IPV6 not routing after update to 23.1.5_4

[agh1701]

The simplest description is that i cannot pint the router using ipv6.  needless to say I cannot contact the internet via ipv6. 23.1.4 worked, other prior versions also had the same problem.

Code: [Select]

Pinging rtr.bs.net [xxxx:xxxx:xxxx:xxxx:201:2eff:fea3:a866] with 32 bytes of data:
Destination host unreachable.
Request timed out.
Destination host unreachable.
Destination host unreachable.

Ping statistics for xxxx:xxxx:xxxx:xxxx:201:2eff:fea3:a866:
    Packets: Sent = 4, Received = 0, Lost = 4 (100% loss),

[agh1701]

Update:
I can ping from opnsense to the internet with name resolution.

Code: [Select]

root@rtr:~ # ping -6 www.opnsense.org
PING6(56=40+8+8 bytes) 2607:fcc8:bfc0:30:bd23:9401:4a3e:5d78 --> 2001:1af8:4700:a1fa:3::2
16 bytes from 2001:1af8:4700:a1fa:3::2, icmp_seq=0 hlim=46 time=110.054 ms
16 bytes from 2001:1af8:4700:a1fa:3::2, icmp_seq=1 hlim=46 time=110.434 ms
^C
--- www.opnsense.org ping6 statistics ---
2 packets transmitted, 2 packets received, 0.0% packet loss
round-trip min/avg/max/std-dev = 110.054/110.244/110.434/0.190 ms
I cannot ping a lan address from opnsense.  there is something going on with the lan(re0) interface

Code: [Select]

root@rtr:~ # ping xxxx:xxxx:xxxx:xxxx::148
ping: UDP connect: Network is unreachable

[gpb]

Had a similar problem after update.  Typically, I can resave the WAN interface and it corrects itself.  This time it didn't work.  I then tried restarting radvd and that didn't *seem* to have an effect either.  However, after a few minutes, it then just started working.  I don't know what's going on.  My setup upgraded fine without issue, but a similar server failed to have ipv6 upon update.

[agh1701]

Summery of problem

Ping for client:
   to router LAN RE(0) 2603:: address      No
   to router LAN RE(0) fe80:: address      Yes
   to any other client Lan 2603:: Address         Yes
   to any internet ipv6 address         no
   
Ping from router:
   to router LAN RE(0) 2603:: address      No
   to router LAN RE(0) fe80:: address      Yes
   to any client LAN 2603::            No   ping: UDP connect: Network is unreachable
   to any internet ipv6 address         Yes   with DNS name resolution

The router WAN RE(1) is fully functional.
The router LAN RE(0) ISP assigned address is not honored.

Attached log with errors

[agh1701]

I filed a bug report.

https://github.com/opnsense/core/issues/6475

[sot3]

I have joined you there.  This is not a new problem but recently it's become a higher priority for me.

[sot3]

I just tried gpb's suggestion to re-save the WAN interface again, and this time it worked.  I don't know what that did or why it worked today but not 3 days ago.

[agh1701]

i have found this to be hit or miss.  have you rebooted?  I am staying on .4 for now.

[sot3]

Yes, after IPv6 was working I installed 23.1.6 and rebooted, and it was still working afterwards.

[lucaberta]

Thanks @agh1701 and @sot3.

IPv6 has been quite unstable for me too, and this time an upgrade to 23.1.6 broke things for good.

I would normally be able to make IPv6 work again after a reboot by running:

/usr/local/etc/rc.configure_interface wan

from the FreeBSD shell as root, but this time it did not work out.

I reverted back to 23.1.4 as hinted by @AdSchellevis on the GitHub bug thread by using:

opnsense-revert -r  23.1.4 opnsense

and without rebooting, IPv6 came back to life after giving the reconfigure_interface wan command listed above.

I will NOT update my OPNsense box any further, as IPv6 is really unstable at this time, and I'd rather have it than not.

Many thanks,

Luca

[ProximusAl]

Interesting, I've updated to 23.1.6 and just for interest went to test my IPv6, and it's not working either.

It definitely was.

All my IPv6 is statically assigned.

I'll go reboot some things to see what happens.

[ProximusAl]

Turns out my IPv6 on OPNSense was working fine.

Stoopid UDM Pro downstream of it that's borked.

[gpb]

Stoopid UDM Pro downstream of it that's borked.

A while back, I had an ipv6 issue that was caused by a specific model TP-Link AP (Omada) bug that I attributed to opnsense/freebsd.  Spent a day on it...neglected to run a test scenario using a wired device...or different AP.  It happens.