Hello all,
I tried searching for my question, and I apologize if I missed the answer. I signed up for the pool with a hostname that only had an A record. I have since enabled IPv6 and added a AAAA record. Since I signed up with a hostname, will I eventually get IPv6 traffic and my stats will always reflect the IPv4 address? I assume I do not need to submit the IPv6 address separately as that would really be a duplication if the hostname is used. Hostname is ntp.cogitoergosum.io
Thanks,
Nathan
Hi and welcome 
The hostname is just for “you”. The system is using IPv4 & IPv6 only.
You can just try to add the system via hostname again or just add the IPv6
Thank you very much, apuls!
You seem to have solved it yourself; everything is looking good:
https://www.ntppool.org/a/cogitoergosum
What exactly did you do? Removed the server and re-added it perhaps? Because I am not sure if the system will automatically add the IPv6-address later on, after the server was initially added with just an A-record.
Hi Marco,
I am not sure if the system will automatically add the IPv6-address later on, after the server was initially added with just an A-record.
This was my question, and my concern was that if I added IPv6 to the pool manually, have I now created a duplicate record if the system automatically polls both A and AAAA when a hostname is provided.
The solution apparently was to just add the IPv6 address as another server to my account, and it seems to function just like @apuls said. The alphanumeric/numeric address is source of truth. So I think it is working and I have not introduced any unnecessary or duplicate sources.
I think it is wasteful of monitoring resources to monitor both ip versions … if we could list the two addresses for each server the monitor could alternate the queries for each ip version.
It doesn’t quite work that way in the real world. Ipv6 addresses will probably tend to score higher. This would appear to be due to better routing and fewer alligator moats. I could be wrong though.
Did I already mention that I’m in favour of adding AAAA-records to the entire pool-domain? Not only to 2.pool.ntp.org, but also to 0,1 en 3 and their country and vendor equivalents.
(I’m not intending to pollute this thread with this matter, so please do not respond. There is already a nice thread about this topic here where you can share your opinions.)
That assumption does not stand any more on all case. My servers score better over IPv4 than IPv6 from San Jose monitoring:
Not anymore, right? pool.ntp.org: NTPman's pool servers
Similar here. My “best” server from scoring point of view is an IPv6 one, but especially from offset spread point of view. Seeing points with more than +/-10 ms is the exception, like the two or three in this snapshot.
Not a systematic analysis, and there’s obviously exceptions, but from glancing at the graphs, the offsets for the IPv4 servers tend to have more pronounced offset outliers into the upper double-digit offset range while the IPv6 ones tend to remain in the lower double-digits.
I guess the monitoring station in San Jose for IPv4 is ussjc1-1a6a7hp, and for IPv6 it is ussjc2-1a6a7hp. And that is right, the monitoring over IPv6 is very good now.
Your servers are in Swiss…why do you care about a San-Jose monitor?
https://www.ntppool.org/a/vm9onk3s5v9mcxq5dc4
As my monitors: 19sfa9p
Are back online, Belgg returned to IPv6, score you pretty well.
Took me a coolple of months to find the problem.
But your servers are ok. My Belgg-IPv6 should be on par in a few days.
Why is a Swiss NTP provider looking at San-Jose? I do not get it.
Inside Europe you are fine.
That was a follow-up answer for a 3 years old discussion thread. At that time, there was one single (but with variable location) monitor assigning the score, and sometimes it was San Jose. If you read carefully the whole thread, you will see this.
Yeah saw that, that got me puzzled about the update 
