Modify

Opened 4 years ago

Closed 4 years ago

#14958 closed defect (fixed)

IPv6-in IPv4 stopped working

Reported by: lbthomsen Owned by: developers
Priority: normal Milestone: Chaos Calmer 15.05
Component: base system Version: Trunk
Keywords: IPv6 Cc:

Description

The last couple of weeks I have been struggling a bit with my WDR4300 router. I am keeping it reasonably up-to-date with builds from trunk every couple of weeks or so. A few weeks back the router never came up after a flash and I had to go through fail-safe to get it up and running again.

When configuring it again what goes wrong is when I ad my Hurricane Electric tunnel (IPv6 in IPv4). It is a little hard to debug because the moment I add the tunnel I get locked out completely, so I assume it might be the firewall that does not get loaded.

The following lines in /etc/config/network:

config interface 'he'
        option proto '6in4'
        option peeraddr '216.218.221.42'
        option ip6addr '2001:470:35:7fb::2'
        option ip6prefix '2001:470:36:7fb::/64'

Is enough to ensure the router - otherwise fully functional - is unreachable.

Now, I know this is a shitty description, but without actually soldering a console port on my router, I am a bit at a loss how actually to debug this problem. Any hints would be much appreciated.

Attachments (0)

Change History (5)

comment:1 Changed 4 years ago by lbthomsen

I should add that it is quite easy for me to reboot in fail-safe, remove the offending lines and get the router back up, so if anybody got any suggestions I can test it quickly.

comment:2 Changed 4 years ago by lbthomsen

I should probably ALSO add that my ISP since I started running the HE tunnel, have in fact started providing dynamic IPv6 using 6to4 and that works fine.

comment:3 Changed 4 years ago by hnyman

Sounds a bit similar as #14858

The reason might be a missing /64 after the ip6addr parameter. (Maybe odchpd requires it, while 6relayd did not? or something like that has recently changed in netifd.)

config interface 'sixxs'

option proto '6in4'
option mtu '1424'
option peeraddr '62.78.96.38'
option ip6addr '2001:tunnel:only:prefix::2/64'
option ip6prefix '2001:public:prefix::/48'

comment:4 Changed 4 years ago by lbthomsen

Ok, now it seems to turn into a proper bug report :) Last comment actually managed to fix it - adding /64 to the local ip6addr made the tunnel start up and the system remained reachable.

However, this IS quite a serious bug since an upgrade will render the system unreachable, AND luci got the following explanation right under the address field:

This is the local endpoint address assigned by the tunnel broker, it usually ends with :2

Having to end it with the netmask /64 is not :2

comment:5 Changed 4 years ago by cyrus

  • Resolution set to fixed
  • Status changed from new to closed

Fixed in r39586.

Add Comment

Modify Ticket

Action
as closed .
The resolution will be deleted. Next status will be 'reopened'.
Author


E-mail address and user name can be saved in the Preferences.

 
Note: See TracTickets for help on using tickets.