> On of the contacts for the specific IP addresses listed above > (that is a contact for the Roulette sensors) wrote: > > "Shutting down all nodes in the meantime." > but no explanation was provided concerning > why all the NTP requests were being made...
The Roulette Testbed, from what I can see in Google caches, is built out of Soekris hardware, and Soekris hardware is a very common choice for high-stability PPS-locked stratum 1 NTP servers or other places where network timing experiments are made. The networking experiments being done clearly might need some high-resolution timing sources. Of course, why would they use random pool servers to set time on such a high-sensitivity testbed? And why would they be configured so poorly? Probably just a config mistake. Tim. _______________________________________________ timekeepers mailing list [email protected] https://fortytwo.ch/mailman/cgi-bin/listinfo/timekeepers
