Robert, thanks for your explanations. I now do understand this for all measurements exept for those of type DNS.
DNS measurements with a central resolver somehow seem not that useful to me. Am I missing the use case here?

Context: I want to measure the consistency of DNS records or how they are seen inside the probes networks.
In my measurement I activated the option to use the probes resolver(s) and left the option to resolve on the probe deactivated.
Does this simulate the scenario of a client asking its local DNS resolver (e.g. assigned by DHCP)?

Thanks, Tim

Tim Wattenberg
+49 1578 8248731

2017-12-01 10:45 GMT+01:00 Robert Kisteleki <robert@ripe.net>:

On 2017-11-30 20:27, Tim Wattenberg wrote:
> Hi everyone,
>
> I think I don’t quite understand the effect of the „Resolve on Probe“ option when creating a measurement. The form says it forces the probe to do DNS resolution, the API reference says that it indicates that a name should be resolved (using DNS) on the probe otherwise it will be resolved on the RIPE Atlas servers.
>
> Could someone explain what this means for example if I have a simple measurement for querying the A record of a given domain via the probe’s resolvers?
>
> Thanks, Tim

Hi,

When you measure something given with a DNS name and leave this option
to its default settings, then the DNS resolution happens once, in the
infrastructure (somewhere in Amsterdam, NL), and the probes are told to
measure towards the resolved *IP*. This is more efficient, prevents DNS
errors on the edges, but only works if DNS can only give one answer.

If you turn on "resolve on probe", then the probes get to measure the
*name* you entered, and do the DNS resolution themselves every single
time they measure. This has a somewhat higher chance of failure, but
it's needed if the resolved IP depends on the location of the vantage point.

Hope this helps,
Robert