• Connection Tests

    From Rob Swindell@1:103/705 to Michiel van der Vlist on Tue Apr 4 15:18:23 2023
    Re: Connection Tests
    By: Michiel van der Vlist to Rob Swindell on Tue Apr 04 2023 04:22 pm

    To create a fixed address:

    netsh int ipv6 add address Internet 2600:6c88:8c40:5b:f1do:1:103:705

    That worked (after changing 'o' to '0' of course).

    For incoming you should create a pinhole for port 24554 and maybe port 23.

    With the DMZ set, no pin-holes needed.
    --
    digital man (rob)

    Sling Blade quote #9:
    Doyle Hargraves: Morris here is a modern-day poet, kinda like in olden times. Norco, CA WX: 60.7øF, 24.0% humidity, 8 mph SSE wind, 0.00 inches rain/24hrs --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Michiel van der Vlist@2:280/5555 to Rob Swindell on Wed Apr 5 11:52:31 2023
    Hello Rob,

    On Tuesday April 04 2023 15:18, you wrote to me:

    netsh int ipv6 add address Internet 2600:6c88:8c40:5b:f1do:1:103:705

    That worked (after changing 'o' to '0' of course).

    Of course. ;-)

    And your binkp server answers on that address:

    + 09:37 [3496] call to 1:103/705@fidonet
    09:37 [3496] trying 2600:6c88:8c40:5b:f1d0:1:103:705
    [2600:6c88:8c40:5b:f1d0:1:103:705]...
    09:37 [3496] connected
    + 09:37 [3496] outgoing session with 2600:6c88:8c40:5b:f1d0:1:103:705:24554
    - 09:37 [3496] OPT CRAM-MD5-357f8433fbc4ded531732d8e1c7c8842
    + 09:37 [3496] Remote requests MD mode
    - 09:37 [3496] SYS Vertrauen
    - 09:37 [3496] ZYZ Rob Swindell
    - 09:37 [3496] LOC Riverside County, California
    - 09:37 [3496] NDL 115200,TCP,BINKP
    - 09:37 [3496] TIME Wed Apr 05 2023 00:37:54 GMT-0700 (Pacific Daylight Time)
    - 09:37 [3496] VER BinkIT/2.41,JSBinkP/4,sbbs3.20a/Win32 binkp/1.1
    + 09:37 [3496] addr: 1:103/705@fidonet

    So all you have to do to earn your 'f' in the list is update the DNS for the mew IPv6 address.

    Binkd has the possibility to specify the address to use for outgoing calls. It overrides the OS preference.

    bindaddr 2600:6c88:8c40:5b:f1do:1:103:705

    I don't know about BinkIT.

    For incoming you should create a pinhole for port 24554 and maybe
    port 23.

    With the DMZ set, no pin-holes needed.

    I am still a bit puzzled about your setup. Do you have /any/ barrier between the big bad InterNet and your Fido Machine?


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20170303
    * Origin: he.net certified sage (2:280/5555)
  • From Wilfred van Velzen@2:280/464 to Michiel van der Vlist on Wed Apr 5 12:57:58 2023
    Hi Michiel,

    On 2023-04-05 11:52:31, you wrote to Rob Swindell:

    MvdV> bindaddr 2600:6c88:8c40:5b:f1do:1:103:705
    ^
    Again! ;-)


    Bye, Wilfred.

    --- FMail-lnx64 2.2.0.0
    * Origin: FMail development HQ (2:280/464)
  • From Michiel van der Vlist@2:280/5555 to Wilfred van Velzen on Wed Apr 5 14:14:29 2023
    Hello Wilfred,

    On Wednesday April 05 2023 12:57, you wrote to me:


    MvdV>> bindaddr 2600:6c88:8c40:5b:f1do:1:103:705
    ^
    Again! ;-)

    Arghh! Copy/paste... :(


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20170303
    * Origin: he.net certified sage (2:280/5555)
  • From Rob Swindell@1:103/705 to Michiel van der Vlist on Wed Apr 5 10:08:44 2023
    Re: Connection Tests
    By: Michiel van der Vlist to Rob Swindell on Wed Apr 05 2023 11:52 am

    Hello Rob,

    On Tuesday April 04 2023 15:18, you wrote to me:

    netsh int ipv6 add address Internet 2600:6c88:8c40:5b:f1do:1:103:705

    And your binkp server answers on that address:

    + 09:37 [3496] outgoing session with 2600:6c88:8c40:5b:f1d0:1:103:705:24554

    So all you have to do to earn your 'f' in the list is update the DNS for the mew IPv6 address.

    It probably makes more sense to just create a new hostname for fido traffic and have that point to that IPv6 address. Is this just a vanity address or does it have some functional puporse?

    Binkd has the possibility to specify the address to use for outgoing calls. It overrides the OS preference.

    bindaddr 2600:6c88:8c40:5b:f1do:1:103:705

    I don't know about BinkIT.

    Maybe. Outbound IPv6 interface control isn't very strong in Synchronet right now.

    For incoming you should create a pinhole for port 24554 and maybe
    port 23.

    With the DMZ set, no pin-holes needed.

    I am still a bit puzzled about your setup. Do you have /any/ barrier between the big bad InterNet and your Fido Machine?

    Just the Windows firewall.
    --
    digital man (rob)

    Synchronet "Real Fact" #18:
    Rob Swindell first learned to program in C by hacking on WWIV BBS software Norco, CA WX: 56.3øF, 41.0% humidity, 0 mph ENE wind, 0.00 inches rain/24hrs --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Michiel van der Vlist@2:280/5555 to Rob Swindell on Wed Apr 5 19:42:05 2023
    Hello Rob,

    On Wednesday April 05 2023 10:08, you wrote to me:

    So all you have to do to earn your 'f' in the list is update the
    DNS for the mew IPv6 address.

    It probably makes more sense to just create a new hostname for fido traffic and have that point to that IPv6 address.

    Yes. Always a good idea to have different host names for different services. Then you have the option to split the services over different machines without having to change host names if the need arises. Most DNS providers have no limit on the number of subdomains.

    Is this just a vanity address or does it have some functional puporse?

    It is just vanity, it serves no technical purpose. But it is quit popular among the Fidonet IPv6 sysops. Over a third of them have such a vanity address.

    Binkd has the possibility to specify the address to use for
    outgoing calls. It overrides the OS preference.

    bindaddr 2600:6c88:8c40:5b:f1do:1:103:705

    I don't know about BinkIT.

    Maybe. Outbound IPv6 interface control isn't very strong in Synchronet right now.

    You can always discuss it with the author. ;-)

    I am still a bit puzzled about your setup. Do you have /any/
    barrier between the big bad InterNet and your Fido Machine?

    Just the Windows firewall.

    So you can configure it to pass ICMP6 Ping so that your binkp server address becomes pingable...


    Cheers, Michiel

    -+- GoldED+/W32-MSVC 1.1.5-b20170303
    * Origin: he.net certified sage (2:280/5555)
    ---
    * Origin: he.net certified sage (2:280/5555)
  • From Rob Swindell@1:103/705 to Michiel van der Vlist on Wed Apr 5 13:38:24 2023
    Re: Connection Tests
    By: Michiel van der Vlist to Rob Swindell on Wed Apr 05 2023 07:42 pm

    Hello Rob,

    On Wednesday April 05 2023 10:08, you wrote to me:

    So all you have to do to earn your 'f' in the list is update the
    DNS for the mew IPv6 address.

    It probably makes more sense to just create a new hostname for fido traffic and have that point to that IPv6 address.

    Yes. Always a good idea to have different host names for different services. Then you have the option to split the services over different machines without having to change host names if the need arises. Most DNS providers have no limit on the number of subdomains.

    I'm my own DNS provider, so yeah, no limit. :-) I added binkp.synchro.net for the f1d0 address and updated my _binkp._src SRV record as well. Next up, the Fidonet nodelist.

    Is this just a vanity address or does it have some functional puporse?

    It is just vanity, it serves no technical purpose. But it is quit popular among the Fidonet IPv6 sysops. Over a third of them have such a vanity address.

    It's showing off a "feature" of IPv6. :-)

    Binkd has the possibility to specify the address to use for
    outgoing calls. It overrides the OS preference.

    bindaddr 2600:6c88:8c40:5b:f1do:1:103:705

    I don't know about BinkIT.

    Maybe. Outbound IPv6 interface control isn't very strong in Synchronet right now.

    You can always discuss it with the author. ;-)

    I didn't write most of the IPv6 support in Synchronet, that was Stephen Hurd, but I can certainly look into it.

    I am still a bit puzzled about your setup. Do you have /any/
    barrier between the big bad InterNet and your Fido Machine?

    Just the Windows firewall.

    So you can configure it to pass ICMP6 Ping so that your binkp server address becomes pingable...

    True that.
    --
    digital man (rob)

    Synchronet "Real Fact" #24:
    1584 Synchronet BBS Software registrations were sold between 1992 and 1996 Norco, CA WX: 67.6øF, 23.0% humidity, 0 mph ENE wind, 0.00 inches rain/24hrs --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Michiel van der Vlist@2:280/5555 to Rob Swindell on Wed Apr 5 23:22:14 2023
    Hello Rob,

    On Wednesday April 05 2023 13:38, you wrote to me:

    I'm my own DNS provider, so yeah, no limit. :-) I added
    binkp.synchro.net for the f1d0 address and updated my _binkp._src SRV record as well.

    Check:

    + 22:59 [3480] call to 1:103/705@fidonet
    22:59 [3480] trying binkp.synchro.net [2600:6c88:8c40:5b:f1d0:1:103:705]...
    22:59 [3480] connected
    + 22:59 [3480] outgoing session with binkp.synchro.net:24554
    [2600:6c88:8c40:5b:f1d0:1:103:705]
    - 22:59 [3480] OPT CRAM-MD5-5932afa00556584e121b45fb2b25f3ca
    + 22:59 [3480] Remote requests MD mode
    - 22:59 [3480] SYS Vertrauen
    - 22:59 [3480] ZYZ Rob Swindell
    - 22:59 [3480] LOC Riverside County, California
    - 22:59 [3480] NDL 115200,TCP,BINKP
    - 22:59 [3480] TIME Wed Apr 05 2023 13:59:30 GMT-0700 (Pacific Daylight Time)
    - 22:59 [3480] VER BinkIT/2.41,JSBinkP/4,sbbs3.20a/Win32 binkp/1.1
    + 22:59 [3480] addr: 1:103/705@fidonet

    Next up, the Fidonet nodelist.

    We will see in a day or two...

    Is this just a vanity address or does it have some functional
    puporse?

    It is just vanity, it serves no technical purpose. But it is quit
    popular among the Fidonet IPv6 sysops. Over a third of them have
    such a vanity address.

    It's showing off a "feature" of IPv6. :-)

    Yes, it is. But "playing around" also has a function in climbing the learning curve.

    You can always discuss it with the author. ;-)

    I didn't write most of the IPv6 support in Synchronet, that was
    Stephen Hurd, but I can certainly look into it.

    Staying tuned...


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20170303
    * Origin: he.net certified sage (2:280/5555)
  • From Michiel van der Vlist@2:280/5555 to Rob Swindell on Thu Apr 6 08:56:04 2023
    Hello Rob,

    Wednesday April 05 2023 23:22, I wrote to you:

    Next up, the Fidonet nodelist.

    We will see in a day or two...

    BTW, I also tested vert.synchro.net at the Telnet port.

    [quote]

    Synchronet BBS for Win32 Version 3.20
    Telnet connection from: 2001:1c02:1105:4500:f1d0:2:280:5556
    Resolving hostname...
    [..]
    _
    Synchronet BBS for Win32 Version 3.20 Copyright 2022 Rob Swindell

    V E R T R A U E N

    [/quote]

    So that is OK as well. :-)


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20170303
    * Origin: he.net certified sage (2:280/5555)
  • From Victor Sudakov@2:5005/49 to Rob Swindell on Thu Apr 6 18:30:06 2023
    Dear Rob,

    03 Apr 23 10:35, you wrote to deon:

    https://1drv.ms/i/s!ApZPvWcrEaRQ5_wrKOnYR4bZu_jJ3Q?e=8f5cy5

    There are also options for Port Forwarding, Firewall, IPv6 Pin-holing, IPv6 DMZ, but I've never used any of those (or similar) features for

    Does anyone know what software clients actually support IPv6 Pin-holing? I would think Bittorrent clients and utilities like syncthing should (because they support UPnP for IPv4), but I guess they don't do IPv6 Pin-holing. Maybe some games?

    Victor Sudakov, VAS4-RIPE, VAS47-RIPN
    --- GoldED+/BSD 1.1.5-b20170303-b20170303
    * Origin: Ulthar (2:5005/49)
  • From Michiel van der Vlist@2:280/5555 to Victor Sudakov on Thu Apr 6 15:38:43 2023
    Hello Victor,

    On Thursday April 06 2023 18:30, you wrote to Rob Swindell:

    Does anyone know what software clients actually support IPv6
    Pin-holing?

    I am not sure what you mean by "software clients" in this context. IPv6 pin-holing is something that is applied to a firewall. Firewalls are found in routers and OSs.

    I would think Bittorrent clients and utilities like syncthing should (because they support UPnP for IPv4), but I guess they don't do IPv6 Pin-holing. Maybe some games?

    Please eleborate...


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20170303
    * Origin: he.net certified sage (2:280/5555)
  • From Nigel Reed@1:124/5016 to All on Thu Apr 6 09:59:21 2023
    On Thu, 6 Apr 2023 15:38:43 +0200
    "Michiel van der Vlist" (2:280/5555) <Michiel.van.der.Vlist@f5555.n280.z2.fidonet> wrote:

    Hello Victor,

    On Thursday April 06 2023 18:30, you wrote to Rob Swindell:

    Does anyone know what software clients actually support IPv6
    Pin-holing?

    I am not sure what you mean by "software clients" in this context.
    IPv6 pin-holing is something that is applied to a firewall. Firewalls
    are found in routers and OSs.

    I would think Bittorrent clients and utilities like syncthing
    should (because they support UPnP for IPv4), but I guess they
    don't do IPv6 Pin-holing. Maybe some games?

    I think he means software that uses uphp to open ports on a firewall as
    and when they are needed, and the firewalls that respond to such
    requests by opening the port. Therefore the app is poking a pinhole
    through the firewall.
    --
    End Of The Line BBS - Plano, TX
    telnet endofthelinebbs.com 23
    --- SBBSecho 3.20-Linux
    * Origin: End Of The Line BBS - endofthelinebbs.com (1:124/5016)
  • From Victor Sudakov@2:5005/49 to Michiel van der Vlist on Fri Apr 7 00:25:08 2023
    Dear Michiel,

    06 Apr 23 15:38, you wrote to me:

    Does anyone know what software clients actually support IPv6
    Pin-holing?

    I am not sure what you mean by "software clients" in this context.
    IPv6 pin-holing is something that is applied to a firewall. Firewalls
    are found in routers and OSs.

    A "software client" in this context is a software like Transmission or uTorrent for example, which is capable of requesting a router to open a port to allow incoming packets to the address/ports where the software is listening.

    In other words, a "software client" is something that "applies" pin-holing to a firewall.

    I would think Bittorrent clients and utilities like syncthing
    should (because they support UPnP for IPv4), but I guess they
    don't do IPv6 Pin-holing. Maybe some games?

    Please eleborate...

    The Transmission torrent client, and the syncthing file synchronization utility can use the UPnP protocol to request a firewall to pass *IPv4* incoming traffic (and create a port porwarding for IPv4 NAT). They cannot however (at least to my knowledge) use UPnP or any other protocol to request a router to open a hole for incoming traffic in an *IPv6* firewall.

    Victor Sudakov, VAS4-RIPE, VAS47-RIPN
    --- GoldED+/BSD 1.1.5-b20170303-b20170303
    * Origin: Ulthar (2:5005/49)
  • From Victor Sudakov@2:5005/49 to Nigel Reed on Fri Apr 7 00:32:20 2023
    Dear Nigel,

    06 Apr 23 09:59, you wrote to All:

    Does anyone know what software clients actually support IPv6
    Pin-holing?

    I am not sure what you mean by "software clients" in this context.
    IPv6 pin-holing is something that is applied to a firewall.
    Firewalls are found in routers and OSs.

    I would think Bittorrent clients and utilities like syncthing
    should (because they support UPnP for IPv4), but I guess they
    don't do IPv6 Pin-holing. Maybe some games?

    I think he means software that uses uphp to open ports on a firewall
    as and when they are needed, and the firewalls that respond to
    such requests by opening the port. Therefore the app is poking a
    pinhole through the firewall.

    Correct. I know some software that can have ports opened on an IPv4 firewall, but none so far which can do that to an IPv6 firewall (even if the firewall claims that it supports IPv6 pinholing, who can make use of it?).

    Victor Sudakov, VAS4-RIPE, VAS47-RIPN
    --- GoldED+/BSD 1.1.5-b20170303-b20170303
    * Origin: Ulthar (2:5005/49)
  • From Michiel van der Vlist@2:280/5555 to Victor Sudakov on Mon Apr 10 15:46:23 2023
    Hello Victor,

    On Friday April 07 2023 00:25, you wrote to me:

    Please eleborate...

    The Transmission torrent client, and the syncthing file
    synchronization utility can use the UPnP protocol to request a
    firewall to pass *IPv4* incoming traffic (and create a port porwarding
    for IPv4 NAT). They cannot however (at least to my knowledge) use UPnP
    or any other protocol to request a router to open a hole for incoming traffic in an *IPv6* firewall.

    I see. Or so I think. You ask for some kind of "IPv6 equivalent" for UPnP. But why would you want that? UpNP is a questionable idea anyway. For IPv4 it creates an entry in de NAT table and as a side effect creates a hole in the firewall.

    But why would you need that for IPv6?

    For IPv6 there (normally) is no NAT, so no need to create an entry in a NAT table. In IPv6 avery device has a Unique Global Address, so one can simply create pinholes in advance as needed for the address in question.


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20170303
    * Origin: he.net certified sage (2:280/5555)
  • From Victor Sudakov@2:5005/49 to Michiel van der Vlist on Tue Apr 11 09:33:44 2023
    Dear Michiel,

    10 Apr 23 15:46, you wrote to me:

    Please eleborate...

    The Transmission torrent client, and the syncthing file
    synchronization utility can use the UPnP protocol to request a
    firewall to pass *IPv4* incoming traffic (and create a port
    porwarding for IPv4 NAT). They cannot however (at least to my
    knowledge) use UPnP or any other protocol to request a router to
    open a hole for incoming traffic in an *IPv6* firewall.

    I see. Or so I think. You ask for some kind of "IPv6 equivalent" for
    UPnP. But why would you want that? UpNP is a questionable idea anyway.
    For IPv4 it creates an entry in de NAT table and as a side effect
    creates a hole in the firewall.

    But why would you need that for IPv6?

    For IPv6 there (normally) is no NAT, so no need to create an entry in
    a NAT table.

    The "IPv6 equivalent" for UPnP is not for creating entries in the NAT table (which is absent in IPv6). It is for creating rules in an IPv6 firewall allowing incoming traffic to an application running on an IPv6-enabled host. A firewall (IPv4 or IPv6) is usually configured to block incoming traffic which is not part of an established outgoing connection.

    In IPv6 avery device has a Unique Global Address, so one
    can simply create pinholes in advance as needed for the address in question.

    Only when you know the IPv6 address and port beforehand. Usually an IPv6 address on the home LAN is dynamic (SLAAC), and the port in peer-to-peer applications, VoIP applications etc is often dynamic too.

    Victor Sudakov, VAS4-RIPE, VAS47-RIPN
    --- GoldED+/BSD 1.1.5-b20170303-b20170303
    * Origin: Ulthar (2:5005/49)
  • From Victor Sudakov@2:5005/49 to Michiel van der Vlist on Tue Apr 11 09:47:00 2023
    Dear Michiel,

    10 Apr 23 15:46, you wrote to me:

    Please eleborate...

    The Transmission torrent client, and the syncthing file
    synchronization utility can use the UPnP protocol to request a
    firewall to pass *IPv4* incoming traffic (and create a port
    porwarding for IPv4 NAT). They cannot however (at least to my
    knowledge) use UPnP or any other protocol to request a router to
    open a hole for incoming traffic in an *IPv6* firewall.

    I see. Or so I think. You ask for

    It is not even that I *ask for* it. I've read here, some messages ago, that some home router declared "IPv6 punch-holing support." Infortunately I could not find more information either about the model of the router or its features.


    for some kind of "IPv6 equivalent" for
    UPnP. But why would you want that? UpNP is a questionable idea anyway.
    For IPv4 it creates an entry in de NAT table and as a side effect
    creates a hole in the firewall.

    But why would you need that for IPv6?

    For IPv6 there (normally) is no NAT, so no need to create an entry in
    a NAT table.

    The "IPv6 equivalent" for UPnP is not for creating entries in a NAT table (which is absent in IPv6). It is for creating rules in an IPv6 firewall allowing incoming traffic to an application running on an IPv6-enabled host. A firewall (IPv4 or IPv6) is usually configured to block incoming traffic which is not part of an established outgoing connection.

    In IPv6 avery device has a Unique Global Address, so one
    can simply create pinholes in advance as needed for the address in question.

    Only when you know the IPv6 address and port beforehand. Usually an IPv6 address on the home LAN is dynamic (SLAAC), and the port in peer-to-peer applications, VoIP applications etc is often dynamic too.

    The situation is different of course when you are hosting an IPv6 web-server or something like that. It would have a fixed IPv6 address and port anyway, so there is no need for punch-holing the firewall.

    Victor Sudakov, VAS4-RIPE, VAS47-RIPN
    --- GoldED+/BSD 1.1.5-b20170303-b20170303
    * Origin: Ulthar (2:5005/49)
  • From Michiel van der Vlist@2:280/5555 to Victor Sudakov on Sat Apr 15 09:28:09 2023
    Hello Victor,

    On Tuesday April 11 2023 09:47, you wrote to me:

    In IPv6 avery device has a Unique Global Address, so one
    can simply create pinholes in advance as needed for the address
    in question.

    Only when you know the IPv6 address and port beforehand.

    When runing servers you normally do...

    Usually an IPv6 address on the home LAN is dynamic (SLAAC),

    No. SLAAC addresses are not dynamic. They are derived from the MAC address.

    and the port in peer-to-peer applications, VoIP applications etc is
    often dynamic too.

    VOIP normally uses standard ports.

    The situation is different of course when you are hosting an IPv6 web-server or something like that. It would have a fixed IPv6 address
    and port anyway, so there is no need for punch-holing the firewall.

    Indeed.


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20170303
    * Origin: he.net certified sage (2:280/5555)
  • From Michiel van der Vlist@2:280/5555 to Rob Swindell on Tue Apr 18 11:44:31 2023
    Hello Rob,

    Wednesday April 05 2023 23:22, I wrote to you:

    Next up, the Fidonet nodelist.

    We will see in a day or two...

    Hmmm.... it seems to take a bit longer than just a couple of days. Almost two weeks later and still no binkp.synchro.net in the nodelist for 1:103/705. :(


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20170303
    * Origin: he.net certified sage (2:280/5555)
  • From Rob Swindell@1:103/705 to Michiel van der Vlist on Tue Apr 18 13:57:12 2023
    Re: Connection Tests
    By: Michiel van der Vlist to Rob Swindell on Tue Apr 18 2023 11:44 am

    Hello Rob,

    Wednesday April 05 2023 23:22, I wrote to you:

    Next up, the Fidonet nodelist.

    We will see in a day or two...

    Hmmm.... it seems to take a bit longer than just a couple of days. Almost two weeks later and still no binkp.synchro.net in the nodelist for 1:103/705. :(

    I haven't requested the change from NC/RC yet. That's on me.
    --
    digital man (rob)

    Synchronet "Real Fact" #59:
    Synchronet swag used to be available for purchase at cafepress.com/synchronet Norco, CA WX: 63.3øF, 62.0% humidity, 9 mph S wind, 0.00 inches rain/24hrs
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Victor Sudakov@2:5005/49 to Michiel van der Vlist on Mon Apr 24 01:20:16 2023
    Dear Michiel,

    15 Apr 23 09:28, you wrote to me:

    In IPv6 avery device has a Unique Global Address, so one
    can simply create pinholes in advance as needed for the address
    in question.

    Only when you know the IPv6 address and port beforehand.

    When runing servers you normally do...

    P2P apps like Transmission are not really servers.

    Well they are in the strict sense of the word, but people just start them up and hope for them to work out of the box, and they are often configured by default to randomize port numbers on each start.

    Usually an IPv6 address on the home LAN is dynamic (SLAAC),

    No. SLAAC addresses are not dynamic. They are derived from the MAC address.

    Not any more. AFAIK the recent implementation of SLAAC uses the privacy extensions which do not use the MAC address but some random numbers to derive the IPv6 host address.

    and the port in peer-to-peer applications, VoIP applications etc
    is often dynamic too.

    VOIP normally uses standard ports.

    SIP (the signalling protocol) does, but the RTP uses random ports. A firewall has no way to know the RTP dynamic port numbers unless it inspects the SIP protocol.

    The situation is different of course when you are hosting an IPv6
    web-server or something like that. It would have a fixed IPv6
    address and port anyway, so there is no need for punch-holing the
    firewall.

    Indeed.

    I don't really understand your point. If we decide that UPnP (think "automatic firewall configuration from the inside") is desirable for IPv4, then it's desirable for IPv6 too. If we decide that UPnP is not desirable, you can do without it in IPv4: just configure a static RFC1918 address and port on your internal "server" and create a static NAT/portmapping entry on the router.

    Victor Sudakov, VAS4-RIPE, VAS47-RIPN
    --- GoldED+/BSD 1.1.5-b20170303-b20170303
    * Origin: Ulthar (2:5005/49)
  • From Michiel van der Vlist@2:280/5555 to Victor Sudakov on Mon Apr 24 16:22:01 2023
    Hello Victor,

    On Monday April 24 2023 01:20, you wrote to me:

    Only when you know the IPv6 address and port beforehand.

    When runing servers you normally do...

    P2P apps like Transmission are not really servers.

    Well they are in the strict sense of the word, but people just start
    them up and hope for them to work out of the box,

    That's their problem...

    and they are often configured by default to randomize port numbers on
    each start.

    Bad practise...

    Usually an IPv6 address on the home LAN is dynamic (SLAAC),

    No. SLAAC addresses are not dynamic. They are derived from the
    MAC address.

    Not any more. AFAIK the recent implementation of SLAAC uses the
    privacy extensions which do not use the MAC address but some random numbers to derive the IPv6 host address.

    Privacy extensions use random numbers for the host part. AFAIK SLAAC still uses the MAC address. What I do see is that DHCP6 is often preferred over SLAAC and the host part of a DHCP6 address also looks random. But it definitely is a fixed address. So no problem.

    and the port in peer-to-peer applications, VoIP applications etc
    is often dynamic too.

    VOIP normally uses standard ports.

    SIP (the signalling protocol) does, but the RTP uses random ports. A firewall has no way to know the RTP dynamic port numbers unless it inspects the SIP protocol.

    If those "random" ports are previously initaiated by the SIP protocol there should be no problem.

    The situation is different of course when you are hosting an
    IPv6 web-server or something like that. It would have a fixed
    IPv6 address and port anyway, so there is no need for
    punch-holing the firewall.

    Indeed.

    I don't really understand your point. If we decide that UPnP (think "automatic firewall configuration from the inside") is desirable for
    IPv4,

    That "we" does not include me. I have never used UPnP, have always had it disabled in my routers and never had any need for it.

    I consider UPnP a security risk.

    So maybe I am not the right person to discuss this "issue".

    then it's desirable for IPv6 too. If we decide that UPnP is not
    desirable, you can do without it in IPv4: just configure a static
    RFC1918 address and port on your internal "server" and create a static NAT/portmapping entry on the router.

    Works for me...


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20170303
    * Origin: he.net certified sage (2:280/5555)