I have a hobby project running DNS service to people looking for NTP public servers. I noticed that the DNS servers apparently get ~5 thousand queries per day from IPs that the GeoIP database we use claim are in in Antarctica. It’s less than 0.0001% of the overall DNS queries, but it made me curious what it’d take to make the service work better there.
I imagine the internet service is fragmented between the various stations with each being best connected to a particular country? Does anyone have contacts there that I could talk to? I imagine (some of?) the stations would have a local NTP service as part of their compute facilities.
Back in the early nineties Andrew Tridgell (Samba, rsync etc) put at
least three Linux-based satellite receivers into Antarctica at the
behest of the Oz government. Dunno if they were or are NTP servers.
I have a hobby project running DNS service to people looking for NTP public servers. I noticed that the DNS servers apparently get ~5 thousand queries per day from IPs that the GeoIP database we use claim are in in Antarctica. It’s less than 0.0001% of the overall DNS queries,
My apologies for my sideline question, where did you get the number of
the overall DNS queries? or just said a random number to the air?
One of my buddies was a network engineer at Palmer Station for a winter. Let me reach out to him.
from 1995-1996, i placed a DNS root server in Antarctica. Funding for the bandwidth cost was high enough that I pulled the service. Never really delved into the actual requirement for “real-time” interactions that could not be localized. caching and batch transfers cover most of the need. for more recent work on high bandwidth delay environments, check out: http://ipnsig.org/
It would be really hard to quantify antarctic IPs as actually being from there. I know some of the people who’ve operated the geostationary links to McMurdo and to the pole (inclined orbit satellite visible only part of the day).
Their WAN links go through geostationary transponder capacity and earth stations elsewhere on the planet within the same C-band hemispheric beams. This means that the IPs which Antarctic research stations exit to the internet from after often part of commercial teleport operators or university/research groups, indistinguishable from ordinary ARIN or RIPE blocks assigned to that entity.
For a while a number of links went through a teleport in Florida.