Dear RIPE Atlas team,I believe we may have discovered a corner case bug in the way RIPE Atlas handles NTP measurements.
The issue surfaced during experiments with my NTP simulator — a small tool that mimics an NTP server and allows for controlled manipulation of its behavior.
In one particular scenario, the simulator randomly varies the poll value in each response. While this doesn’t occur frequently in real-world implementations, it is technically valid behavior.
It appears, however, that either the probes or the measurement system itself do not expect the poll value to change across responses. When initiating a measurement with the default setting of 3 packets, and each response has a different poll value, the result is that the measurement ends up reporting nothing.
I first described this on the RIPE Atlas mailing list: https://mailman.ripe.net/archives/list/[email protected]/thread/33KNP7KPXN3YKUSEFTQGJMKVEHJAKX3Q/ A detailed analysis can be found here: https://mailman.ripe.net/archives/list/[email protected]/message/OSRNUJ2YN7WFWLIUG2AEJV2FBGZI34NH/ The NTP simulator is publicly reachable at: ntp1.testdns.nlI'd appreciate it if you could take a look. Thank you in advance for your time and consideration!
Kind regards, -- Marco Davids SIDN Labs
smime.p7s
Description: S/MIME-cryptografische ondertekening
----- To unsubscribe from this mailing list or change your subscription options, please visit: https://mailman.ripe.net/mailman3/lists/ripe-atlas.ripe.net/ As we have migrated to Mailman 3, you will need to create an account with the email matching your subscription before you can change your settings. More details at: https://www.ripe.net/membership/mail/mailman-3-migration/
