Trouble reaching whois.ripe.net and www.ietf.org

Jeroen Massar jeroen at unfix.org
Tue Oct 17 18:29:32 CEST 2006


Williams, Marc wrote:
> Alexander / Stig,
> 
> 	Currently, access to www.ietf.org is via a single tunnel from
> OCCAID in Ashburn, VA.   Are you aware of any carrier out there who
> would be willing and well equiped to provide redundant access to IETF
> for cases like this?  Perhaps we could have supressed advertisement of
> our prefixes toward the trouble. 

Checking GRH (http://www.sixxs.net/tools/grh/):

grh.sixxs.net> show bgp 2001:503:C779:B:0:0:D1AD:35B4
BGP routing table entry for 2001:503:c779::/48
Paths: (85 available, best #18, table Default-IP-Routing-Table)

Note that GRH has 104 neighbors, clearly a large number don't even
receive your prefix at all.

That /48 rings a bell to me: you are getting filtered at certain places
and thus run into blackholes. Filtering and especially people who don't
update is playing you bad tricks again.

If I have some time again once I'll implement a 'find the broken
neighbour' option into GRH, but upto then you can use the compare
feature. As you are 'single-homed' to OCCAID, paths should be quite
similar as their 2001:4830::/32 prefix, as such check:

https://noc.sixxs.net/tools/grh/compare/?when=now&a=2001:503:c779::/48&b=2001:4830::/32&i=7786

Which shows some discrepancies; most notably the missing prefixes at
NIIF/HUNGARNET/AS1955, Easynet/AS4589 (who also miss out on the root
server blocks), and freenet/AS5430.

Notably Level3/AS3356 seems to perfectly have your prefix directly from
 OCCAID similar to how they get the OCCAID prefix.

I suggest that you start annoying people who are filtering. That
unfortunately is the only way to get this problem resolved.

Greets,
 Jeroen

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 311 bytes
Desc: OpenPGP digital signature
Url : http://lists.cluenet.de/pipermail/ipv6-ops/attachments/20061017/cd1fac1d/signature.bin


More information about the ipv6-ops mailing list