Slow IN-ADDR.ARPA responses

Anurag Bhatia me at anuragbhatia.com
Thu Feb 9 09:41:14 UTC 2012


Hi John


I have seen similar cases in past with root servers itself. Usually
problems are that local anycasted node here goes down and thus traffic is
redirected to other nearest server in Europe causing high latency. Can you
share traceroute result of a good Vs bad node say A Vs C. Can you see both
coming from within your country?

On Thu, Feb 9, 2012 at 8:16 AM, John Levine <johnl at iecc.com> wrote:

> >I checked the traffic graphs for the server we operate
> >(a.in-addr-servers.arpa) and it has normal traffic loads. Have not heard
> >of any report of issues with the other operators.
>
> Actually, the A server is the only one that's responding quickly,
> viewed from my DSL line hanging off gblx:
>
> A   26ms
> B   83ms
> C  308ms
> D  136ms
> E  248ms
> F   96ms
>
> An acquaintance at LinkedIn tells me that they're seeing the same issue.
>
> Doing a traceroute to the C server, it looks pretty snappy as far as
> London, then slow as soon as it's in Afrinic's network.  I think it's
> usually faster than that.  For the E server, the link between HE in
> California and Australia seems slow.
>
> R's,
> John
>
>


-- 

Anurag Bhatia
anuragbhatia.com
or simply - http://[2001:470:26:78f::5] if you are on IPv6 connected
network!

Twitter: @anurag_bhatia <https://twitter.com/#!/anurag_bhatia>
Linkedin: http://linkedin.anuragbhatia.com



More information about the NANOG mailing list