<html><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; "><br><div><blockquote type="cite"><blockquote type="cite"><font class="Apple-style-span" color="#000000"><br></font><br></blockquote><blockquote type="cite">On our end, this is what we've got:<br></blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite">Dedicated box doing nothing other than 6to4. It's a dual P3 866 Xeon,<br></blockquote><blockquote type="cite">and it's pretty much got 99% idle time on it.<br></blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite">This box uses Quagga to announce 192.88.99.0/24 and 2002::/16 to our<br></blockquote><blockquote type="cite">core router. This way if the box dies, our announcements get withdrawn.<br></blockquote>A quagga conf file (example or real) would be helpful to look at...<br><br></blockquote><div><br></div><div>It's pretty simple really:</div><div><br></div><div><div>router bgp 19255</div><div> bgp router-id 192.88.99.1</div><div> network 192.88.99.0/24</div><div> neighbor 216.14.98.5 remote-as 19255</div><div> neighbor 216.14.98.5 update-source 216.14.98.26</div><div> neighbor 2001:4978:1:410::ffff remote-as 19255</div><div> no neighbor 2001:4978:1:410::ffff activate</div><div>!</div><div> address-family ipv6</div><div> network 2002::/16</div><div> neighbor 2001:4978:1:410::ffff activate</div><div> exit-address-family</div><div><div>!</div><div><br></div><div>216.14.98.5 is our core router's v4 address for this link. 2001:4978:1:410::ffff is it's v6 address.</div><div><br></div><div><br></div></div></div><blockquote type="cite">I note that everybody doing this seems utterly reliant on BGP, in terms<br>of distributing the anycast address to the world. In the inside the<br>smaller (wireless) ISP case, BGP is not in use. I wonder what will<br>happen (router trafficwise) if I use another protocol... or don't use<br>one at all. what additional traffic would dns udp over ipv6 generate...<br><br></blockquote><div><br></div><div>There's nothing special about BGP, it's just what a lot of us are using elsewhere, so it's easier to reuse that infrastructure than run another routing protocol just for this. Nothing says you can't use OSPF or IS-IS or even static routes that I'm aware of. You just need to concern yourself with BGP if you want people outside your network to be able to use your relay.</div><div><br></div><div>If you're just doing this for purely internal use and don't care if you break connectivity to your own clients if your 6to4 relay goes down, you can just static route all of it. </div><br><blockquote type="cite">I assume you are doing BGP announcements to the core router from quagga.<br><br>Similarly, I assume your core router filters out bogus announcements of<br>other 6to4 routers (for example, someone as crazy as I am, inside your<br>network, mistakenly announcing they have 6to4 with a better metric than<br>you do...<br><br></blockquote><div><br></div><div>Yep, prefix filters on every BGP session will make sure we're only accept what we're expecting from anyone we have a BGP session with. BGP sessions have to be explicitly configured though, so there's not much chance of a random user setting up a BGP router and breaking our network.</div><div><br></div><blockquote type="cite"><blockquote type="cite">The occasional burst over 100mbps often enough to justify a GigE port<br></blockquote><blockquote type="cite">for this, but average use for our Chicago relay is less than 10mbps.<br></blockquote><blockquote type="cite">Average use for our Amsterdam relay is less than 30mbps.<br></blockquote>I look forward to the results of your ipv6experiment! :)</blockquote><div><br></div>It's coming.. I know i've been saying that for months, but it's really almost ready now. :)</div><div><br></div><div>-- Kevin</div><div><br></div></body></html>