cleanest way to go. Hence I also 
use  it on all my systems

I am only looking at ifup/down as that is the default in bookworm for a 
non-graphical install.

I did a bit of experimentation with the installer and you don't get any 
options during the install to select networking method. You either get 
NetworkManager if you do a graphical install or ifup ifdown non graphical.

I looked at the if-up code in /etc/network/if-up.d/resolved

#!/bin/sh
#
# Script fragment to make ifupdown supply DNS information to resolved
#

I did an graphical install as a test and NetworkManager.service 
communicates to systemd-resolved by the internal message bus and I saw 
dhcp provided DNS being passed and used by systemd-resolved

Reading the documentation, systemd-resolved is tightly integrated with 
systemd-networkd as resolved reads all the systemd-networkd 
configuration files. The reverse may not be so true.

The situation now appears to be that networking.service (ifup/ifdown), 
NetworkManager.service, systemd-networkd.service all work (or can work) 
with systemd-resolved

systemd-networkd *should* be able to update systemd-resolved with DNS 
from DHCP but I haven't seen evidence it does so far

-- 
Jeremy
(Lists)

[toc] | [prev] | [next] | [standalone]

#256037

On Tue 14 Mar 2023 at 15:00:20 (+0800), Jeremy Ardley wrote:
> On 14/3/23 13:15, David Wright wrote:
> > 
> > > cat /etc.network/interfaces
> > > # This file describes the network interfaces available on your system
> > > # and how to activate them. For more information, see interfaces(5).
> > > 
> > > source /etc/network/interfaces.d/*
> > > 
> > > # The loopback network interface
> > > auto lo
> > > iface lo inet loopback
> > > 
> > > # The primary network interface
> > > allow-hotplug enp0s3
> > > iface enp0s3 inet dhcp
> > > # This is an autoconfigured IPv6 interface
> > > iface enp0s3 inet6 auto
> >                 ↑↑↑↑↑↑↑↑↑↑
> > 
> > I'm not very familiar with interfaces nowadays, but that looks
> > as if it's asking for enp0s3 to be autoconfigured with an IPv6
> > address. I thought that one got an IPv6 link address autoconfigured
> > anyway—I certainly do.
> > 
> > Did the debian-installer write that line (and comment)?
> 
> Yes. It was a result of the installation process identifying the
> interface and generating the necessary /etc/network/interfaces
> 
> > What's the output from:
> > 
> >    $ ip a
> > 
> > Mine (skipping lo):
> > 
> >    2: enp3s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
> >        link/ether a4:01:23:45:67:89 brd ff:ff:ff:ff:ff:ff
> >        inet 192.168.1.14/24 brd 192.168.1.255 scope global dynamic enp3s0
> >           valid_lft 74247sec preferred_lft 74247sec
> >        inet6 fe80::a601:23ff:fe45:6789/64 scope link      ← this here
> >           valid_lft forever preferred_lft forever
> > 
> root@debian12:/etc/network# 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: enp0s3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel
> state UP group default qlen 1000
>     link/ether 08:00:27:18:b6:ac brd ff:ff:ff:ff:ff:ff
>     inet 10.31.40.166/24 brd 10.31.40.255 scope global dynamic enp0s3
>        valid_lft 76395sec preferred_lft 76395sec
>     inet6 2403:5800:c101:b700:a00:27ff:fe18:b6ac/64 scope global
> dynamic mngtmpaddr
>        valid_lft 6277sec preferred_lft 2676sec
>     inet6 fe80::a00:27ff:fe18:b6ac/64 scope link
>        valid_lft forever preferred_lft forever
> 
> The difference to yours is  I have a routable IPv6 address in addition
> to the local link.
> 
> Your  one has  an autogenerated address
> 
> inet6 fe80::a601:23ff:fe45:6789/64
> 
> This is made as a function of your ethernet interface MAC address
> 
> I have an autogenerated address
> 
> inet6 fe80::a00:27ff:fe18:b6ac/64
> 
> and an address that must have been created as  a result of an RA or
> DHCP message as it is prefixed with my IPv6 address range
> 
> inet6 2403:5800:c101:b700:a00:27ff:fe18:b6ac/64
> 
> I note my two address  have the same tails so I guess my routable IPv6
> address is generated from the interface MAC address
> 
> I conclude there is no IPv6 DHCP involved but there must be something
> that listens to RA announcements and generates a MAC derived address
> from that.
> 
> I still need to know how to control that as there are options relating
> to privacy that can do things like generate an expendable address in a
> different way and later generate another and deprecate earlier
> addresses.

I get my link address without any configuration by me: nothing like
iface enp0s3 inet6 auto   is involved. So that line, unfamiliar to
me as a bullseye user, may be the cause of your routable address
problem. What happens if you take it out?

Cheers,
David.

[toc] | [prev] | [next] | [standalone]

#256031

On Tue, 14 Mar 2023 12:29:22 +0800
jeremy ardley <jeremy@ardley.org> wrote:

> (I'm quite annoyed they have done away with /var/log/syslog)

See the README in /var/log for work-arounds.


-- 
Does anybody read signatures any more?

https://charlescurley.com
https://charlescurley.com/blog/

[toc] | [prev] | [standalone]

Back to top | Article view | linux.debian.user

csiph-web