[72211] in North American Network Operators' Group
Re: ultradns reachability
daemon@ATHENA.MIT.EDU (Joe Abley)
Fri Jul 2 11:19:03 2004
In-Reply-To: <20040702144317.GA90005@ussenterprise.ufp.org>
Cc: "Nanog@Merit. Edu" <nanog@merit.edu>
From: Joe Abley <jabley@isc.org>
Date: Fri, 2 Jul 2004 11:16:08 -0400
To: Leo Bicknell <bicknell@ufp.org>
Errors-To: owner-nanog-outgoing@merit.edu
On 2 Jul 2004, at 10:43, Leo Bicknell wrote:
> Note in the later pages what happens to particular servers under
> packet loss. They all start to show an affinity for a subset of
> the servers. It's been said that by putting some non-anycasted
> servers in with the anycasted servers what can happen is if the
> anycast has issues many things will "latch on" to the non-anycasted
> servers and not go back even when the anycast is fixed.
In my opinion, the primary purpose of anycast distribution of
nameservers is reliability of the service as a whole, and not
performance. Being able to reach a server is much more important than
whether you can get a reply from a particular server in 10ms or 500ms.
So, I think the issue you mention (which is certainly mention-worthy)
is a much smaller problem than the apparently observed problem of all
nameservers in the NS set being unavailable.
Joe
