And now, for your regularly scheduled dose of broken ipv6
cfriacas at fccn.pt
Fri Sep 22 10:16:55 CEST 2006
Yes. IPv6 routing is still a big piece of <you know what!>...
IMHO, current status will only change when people start to put their
money in front of their mouth. As a customer, if you don't care about IPv6
routing as much as you care for IPv4 routing, the status quo will not
One other thing that should help in my view would be shutting down all the
services that freely provide IPv6 transit (and mostly with a dramatic
performance associated with it!). This "availability" not only drives
people to think that IPv6 is still a toy for some, and not a production
driven service. And of course, this fact also doesn't encourage the
biggest transit providers to go into IPv6.
If one notices that a given route is unreachable (or with a very bad
end-to-end performance), the normal behaviour would be complaining
to your upstream provider(s). And how many times that happens when the
routing issue is exclusively related with IPv6??? :-(((
At this point, and on the given circumstances, i'm even surprised to see
how many transit providers have the IPv6 option in their services'
Recently i've started to measure v4 and v6 latency to the same host (or to
the same name with both A and AAAA records) and found that it's really,
really hard to find names with both A and AAAA in some countries (and yes,
*even* in the academic communities...)
For what i've seen so far, the measurements i get with a smaller
difference than 1ms between v4 & v6 are from other NRENs (to be more
accurate... 13 of them to the date). Could be better, could be worse...
./Carlos Skype: cf916183694
Wide Area Network (WAN) Workgroup, CMF8-RIPE, CF596-ARIN
FCCN - Fundacao para a Computacao Cientifica Nacional http://www.fccn.pt
"Internet is just routes (196663/675), naming (millions) and... people!"
On Thu, 21 Sep 2006, Jeroen Massar wrote:
> Nick Hilliard wrote:
>> Oh my!
>> At least the pktloss is down from an irritating 93% to a more acceptable 16%.
>> pancake:/home/nick> traceroute6 www.ripe.net
>> traceroute6 to kite-www.ripe.net (2001:610:240:0:a843::8) from 2001:bb0:ccc0:1::44, 64 hops max, 12 byte packets
>> 1 2001:bb0:ccc0:1::1 0.336 ms 0.163 ms 0.128 ms
>> 2 pe-tun200.ipv6.cra.dublin.eircom.net 44.168 ms 41.373 ms 43.009 ms
>> 3 sl-bb1v6-bru-t-8.sprintv6.net 62.928 ms 58.672 ms 59.848 ms
>> 4 sl-bb1v6-rly-t-1001.sprintv6.net 154.586 ms 159.475 ms 156.289 ms
>> 5 2001:440:eeee:ffc8::2 232.806 ms 232.858 ms 231.715 ms
>> 6 2001:450:2001:1000:0:670:1708:219 233.530 ms 310.736 ms 231.747 ms
>> 7 paix.ipv6.he.net 233.496 ms 235.718 ms 237.707 ms
>> 8 3ffe:81d0:ffff:1:: 232.047 ms 229.174 ms 232.789 ms
> Wouldn't you really quickly kick sprint to NOT use HE.net as a transit?
> Clearly HE.net still lives in the times of 6bone age.
> Also why doesn't sprint have a router in Amsterdam? That would provide
> them with the route to RIPE and a lot of other good IPv6 networks.
> As you are a customer, ask them ;)
> (There is one side-effect explanation btw: filters, as the RIPE IPv6
> block is announced as a /48, which is filtered by some, and accepted
> only by the ones who can't care less. Thus quick solution: drop the /48
> and simply use the /32 towards SURFNet for super light speed connectivity ;)
More information about the ipv6-ops