Hello, I was under impression RIPE tries to somewhat anonymize probe information that's displayed publicly, e.g. only showing the probe IP address information down to the subnet or AS level on the maps, providing the option to use "Obfuscated DNS entry", etc. But on the public stats pages e.g. here: https://stat.ripe.net/widget/atlas-probes#w.resource=DE (+ "switch to table view") I am able to see both IPv4 and IPv6 addresses of many or most probes without even having to log-in; is that intended? Also what is the point of choosing to use the Obfuscated DNS option if the probe IPs are available in the clear anyway. -- With respect, Roman
On 7 April 2014 12:45, Roman Mamedov <rm@romanrm.net> wrote:
But on the public stats pages e.g. here: https://stat.ripe.net/widget/atlas-probes#w.resource=DE (+ "switch to table view") I am able to see both IPv4 and IPv6 addresses of many or most probes without even having to log-in; is that intended?
My own probe isn't show in the table for the UK, so maybe it's based on a probe-holder's option somewhere? That may be wrong though, becuase y probe was down yesterday for maintenance window, though is up now, just not for 24h yet. I could be very wrong, just a hunch. Gord -- sent via Gmail web interface, so please excuse my gross neglect of Netiquette
Hi Roman, On 07/04/14 13:45, Roman Mamedov wrote:
Hello,
I was under impression RIPE tries to somewhat anonymize probe information that's displayed publicly, e.g. only showing the probe IP address information down to the subnet or AS level on the maps, providing the option to use "Obfuscated DNS entry", etc.
Indeed, that is true for private probes as well as their measurements.
But on the public stats pages e.g. here: https://stat.ripe.net/widget/atlas-probes#w.resource=DE (+ "switch to table view") I am able to see both IPv4 and IPv6 addresses of many or most probes without even having to log-in; is that intended? Also what is the point of choosing to use the Obfuscated DNS option if the probe IPs are available in the clear anyway.
For a private probe you will only see the probe's ID and the ASN but not the IP address. Best regards, Christian Teuschel
Dear Roman, all, On 07-apr.-14 13:45, Roman Mamedov wrote:
Hello,
I was under impression RIPE tries to somewhat anonymize probe information that's displayed publicly, e.g. only showing the probe IP address information down to the subnet or AS level on the maps, providing the option to use "Obfuscated DNS entry", etc.
But on the public stats pages e.g. here: https://stat.ripe.net/widget/atlas-probes#w.resource=DE (+ "switch to table view") I am able to see both IPv4 and IPv6 addresses of many or most probes without even having to log-in; is that intended?
As my colleague pointed out already, we display all network-related information about public probes (with the exception of personal data of the probe host). Hosts can choose to not mark their probe public, however, this does not make the probe "private", nor does it hide all the networking-related information in the measurement results involving that probe. You can find out what is the difference between the data presented about public and not-public probes in this RIPE Labs article: https://labs.ripe.net/Members/becha/proposing-making-ripe-atlas-data-more-pu...
Also what is the point of choosing to use the Obfuscated DNS option if the probe IPs are available in the clear anyway.
These are two different use cases, and different levels of trying to hide your probe data. Obfuscated DNS is used if you do not want the hostname for the probe to be generated by us, based on the probe-ID, and therefore predictable, even if you do not mark your probe as public. But, indeed, IP address of the probe will be visible *in the measurements results* even if the probe is not marked public. The only difference is that the probes not marked public will not be _listed_ in our tolls - both RIPE Atlas web interfaces, and RIPEstat. We still encourage users to keep the defaults (public), since the goal of the RIPE Atlas is to share measurements results, for the benefit of the whole community. We are working on producing more documentation based on the above article, and based on the results of the discussion with the community, as summarized in the massages to this & MAT-wg mailing lists: https://www.ripe.net/ripe/mail/archives/mat-wg/2014-March/000436.html https://www.ripe.net/ripe/mail/archives/ripe-atlas/2014-March/001359.html Please let me know if you have any further questions. Regards, Vesna
participants (4)
-
Christian Teuschel -
Gord Slater -
Roman Mamedov -
Vesna Manojlovic