Добро пожаловать, Гость. Пожалуйста авторизуйтесь здесь.
FGHIGate на GaNJa NeTWoRK ST@Ti0N - Просмотр сообщения в эхоконференции IPV6
Введите FGHI ссылку:


Присутствуют сообщения из эхоконференции IPV6 с датами от 31 Jul 11 14:37:00 до 03 Oct 24 21:46:09, всего сообщений: 7440
Ответить на сообщение К списку сообщений Предыдущее сообщение Следующее сообщение
= Сообщение: 3423 из 7440 ============================================= IPV6 =
От   : Michiel van der Vlist            2:280/5555         18 Aug 16 00:45:40
Кому : Tony Langdon                                        18 Aug 16 00:45:40
Тема : livewirebbs and IPv6
FGHI : area://IPV6?msgid=2:280/5555+57b4ed60
На   : area://IPV6?msgid=1877.fido-ipv6@3:633/410+1c1a228f
= Кодировка сообщения определена как: CP850 ==================================
==============================================================================
Hello Tony,

On Thursday August 18 2016 08:13, you wrote to me:

MvV>> Having two different host names for IPv4 and IPv6 is
MvV>> undesirable. Just add both the A an the AAAA record to
MvV>> fido.livewirebbs.com and callers will sort out automatically if
MvV>> they should use IPv4 or IPv6.

TL> I agree that the primary hostname should have both A and AAAA records.

Yes, that would be the normal situation wouldn't it? I did not want to confuse Allen with the exceptions. Yet...

TL> There are reasons to have extra IPv4 and possibly IPv6 only records.
TL> The most common being where some of your services only run on IPv4,

Or only on IPv6...

TL> having an A record (e.g. ipv4.yourdomain.com) can prevent delays for
TL> IPv6 capable hosts that attempt to connect.  But the main hostname
TL> should have A and AAAA records.

TL> I'm using separate hostnames for binkp and telnet for this reaason.
TL> When I get around to upgrading the BBS and it can use IPv6, I won't
TL> need to do this.

Yes, if not alle services support both IPv4 and IPv6 a seperate host name is a good idea.

TL>  Similarly, one of my feeds is IPv6 capable.  However,
TL> my feed was on a secondary mailer, which didn't support IPv6 at the
TL> time.  This caused long delays on my end,

Yes, that can happen when software is led to believe IPv6 is available when it is not.

TL> until he added an IPv4 only hostname, which I switched to. The
TL> software has since been upgraded on his end, and I've gone back to
TL> polling the main IPv4/IPv6 hostname, now that his end answers on IPv6.

I am aware of it. In fact Kees van Eeten and me in our joint hunt for Fidonet potential IPv6 nodes often see this. He has a script that hunts for AAAA records in the nodelist that have not been tagged as IPv6 capable yet.

I then poll the node and sometimes I get an IPv6 connect. But more often this is what happens:

+ 00:49 [1716] call to 1:214/22@fidonet
  00:49 [1716] trying f22.n214.z1.binkp.net [2602:306:bd31:aa50:e23f:49ff:fead:fa9]...
? 00:49 [1716] connection to 1:214/22@fidonet failed: {W32 API error
               10060} Connection timed out
  00:49 [1716] trying f22.n214.z1.binkp.net [104.49.12.129]:24554...
  00:49 [1716] connected
+ 00:49 [1716] outgoing session with f22.n214.z1.binkp.net:24554 [104.49.12.129]
- 00:49 [1716] OPT CRAM-MD5-09ad8527aabc6275ef9509f2545471ee
+ 00:49 [1716] Remote requests MD mode
- 00:49 [1716] SYS US 99 BBS

There is an AAAA record, but no server answering on IPv6. Robots have patience so in this case it is no big deal. But if the caller is human, the one minute delay can be very annoying.

When this happens, I get in touch with the sysop. Some times it leads to nothing. Some other times it leads to another entry in the list of IPv6 capable nodes. ;-)


Cheers, Michiel

--- GoldED+/W32-MSVC 1.1.5-b20130111
* Origin: he.net certified sage (2:280/5555)

К главной странице гейта
Powered by NoSFeRaTU`s FGHIGate
Открытие страницы: 0.052801 секунды