[OpenWrt-Devel] disappearing on-link route with RA
Gui Iribarren
gui at altermundi.net
Mon Sep 8 08:59:21 EDT 2014
Hello fine folks,
i'm testing bb-rc, compiled by myself from:
git://git.openwrt.org/14.07/openwrt.git at exactly this commit:
git-svn-id: svn://svn.openwrt.org/openwrt/branches/barrier_breaker@42398
so, a little bit more recent than the published bb-rc3 from
http://downloads.openwrt.org/barrier_breaker/14.07-rc3/
and i encounter a behaviour i'm already getting used to :P [0]
with network.iface.proto=dhcpv6, the first RA that arrives creates the
on-link route correctly (2001:db8::/64 on dev blah)
and the second RA, makes it disappear :(
leaving only the "default" route, and producing "triangular" routing
between nodes on the same link-local
you might not notice this immediately on a wired LAN, but on a
geographically widespread batman-adv mesh network like DeltaLibre, it's
the first thing i notice after sysupgrading, and ssh'ing into a node :)
root at palmeras:~# ip -6 r
default from :: via fe80::a2f3:c1ff:fe86:425 dev br-lan proto static
metric 1024
default from 2a00:1508:1:f004::/64 via fe80::a2f3:c1ff:fe86:425 dev
br-lan proto static metric 1024
2a00:1508:1:f004::/64 dev br-lan proto static metric 256
fddf:ca97:c78b::/64 dev br-lan proto kernel metric 256
fe80::/64 dev br-lan proto kernel metric 256
fe80::/64 dev eth0-167 proto kernel metric 256
fe80::/64 dev wlan0_ap proto kernel metric 256
fe80::/64 dev wlan0_adhoc proto kernel metric 256
fe80::/64 dev wlan0_adhoc-167 proto kernel metric 256
root at palmeras:~# rdisc6 br-lan
Soliciting ff02::2 (ff02::2) on br-lan...
Hop limit : 64 ( 0x40)
Stateful address conf. : No
Stateful other conf. : No
Mobile home agent : No
Router preference : medium
Neighbor discovery proxy : No
Router lifetime : 1800 (0x00000708) seconds
Reachable time : unspecified (0x00000000)
Retransmit time : unspecified (0x00000000)
Prefix : 2a00:1508:1:f004::/64
On-link : Yes
Autonomous address conf.: Yes
Valid time : 3600 (0x00000e10) seconds
Pref. time : 3600 (0x00000e10) seconds
MTU : 1500 bytes (valid)
Source link-layer address: A0:F3:C1:86:04:25
Recursive DNS server : 2a00:1508:1:f004::1
DNS server lifetime : 1200 (0x000004b0) seconds
from fe80::a2f3:c1ff:fe86:425
root at palmeras:~# ip -6 r
default from :: via fe80::a2f3:c1ff:fe86:425 dev br-lan proto static
metric 1024
default from 2a00:1508:1:f004::/64 via fe80::a2f3:c1ff:fe86:425 dev
br-lan proto static metric 1024
fddf:ca97:c78b::/64 dev br-lan proto kernel metric 256
fe80::/64 dev br-lan proto kernel metric 256
fe80::/64 dev eth0-167 proto kernel metric 256
fe80::/64 dev wlan0_ap proto kernel metric 256
fe80::/64 dev wlan0_adhoc proto kernel metric 256
fe80::/64 dev wlan0_adhoc-167 proto kernel metric 256
root at palmeras:~# opkg list odhcp6c netifd
odhcp6c - 2014-08-25-0300fe7589a1701361735ac068e4b57bb1a1896f
root at palmeras:~# opkg list netifd
netifd - 2014-08-31-b2dcb02570939d98b92c7c55db1c328693a5d52a
[0]: i recall encountering a similar regression twice already, in the
last couple of years or so, and in both cases it was promptly fixed by
cyrus: much respect for your impressive work on ipv6! /me always thankful :)
[0a] in fact, i was just looking for the last report
https://lists.openwrt.org/pipermail/openwrt-devel/2014-February/023640.html
and I don't recall specifically, but i guess it worked after that fix,
(or i would have noticed). Anyway i'm not sure anymore :$ sorry for not
following up that original thread, now it's gone from my memory
happy to debug this further, provide pcaps, or anything else needed
cheers! gui
_______________________________________________
openwrt-devel mailing list
openwrt-devel at lists.openwrt.org
https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
More information about the openwrt-devel
mailing list