Hello, Jared!
I stumble on IPv6 servers that are allowed in the pool while being no more
than just a tunnel to a IPv6 broker, with entry nodes no closer than a few
countries away, so resulting delays are as large as 110 milliseconds.
Why are those considered any better than 6to4, if they are inferior?
I similarly have found this frustrating with people who pretend that a tunnel
to someone is valuable enough to be in the pool vs native connectivity.
If I remember correctly, when a server is first added into the pool
management
system, it is assigned a score of zero (or even -100) and therefore is
forced
to prove its capability of being useful. Unsalvageably poor servers simply
will not surpass the watermark and perhaps will be speedy deleted.
So, the presence of a [tunneled] server in the pool is a clear indication
that the system considers it a valuable item, -- the clients will discard it
anyway if it performs not so well. And I was wondering why the system
does not
even want to give a chance to servers with "lower-class" addresses.
6to4 is deprecated in RFC 7526.
Thanks for the reference, I was not aware that deprecation is official.
Regarding 6to4 addresses, these are not widely reachable on networks.
There were a lot of issues with 6to4 relays...
Well, for me it "just works": I could reach all well-known IPv6 web sites,
pass connectivity tests, and poll public and private NTP servers.
In contrast, I must add, plain old IPv4 may have reachability problems
when it comes to NTP. Around the month of March, for example, there were
a major NTP blackout on the international links [that my ISPs are using]:
the pool monitoring station in Los Angeles could not reach my server,
while I was not able to reach most servers in other countries -- including
those that operated for years without downtime, and also plenty of others
I selected just for the test. This lasted for more than a month, and then
repeated in the summer, although in a much smaller scale.
My point is that prejudice is not a good thing, especially when automatic
monitoring is employed anyway. If something actually works, then why
not to use it? If something does not work, then who cares how "right" it is?
Most networks are getting IPv6 these days. You should try and seek out
networks that do IPv6 and make it available to you.
Not in the residential sector, especially in my locality. The only exception
are bigger ISPs that occasionally replace broken hardware with newer one
that is already capable of IPv6; but they do not provide it as a service,
rather than as a unwarranted and unsupported bonus. However, bigger ISPs
usually discourage their customers to have a publicly accessible servers:
by not providing static IP addresses, by not giving ability to assign PTR,
by blocking inbound connections. Smaller ISPs, on the other side, do not
have plans for IPv6 at all. It is a vicious circle: ISPs do not consider
IPv6 as there is no demand for it (at least how they think), customers
do not consider IPv6 because they are sure their ISP will never have it.
As for having a server at the office, I have never worked in a company
that has its own servers exposed to the public, or has redundant links,
or has more bandwidth than at home.
There’s no good way for me to link my v4 and v6 host that are the same.
This means you might get answers from 3.pool.ntp.org and 2.pool.ntp.org
that are the same host which is not desirable for a robust set of answers.
If I am not mistaken, I read somewhere here in the list archives that this
is more a "feature" than a "bug", as a multi-homed server is reachable
over multiple networks and therefore deserves several entries in the pool.
However, I personally find it incorrect, too, that all my IPv4 addresses
are considered as independent servers, despite the same CNAME / PTR.
But what would be a perfect solution to this?
Thank you for an in-depth explanation anyway.
--
With best regards,
Anton
_______________________________________________
pool mailing list
[email protected]
http://lists.ntp.org/listinfo/pool
