2Degrees, Fritz!Box, IPv6 & Linux: Won't route.

Forums › 2degrees (including Slingshot, Orcon, Flip, Stuff Fibre, MyRepublic, 2talk and Vocus) › 2Degrees, Fritz!Box, IPv6 & Linux: Won't route.

HamishMacEwan

28 posts

Geek

Trusted

#239355 13-Jul-2018 11:40

Hi,

Switched over my fibre RSP (Retail Service Provider) from Bigpipe,200/20,MikroTik RB850Gx2 to 2Degrees Broadband,Max/max, FRITZ!Box 7560. I changed the router due to reachability issues for some WWW sites. And its always easier to diagnose issues if you use their issued router. I have turned off the FRITZ!Box DHCP server, Wi-Fi AP, and renumbered it to 192.168.1.249 because the home network is not exactly greenfields...

Things have gone swimmingly for the Window 10 box, but Linux (3.16.0-5-amd64 #1 SMP Debian 3.16.51) isn't faring so well.

Addresses and routes are appearing but packets aren't flowing:

ip -6 addr dev eth1

2: eth1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qlen 1000
inet6 2406:e006:28a6:0:960c:6dff:fe80:61b/64 scope global noprefixroute dynamic
valid_lft 2528064sec preferred_lft 540864sec
inet6 2406:e006:1c6:1:960c:6dff:fe80:61b/64 scope global noprefixroute dynamic
valid_lft 7110sec preferred_lft 3510sec
inet6 fe80::960c:6dff:fe80:61b/64 scope link
valid_lft forever preferred_lft forever

ip -6 route list dev eth1

2406:e006:1c6:1::/64 proto ra metric 1
2406:e006:1c6::/48 via fe80::eadf:70ff:fe31:3452 proto ra metric 1
2406:e006:28a6::/64 proto ra metric 1
fe80::/64 proto kernel metric 256 mtu 1492
default via fe80::eadf:70ff:fe31:3452 proto static metric 1024

I can reach the default next hop:

ping6 fe80::eadf:70ff:fe31:3452%eth1
PING fe80::eadf:70ff:fe31:3452%eth1(fe80::eadf:70ff:fe31:3452) 56 data bytes
64 bytes from fe80::eadf:70ff:fe31:3452: icmp_seq=1 ttl=64 time=0.997 ms
64 bytes from fe80::eadf:70ff:fe31:3452: icmp_seq=2 ttl=64 time=0.676 ms
64 bytes from fe80::eadf:70ff:fe31:3452: icmp_seq=3 ttl=64 time=0.752 ms
64 bytes from fe80::eadf:70ff:fe31:3452: icmp_seq=4 ttl=64 time=0.719 ms

But traceroute6 doesn't get very far:

traceroute6 -n 2001:8b0:0:30::666:102
traceroute to 2001:8b0:0:30::666:102 (2001:8b0:0:30::666:102) from 2406:e006:28a6:0:960c:6dff:fe80:61b, 30 hops max, 24 byte packets
1 * * *
2 * * *
3 * * *

etc. etc.

So Windows 10 is working just fine in exactly the same network, but Debian (perhaps due to greater compliance or misconfiguration in the past) is struggling in a way that it didn't when using the MikroTik.

Any suggestions or requests for additional information gratefully received.

Hamish.

http://protopage.com/Hamish.MacEwan

stinger

628 posts

Ultimate Geek
Inactive user

#2055719 13-Jul-2018 13:33

Not that it will help you., but IPv6 worked out of the box for my 2 degrees connection (with a Fritzbox 7490). A combination of pi3, Ubuntu Server and Fedora machines. What OS are you using on the Linux box?

Quic Broadband

Move to New Zealand's best fibre broadband service (affiliate link). Note that to use Quic Broadband you must be comfortable with configuring your own router.

NickMack

948 posts

Ultimate Geek

Trusted

Lifetime subscriber

#2055754 13-Jul-2018 13:45

@michaelmurfy

https://nick.mackechnie.co.nz | NZ ISP latency monitoring - https://smokeping.thenet.gen.nz

michaelmurfy

meow
13216 posts

Uber Geek

Moderator

ID Verified

Trusted

Lifetime subscriber

#2055857 13-Jul-2018 15:05

Based on that kernel version it appears you're using Debian Jessie? I've just tested this on Stretch and it is working as intended:

:~$ ip -6 route list dev enp1s0
2406:e001:1:df01::/64 proto kernel metric 256 expires 86001sec pref medium
fe80::/64 proto kernel metric 256 pref medium
default via fe80::7a8a:20ff:fe7d:6557 proto ra metric 1024 expires 1401sec hoplimit 64 pref medium

:~$ ip -6 route list dev enp1s0
2406:e001:1:df01::/64 proto kernel metric 256 expires 85956sec pref medium
fe80::/64 proto kernel metric 256 pref medium
default via fe80::7a8a:20ff:fe7d:6557 proto ra metric 1024 expires 1356sec hoplimit 64 pref medium

:~$ traceroute6 -n 2001:8b0:0:30::666:102
traceroute to 2001:8b0:0:30::666:102 (2001:8b0:0:30::666:102), 30 hops max, 80 byte packets
1 2406:e001:1:df01::1 0.621 ms 0.582 ms 0.550 ms
2 2406:e000:2801:c::1 1.080 ms 1.061 ms 1.015 ms
3 * * *
4 2001:1900:2100::20e 166.342 ms 166.091 ms 166.471 ms
5 2001:1900:2100::20d 166.682 ms 165.863 ms 166.078 ms
6 2001:1900::4:6 285.263 ms 284.669 ms 285.228 ms
7 2001:8b0:0:53::48 295.503 ms 294.971 ms *
8 2001:8b0:0:89a::1 294.142 ms 294.466 ms 293.654 ms
9 2001:8b0:0:30::51bb:1e02 296.416 ms 295.569 ms 295.427 ms
10 2001:8b0:0:30::666:102 295.017 ms 295.127 ms 296.362 ms

Try this on a more recent distro like Stretch or even boot to Ubuntu 18.04 (Live is fine). It appears that your host may have a broken /etc/dhcpcd.conf configuration. Here is the default (working) configuration for Stretch:

# A sample configuration for dhcpcd.
# See dhcpcd.conf(5) for details.

# Allow users of this group to interact with dhcpcd via the control socket.
#controlgroup wheel

# Inform the DHCP server of our hostname for DDNS.
hostname

# Use the hardware address of the interface for the Client ID.
#clientid
# or
# Use the same DUID + IAID as set in DHCPv6 for DHCPv4 ClientID as per RFC4361.
# Some non-RFC compliant DHCP servers do not reply with this set.
# In this case, comment out duid and enable clientid above.
duid

# Persist interface configuration when dhcpcd exits.
persistent

# Rapid commit support.
# Safe to enable by default because it requires the equivalent option set
# on the server to actually work.
option rapid_commit

# A list of options to request from the DHCP server.
option domain_name_servers, domain_name, domain_search, host_name
option classless_static_routes
# Most distributions have NTP support.
option ntp_servers
# Respect the network MTU. This is applied to DHCP routes.
option interface_mtu

# A ServerID is required by RFC2131.
require dhcp_server_identifier

# Generate Stable Private IPv6 Addresses instead of hardware based ones
slaac private

Michael Murphy | https://murfy.nz
Referral Links: Quic Broadband (use R122101E7CV7Q for free setup)

Are you happy with what you get from Geekzone? Please consider supporting us by subscribing.
Opinions are my own and not the views of my employer.

HamishMacEwan

28 posts

Geek

Trusted

#2055897 13-Jul-2018 16:30

Thank you both for the responses.

michaelmurfy:

Based on that kernel version it appears you're using Debian Jessie?

Yes.

I'll digest the suggestions and see what they spur.

Can you, if its brief, tell me how you got your "$ip -6 route list dev enp1s0" to show hop count and pref in my version or is it a Stretch thing only?

My Jessie doesn't appear to have an /etc/dhcpd.conf which may be extremely broken, but as noted, the problems occurred only since the Fritz!Box, it was all good with the MikroTik.

Hamish.

http://protopage.com/Hamish.MacEwan

visentens

7 posts

Wannabe Geek

#2055905 13-Jul-2018 17:15

Is it a fresh install? could maybe spin up a virtual machine of your distro to see if a fresh install faces the same problems, and if not then swap out the default config files till you find the culprit.

michaelmurfy

meow
13216 posts

Uber Geek

Moderator

ID Verified

Trusted

Lifetime subscriber

#2055999 13-Jul-2018 22:12

HamishMacEwan:

Can you, if its brief, tell me how you got your "$ip -6 route list dev enp1s0" to show hop count and pref in my version or is it a Stretch thing only?

My Jessie doesn't appear to have an /etc/dhcpd.conf which may be extremely broken, but as noted, the problems occurred only since the Fritz!Box, it was all good with the MikroTik.

Yes - there has been some improvements in Stretch in regards to networking.

My Desktop machine (running Solus Linux - not Debian based) has the following:

It doesn't use the dhcpd like Debian Stretch and instead just uses good ole Network Manager. If I was going to take a punt in what the issue is I'd say you've got a broken dhcpd configuration. I would honestly recommend doing a "apt-get dist-upgrade" on that box and bring it from legacy to stable.

HamishMacEwan

28 posts

Geek

Trusted

#2056040 14-Jul-2018 07:40

michaelmurfy:

Try this on a more recent distro

From a couple of other linux distributions that worked in the past.

Distributor ID: LinuxMint
Description: Linux Mint 18 Sarah
Release: 18
Codename: sarah

ip -6 route list dev wlp3s0

2406:e006:1c6:1::/64 proto ra metric 600 pref medium
2406:e006:1c6::/48 via fe80::eadf:70ff:fe31:3452 proto ra metric 600 pref medium
2406:e006:f6c::/64 proto kernel metric 256 expires 518315sec mtu 1492 pref medium
2406:e006:1260::/64 proto kernel metric 256 expires 520531sec mtu 1492 pref medium
2406:e006:1555::/64 proto kernel metric 256 expires 466155sec mtu 1492 pref medium
2406:e006:194a::/64 proto kernel metric 256 expires 522743sec mtu 1492 pref medium
2406:e006:1ac0::/64 proto kernel metric 256 expires 530058sec mtu 1492 pref medium
2406:e006:28a6::/64 proto kernel metric 256 expires 543085sec mtu 1492 pref medium
2406:e006:2acf::/64 proto kernel metric 256 expires 525513sec mtu 1492 pref medium
2406:e006:2dc6::/64 proto kernel metric 256 expires 475809sec mtu 1492 pref medium
2406:e006:375c::/64 proto kernel metric 256 expires 519774sec mtu 1492 pref medium
2406:e006:3ceb::/64 proto kernel metric 256 expires 473590sec mtu 1492 pref medium
fe80::/64 proto kernel metric 256 mtu 1492 pref medium
default via fe80::eadf:70ff:fe31:3452 proto static metric 600 pref medium

traceroute6 -n 2001:8b0:0:30::666:102
traceroute to 2001:8b0:0:30::666:102 (2001:8b0:0:30::666:102) from 2406:e006:f6c:0:711a:bdbb:44d0:3812, 30 hops max, 24 byte packets
1 *^C

Distributor ID: Raspbian
Description: Raspbian GNU/Linux 8.0 (jessie)
Release: 8.0
Codename: jessie

ip -6 addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever

No eth0 -6 addr at all.

ip -6 route list dev eth0
2403:9800:c048:a10::/64 proto kernel metric 202
2406:e006:1c6:1::/64 proto kernel metric 202 mtu 1492
2406:e006:f6c::/64 proto kernel metric 202
2406:e006:1260::/64 proto kernel metric 202
2406:e006:1555::/64 proto kernel metric 202
2406:e006:194a::/64 proto kernel metric 202
2406:e006:1ac0::/64 proto kernel metric 202
2406:e006:28a6::/64 proto kernel metric 202
2406:e006:2acf::/64 proto kernel metric 202
2406:e006:2dc6::/64 proto kernel metric 202
2406:e006:375c::/64 proto kernel metric 202
2406:e006:3ceb::/64 proto kernel metric 202
fd00::/64 proto kernel metric 202 mtu 1492
default via fe80::e68d:8cff:fe26:a1ee metric 202

Still retaining the default from the MikroTik (fe80::e68d:8cff:fe26:a1ee) and the new route doesn't appear.

traceroute6 -n 2001:8b0:0:30::666:102
traceroute to 2001:8b0:0:30::666:102 (2001:8b0:0:30::666:102), 30 hops max, 80 byte packets
1 * * *
2 * * *
3 * * *
4 * * *
5 * * *
6 *^C

However this distribution is the only one I have that has /etc/dhcpcd.conf and it is at a glance identical to the one you provided.

It is possibly worth repeating that all of these hosts worked with the Bigpipe/MikroTik, if not out of the box, after I had, IIRC, corrected the hop limit in Neighbor Discovery to 63 from 0.

http://protopage.com/Hamish.MacEwan

michaelmurfy

meow
13216 posts

Uber Geek

Moderator

ID Verified

Trusted

Lifetime subscriber

#2056058 14-Jul-2018 09:38

Interesting... Your router is at fault here.

I don't have a Fritz!Box to test with (didn't opt in to the $15 Fritz when I signed up) so not sure what else to add here.

HamishMacEwan

28 posts

Geek

Trusted

#2056085 14-Jul-2018 09:53

michaelmurfy:

Interesting... Your router is at fault here.

I don't have a Fritz!Box to test with (didn't opt in to the $15 Fritz when I signed up) so not sure what else to add here.

Thank you very much for your endeavours Michael, much appreciated.

I suspect its to do with the shambles my network is, being the accretion of years of routers, modems, wi-fi access points, switches, VLAN10, NAT, port mapping and the like. Plus with more parts of increasing smarts I understand them less and clever v6 interactions via work-arounds, developments and adaptations are what is killing this.

Dumb ole v4 crashes through (more or less, there are issues), well tested Microsoft OS survives, but the installed base and the new stuff is not playing well together.

Might be time to torch the lot and get one box that I do understand that does enough.

If some identifiable thing solves this, I'll be back to report.

Hamish.

http://protopage.com/Hamish.MacEwan

HamishMacEwan

28 posts

Geek

Trusted

#2056118 14-Jul-2018 11:21

For completeness:

C:\Users\Hamish>ipconfig

Windows IP Configuration

Ethernet adapter LAN: (In Settings|Network & Internet this is called "Network 2")

Connection-specific DNS Suffix . :
IPv6 Address. . . . . . . . . . . : 2406:e006:1c6:1:7061:2e2f:aef9:b2c9
Temporary IPv6 Address. . . . . . : 2406:e006:1c6:1:55b3:1ea7:df76:40c5
Link-local IPv6 Address . . . . . : fe80::7061:2e2f:aef9:b2c9%7
IPv4 Address. . . . . . . . . . . : 192.168.1.69
Subnet Mask . . . . . . . . . . . : 255.255.255.0
Default Gateway . . . . . . . . . : fe80::eadf:70ff:fe31:3452%7
192.168.1.249

But I have to configure the IPv4 manually, if I don't and ask for automatic (DHCP):

Ethernet adapter LAN:

Connection-specific DNS Suffix . : macewan.nz
IPv6 Address. . . . . . . . . . . : 2406:e006:1c6:1:7061:2e2f:aef9:b2c9
Temporary IPv6 Address. . . . . . : 2406:e006:1c6:1:55b3:1ea7:df76:40c5
Link-local IPv6 Address . . . . . : fe80::7061:2e2f:aef9:b2c9%7
IPv4 Address. . . . . . . . . . . : 192.168.1.69
Subnet Mask . . . . . . . . . . . : 255.255.255.0
Default Gateway . . . . . . . . . : fe80::eadf:70ff:fe31:3452%7
192.168.1.251

The IPv4 default gateway attracts the old value (192.168.1.251), from where, IHNFI.

But IPv6,

http://protopage.com/Hamish.MacEwan

HamishMacEwan

28 posts

Geek

Trusted

#2056614 15-Jul-2018 19:12

HamishMacEwan:

The IPv4 default gateway attracts the old value (192.168.1.251), from where, IHNFI.

As noted the LAN is a shambles of bits and pieces accumulated over time as interest, need and resources allowed. I got a Netgate SG-1000, on special and played around with it for a while.

It seemed like a nice, reliable home for the DHCP Server on the LAN after I stopped using it as a fully operational router.

That's where I changed the default router for DHCP requests from .251 to .249

But that didn't change the behaviour above. Then I discovered:

The LAN interface static IPv4 configuration gives the option of an IPv4 upstream gateway on the LAN, useful for downloading updates to the device without a WAN Internet connection.

But perhaps this was over-riding the DHCP server gateway setting?

Actually no. I belatedly discovered "DHCP Static Mappings for this Interface" which provides a wide range of bespoke values for individual devices, including... gateway addresses, and sure enough, that's where the feral .251 was.

Leave it blank, and it picks up the general DHCP gateway value. Success.

As for the IPv6 issue, I plugged Jimmy directly into the Fritz!Box, replacing the cable at the same time, and shortly afterward:

jimmy:~$ traceroute6 loopsofzen.uk
traceroute to loopsofzen.uk (2001:8b0:0:30::666:102) from 2406:e006:1c6:1:960c:6dff:fe80:61b, 30 hops max, 24 byte packets
1 2406:e006:1c6:1:eadf:70ff:fe31:3452 (2406:e006:1c6:1:eadf:70ff:fe31:3452) 1.693 ms 0.988 ms 1.137 ms
2 2406:e000:2801:c::1 (2406:e000:2801:c::1) 2.929 ms 1.047 ms 1.383 ms
3 * * *
4 ge-6-13-147.car1.StLouis1.Level3.net (2001:1900:2100::20e) 165.898 ms 165.89 ms 165.923 ms
5 xe-7-1-3.bar1.Toronto1.Level3.net (2001:1900:2100::20d) 166.772 ms 168.209 ms 166.734 ms
6 lo-0-1-v6.edge4.London1.Level3.net (2001:1900::4:6) 285.455 ms 285.752 ms 286.839 ms
7 h.aimless.thn.aa.net.uk (2001:8b0:0:53::48) 296.725 ms 294.944 ms 294.465 ms
8 a.needless.mh.aa.net.uk (2001:8b0:0:89a::1) 297.477 ms 295.105 ms 294.637 ms
9 a.homeless.aa.net.uk (2001:8b0:0:30::51bb:1e02) 296.059 ms 296.139 ms 303.872 ms
10 loopsofzen.uk (2001:8b0:0:30::666:102) 298.493 ms 297.001 ms 297.626 ms

http://protopage.com/Hamish.MacEwan

michaelmurfy

meow
13216 posts

Uber Geek

Moderator

ID Verified

Trusted

Lifetime subscriber

#2056686 15-Jul-2018 20:44

Hah! All this time was a simple Pebcak.

Simpler is better. Maybe look at investing in an Edgerouter or Ubiquiti USG along with some access points to clean up your network. Glad it is working now!