I saw the same today.  About five hours of packet loss starting at 00:04 PDT to about 05:08 PDT.  Seems to be working fine now.

So I have two IPv6 providers - my ISP gives me native IPv6 (but with some limitations such as no reverse DNS, and DHCP prefix delegation which is not guaranteed not to change), and he.net gives me an IPv6 tunnel. I use both on different segments of my home network - native on the guest lan, tunnel where I run most of my (small) servers.

Today the tunnel is having some issues, and it got me to think - in a v4 context, I could very easily switch over to my other provider, by just doing NAT on the router until things go back to normal. But with IPv6, are there any quick solutions when one of the providers unexpectedly goes down ? switching to the other provider isn't as easy because it involves every machine on the lan having to get new addresses somehow. Doable in a matter of hours, but not as a quick switch-over.

I am wondering, are there any practical solutions to this ?

Hi,

My IPv6 tunnel usualy has great performance, but today it is severely degraded to the point of being almost unusable - unusual latency of about 60ms (normally about 5ms), and 30% to 40% packet loss.

Looking into it, it seems that the IPv4 packets are bottlenecking somewhere on HE's network, as they try to go from my router (on att's network) to the tunnel endpoint (72.52.104.74 aka fmt2)

The last hop I can reach with good performance is 10ge4-4.core1.sjc1.he.net ; the next hop 100ge13-2.core4.fmt2.he.net is getting lag and packet loss.

Just reporting the issue; thanks again for the amazing service! it's in those times where it's failing me that I realize how much I've come to depend on it...

Well I switched to bridge mode on the modem but no progress... (still trying to set up a tunnel because Verizon doesn't support IPv6 in bridge mode)

When I set up the tunnel, add an IP, create a default route, I'm able to get a SINGLE ping response packet from google.com, and that's it... stops at that point, and then running another ping gets nothing:

Code: [Select]

# ping6 google.com
PING google.com (2607:f8b0:4023:1004::66): 56 data bytes
64 bytes from 2607:f8b0:4023:1004::66: seq=0 ttl=113 time=20.334 ms
^C
--- google.com ping statistics ---
7 packets transmitted, 1 packets received, 85% packet loss
round-trip min/avg/max = 20.334/20.334/20.334 ms

I'm really not sure what issue I'm running into here...

I would assume that the OpenWrt linux kernel is multi-core supporting. But maybe the support of the SoC is not fully multi-core supporting? But I think other operations work with multi-threading on this unit. How would I check out this interrupt thing?

@hobbsfamily
Doesn't solve missing prefix delegation, but maybe it gets you to where you want to be: Have you tried setting up your router in switch mode instead of as a router?

I have 'figured out' how to put their modem in bridge mode, however it does seem to have some negative consequences, so I fell back to having the router be the only device connected and set it up in the DMZ... I think I will have much better luck with it in bridge mode though... I have to experiment to see if protocol 41 gets through (Verizon's bridge mode is a bit kludgy)

Thanks!

We don't offer 6in6 tunnels with tunnelbroker.net.

Also Verizon doesn't learn route announcements from their residential customers, nor do we issue permission for broker users to have others announce their allocations from our system. So you can't use our IPv6 address space over your native VZ connection, they only work over the tunnel.

Thanks @broquea! That makes sense.

Is your OS kernel multi-core supporting?  Maybe the board routes all interruptes to core zero in hardware and that can’t be changed.  Just things to check out.

@hobbsfamily
Doesn't solve missing prefix delegation, but maybe it gets you to where you want to be: Have you tried setting up your router in switch mode instead of as a router?

I am terminating my tunnel with OpenWrt running on a cheapish router, ARM-based (MediaTek).

I've found that I get quite poor performance of only 300 Mbps max from my 1 Gbps legacy connection. If I run htop I see that only one core is getting all the load (though it shows 4 cores, while the specs actually say it's just dual core?).

So I was wondering if the sit/6in4 kernel module maybe doesn't support multithreading? Does anybody know more about it?