owen at delong.com
Sun Oct 28 17:10:08 UTC 2012
On Oct 27, 2012, at 11:28 , Christopher Morrow <morrowc.lists at gmail.com> wrote:
> On Sat, Oct 27, 2012 at 7:05 AM, Anurag Bhatia <me at anuragbhatia.com> wrote:
>> Hi Blair
>> I guess that's pretty much because they don't really wish to put any info
>> related to routers in public including location & circuit bandwidth which is
>> often given major networks in PTR.
> more over, what help is it? I'm of two minds really about this:
> 1) it's handy to say: the router in elbonia is being 'bad'
> 2) it's just as simple to say: 'your router with interface ip 220.127.116.11
> is being bad'
> (or: "everything through 18.104.22.168 is forked... plstofixkthxbi!")
It's handy to say foo-e1-kcks is hosed.
Not as handy to say 2001:db8:5fe3:139a:6254:03ff:fe19:acf3 is hosed.
> It's often cited as a headache to maintain the PTRs (not really,
> automation ftw!) I think really it gets down to "how does it really
See above? Beyond that, it's also convenient if you're trying to correlate outages
affecting more than just google. For example, if I'm getting complaints about
access to google, yahoo, nymex, and edgar and traceroutes to all three of those
show packet loss between routers in Dallas and routers in Atlanta, then I know
I'm probably facing a fiber or carrier outage or partial outage along the Dallas
to Atlanta path. I may be able to take independent action to reroute my traffic
via a more northerly path to avoid that problem.
More information about the NANOG