[Atlas-anchors-pilot] Services on Atlas anchors as measurement targets

[Stephane Bortzmeyer]

On Fri, Mar 08, 2013 at 09:23:12AM +0100,
 Robert Kisteleki <robert at ripe.net> wrote 
 a message of 26 lines which said:

> At the moment our question to you is: what do you think about adding
> larger pre-defined responses to the DNS service?

It does not work for me yet:

% dig @fr-cdg-as2486.anchors.atlas.ripe.net  TXT 484.6. 

; <<>> DiG 9.8.4-rpz2+rl005.12-P1 <<>> @fr-cdg-as2486.anchors.atlas.ripe.net TXT 484.6.
; (2 servers found)
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: REFUSED, id: 55693
;; flags: qr rd; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 1
;; WARNING: recursion requested but not available

;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags: do; udp: 4096
;; QUESTION SECTION:
;484.6.				IN	TXT

;; Query time: 2 msec
;; SERVER: 2001:67c:217c:4::2#53(2001:67c:217c:4::2)
;; WHEN: Fri Mar  8 09:46:28 2013
;; MSG SIZE  rcvd: 34


> Would you see a benefit to go
> higher, like up to 4k or so? With rate limiting, or without? Obviously, the
> question comes up because the risk of possible DNS reflection attacks.

I won't reply immediately to your question but I have a suggestion
about DNS debugging. Why not installing a DNS looking glass on the
Anchors (shameless plug: <http://www.bortzmeyer.org/dns-lg.html>)? It
is not the same service but it would be useful.