[dns-operations] AAAA record for c.root-servers.net

[Daniel Kalchev]

On 23.04.14 21:06, Stephane Bortzmeyer wrote:
> On Mon, Apr 21, 2014 at 10:33:42AM +0300,
>   Daniel Kalchev <daniel at digsys.bg> wrote
>   a message of 165 lines which said:
>
>> This is apparently an bug in the RIPE Atlas probe management
>> software — it needs to make sure the probe can generally reach it’s
>> own measurement targets, before assigning it to do any public IPv6
>> measurements.
> This has been documented:
>
> https://labs.ripe.net/Members/stephane_bortzmeyer/how-many-atlas-probes-believe-they-have-ipv6-but-are-wrong

Yes, and it is worse, that this is a well documented bug, that exists 
for far too long. In the sense that it makes the IPv6 "measurements" by 
RIPE Atlas useless.
Any probe that has not successfully communicated (recently) over the 
requested protocol with at least RIPE Atlas's own targets and now 
Anchors, should never be assigned any tasks for that protocol, because 
the results will be likely bogus. You don't measure things with 
instruments you know do not work at all. I know you did in fact 
recommend this option, but in my opinion that wasn't strong enough 
conclusion. Same about the 6to4 and Teredo connected probes. This does 
not even require any change to the probe's firmware.

Otherwise RIPE Atlas is an very valuable measurement tool.

Daniel