Connectivity to Latin American NRENs
JORDI PALET MARTINEZ
jordi.palet at consulintel.es
Thu Oct 11 09:49:10 CEST 2007
I've raised this several times in the IPv6 working group of CLARA, the need
for commercial peering. I've copied that WG, hoping that a voice different
than mine can help to convince them.
Totally agree with your points. Not having good commercial peering means
they become unreachable and this is against the CLARA objectives !
I've even offered long time ago free transit because some tier ones offered
to me for CLARA, but nothing happened.
Regards,
Jordi
> De: Bernhard Schmidt <berni at birkenwald.de>
> Responder a: <ipv6-ops-bounces+jordi.palet=consulintel.es at lists.cluenet.de>
> Fecha: Thu, 11 Oct 2007 02:33:26 +0200
> Para: <ipv6-ops at lists.cluenet.de>
> Asunto: Connectivity to Latin American NRENs
>
> Hello everyone,
>
> this is my usual whining about the sad state of the global IPv6 routing
> in certain parts of the DFZ. Today's episode is about the mess one has
> to cross when trying to reach most educational destinations in Latin
> America as well as Canada. I've tried to reach someone responsible by
> mail for the last four weeks, but failed to get any response.
>
> There seems to be a subcontinental educational network called CLARA
> (AS27750) which seems to serve as educational backbone for most if not
> all Latin American NRENs, just like GEANT or Abilene. CLARA is
> interconnected to both GEANT and Abilene which makes, apart from some
> exceptions, IPv6 to perform about equally to IPv4 - seen from the
> educational world.
>
> However, just like e.g. Abilene used to in the beginning, CLARA does not
> provide any transit to the commercial world. This is where our friends
> at UNAM.MX (AS278) jump in and provide that entire CLARA cloud upstream
> to the commercial world. Unfortunately, UNAM.MX itself is a tunnel shop
> specialized in world sightseeing, uplinked to Lava.net (Hawaii) which is
> uplinked to ChungHwa Telecom (Taiwan, maybe with some equipment in US)
> uplinked to Sprint. Example path
>
> 2001:1338::/32 8767 3549 6175 17715 6435 278 18592 27750 27807 20312 i
>
> this is a prefix holding two ccTLD nameservers (ns[12].nic.ve) with glue
> in the root-zone and all other whistles and bells.
>
> As one might imagine, the performance on this path is less than optimal.
> I set up a smokeping when I started to contact UNAM and it looks rather
> horrible -
> http://noc.birkenwald.de/smokeping/smokeping.cgi?target=World.America.nicve
>
> Current mtr output (Smokeping looks rather good at this moment)
> HOST: svr02.teleport-iabg.de Loss% Snt Last Avg Best Wrst StDev
> 1. backbone2-gige-0-3-15.telepo 0.0% 20 0.7 4.6 0.5 47.3 11.8
> 2. mchn-s1-rou-1030.DE.euroring 0.0% 20 2.5 2.6 2.2 2.8 0.2
> 3. hmb-s2-rou-1030.DE.eurorings 0.0% 20 17.5 17.4 17.0 17.8 0.2
> 4. sl-bb1v6-nyc-t-28.sprintv6.n 0.0% 20 108.4 108.4 107.9 108.9 0.2
> 5. sl-bb1v6-rly-t-1003.sprintv6 0.0% 20 184.3 184.5 184.1 185.8 0.4
> 6. 2001:ca0:1::1:1 5.0% 20 495.6 495.6 488.4 499.5 2.9
> 7. tunnel-chttl-lavanoc.lava.ne 5.0% 20 500.1 496.2 489.8 500.1 2.6
> 8. 3ffe:8070:1:13::1 15.0% 20 608.5 570.0 558.8 608.5 13.0
> 9. 2001:1228:11b:90a::1 10.0% 20 811.2 819.2 790.9 974.5 39.6
> 10. 2001:1228:10a:f09::1 5.0% 20 827.9 834.9 803.5 1045. 56.3
> 11. 2001:1228:10a:f02::2 10.0% 20 820.3 859.9 802.9 1118. 97.0
> 12. 2001:1348:1:2::1 10.0% 20 812.8 847.1 797.9 1062. 77.5
> 13. 2001:1348::e 10.0% 20 818.5 842.5 804.9 1006. 58.4
> 14. 2001:1348:1:17::2 15.0% 20 658.1 678.7 653.2 801.4 40.5
> 15. 2800:30:ffff::2 10.0% 20 662.3 687.0 647.7 863.7 54.9
> 16. 2001:1338:ffff:3::1 5.3% 19 656.2 679.7 649.9 937.4 66.4
> 17. 2001:1338:ffff::2 21.1% 19 656.5 700.2 650.3 1032. 108.4
> 18. ns1.nic.ve 15.8% 19 666.0 705.0 651.1 976.8 102.8
>
> Similar issues apply to CA*net (Canadian NREN), which also seems to
> think it does not need any global upstream. Smokeping (seen from
> AS29259) is at
> http://noc.birkenwald.de/smokeping/smokeping.cgi?target=World.America.canet
> the current path is
>
> 2001:410::/32 8767 3549 6175 17715 6435 278 18592 27750 6509
>
> 286 1273 6830 6830 6830 6830 6939 6939 6939 6939 2516 \
> 7660 22388 11537 6509
>
> So why am I writing this? First, I failed to reach anyone at the
> affected ASNs, which makes me think the only hope is some direct
> face-to-face contact of someone reading this mail knowing someone in
> charge. Second, there are at least two networks out there that are
> currently trying hard to become something that could be called a Tier-1
> in the IPv6 world, giving away free transit whereever possible and
> partially resorting to cheap blame games. While I do not think this is
> a bright idea in general and don't approve it, here in this case giving
> (free) commercial transit to CLARA and/or CA*net, even with a short
> tunnel, might be beneficiary (for the educational networks, for the
> willing transit and, ultimately, for the entire IPv6 deployment).
>
> Unfortunately, by publishing AAAA records in DNS that point to such
> badly reachable prefixes they might break _my_ IPv6-deployed clients
> (well, not the clients, the user's experience with IPv6) which is
> something I'd like to avoid.
>
> Does anyone have responsible contacts there and could jump in to provide
> a communication channel with willing upstreams? I did some short checks,
> almost none of the existing commercial (!) IPv4-transits of the Latin
> American networks seem to be active in IPv6, so that road is probably
> blocked.
>
> Thanks,
> Bernhard
**********************************************
The IPv6 Portal: http://www.ipv6tf.org
Bye 6Bone. Hi, IPv6 !
http://www.ipv6day.org
This electronic message contains information which may be privileged or confidential. The information is intended to be for the use of the individual(s) named above. If you are not the intended recipient be aware that any disclosure, copying, distribution or use of the contents of this information, including attached files, is prohibited.
More information about the ipv6-ops
mailing list