Keyboard Shortcuts
Likes
- X25
- Messages
Search
|
Re: Hello and is anything always on in the Telebahn
I've set up some fun bounces off my node. Please be kind, these are for convenience and fun at this point. If abuse happens, they will go away. Each of the bounces is limited to 8 connections, and the menu system is limited to 24.
78323592 -> [StarWars] bounces to towel.blinkenlights.nl via telnet 78323593 -> [TeleHack] bounces to telehack.com via telnet 78323599 -> [Menu] my node's network facing menu of known destinations (some on TELEBAHN, some via Telnet over IP) Once I have my BBS online it will live at 78322799. -DK |
|
Re: JJ 2611 config WAS: RE: [X.25] Cisco 2620 as PAD for a modem bank
I had to add the following lines to get any pad connections to go through on the 2620:
x25 profile TELEBAHN dxe ?x25 facility windowsize 10 10
?x25 facility packetsize 1024 1024
?x25 subscribe packetsize permit 32 1024 target 512 1024
?x25 subscribe windowsize permit 1 16 target 2 10
with those lines applied I was finally able to connect to JJ's and I also managed to get an inbound PAD to VTYs setup -DK |
|
JJ 2611 config WAS: RE: [X.25] Cisco 2620 as PAD for a modem bank
john jones
开云体育X252611#sho ver Cisco Internetwork Operating System Software IOS (tm) C2600 Software (C2600-IS-M), Version 12.2(8)T5,? RELEASE SOFTWARE (fc1) TAC Support: Copyright (c) 1986-2002 by cisco Systems, Inc. Compiled Fri 21-Jun-02 13:48 by ccai Image text-base: 0x80008074, data-base: 0x8140878C ? ROM: System Bootstrap, Version 11.3(2)XA4, RELEASE SOFTWARE (fc1) ? X252611 uptime is 18 hours, 16 minutes System returned to ROM by power-on System restarted at 16:20:22 DST Fri Jul 1 2022 System image file is "flash:c2600-is-mz.122-8.t5.bin" ? cisco 2611 (MPC860) processor (revision 0x203) with 59392K/6144K bytes of memory. Processor board ID JAD05030CPT (694590607) M860 processor: part number 0, mask 49 Bridging software. X.25 software, Version 3.0.0. 2 Ethernet/IEEE 802.3 interface(s) 1 Serial network interface(s) 16 terminal line(s) 32K bytes of non-volatile configuration memory. 16384K bytes of processor board System flash (Read/Write) ? Configuration register is 0x2102 ? X252611#sho run Building configuration... ? Current configuration : 2621 bytes ! ! Last configuration change at 01:10:49 DST Sat Jul 2 2022 by cisco ! NVRAM config last updated at 18:00:29 DST Fri Jul 1 2022 by cisco ! version 12.2 service pad to-xot service pad from-xot service tcp-keepalives-in service tcp-keepalives-out service timestamps debug datetime msec service timestamps log datetime msec no service password-encryption ! hostname X252611 ! logging buffered 100000 debugging ? username x25 privilege 2 password 7 11114B50 memory-size iomem 10 clock timezone EST -5 clock summer-time DST recurring ip subnet-zero ip cef ! ! ip host 111.687.x25.org 1.1.1.1 ip host X252611 1.1.1.1 ip name-server 8.8.8.8 ! ! x25 profile TELEBAHN dxe x25 version 1988 x25 modulo 128 x25 address 687111 x25 win 10 x25 wout 10 x25 ips 1024 x25 ops 1024 x25 routing !???????? xot access-group 10 profile TELEBAHN ! ! ! ! interface Loopback0 ip address 1.1.1.1 255.255.255.255 ! interface Ethernet0/0 ip address 192.168.0.100 255.255.255.0 full-duplex no cdp enable ! interface Serial0/0 no ip address encapsulation x25 x25 address 2611 no cdp enable ! ip classless ip route 0.0.0.0 0.0.0.0 192.168.0.1 no ip http server ! access-list 10 permit any ! x25 route (...)(...).. xot dns \2.\1.x25.org x25 host X252611 687111 ! ! banner login ^CC ? Welcome to JJ's X.25 Host on the TeleBahn X25 network. ? ??You can login with user x25 password x25 ? ??Sorry there isn't anything more interesting than a Cisco router to log in to. ? ??Once you log in, you will auto-connect to the Cisco X28 PAD ??????????? ??^C privilege configure level 14 enable privilege exec level 14 configure ! line con 0 line 33 no exec transport input all line 34 38 line 39 session-timeout 5 ?exec-timeout 60 0 absolute-timeout 30 session-disconnect-warning 180 message Inactivity Warning ! Disconnect to 2 minutes x25 subaddress 89 login local autocommand? x28 line 40 session-timeout 360 ?session-disconnect-warning 240 message WARNING.? Your Session will terminate in 2 minutes rotary 99 no exec transport input all line 41 48 line aux 0 modem InOut autocommand? x28 no exec line vty 0 4 exec-timeout 240 0 login local transport input pad v120 telnet rlogin udptn ! ntp clock-period 17180322 ntp server 185.216.231.84 ! end ? From: [email protected] <[email protected]> On Behalf Of john jones via groups.io
Sent: Saturday, July 2, 2022 9:40 AM To: [email protected] Subject: Re: [X.25] Cisco 2620 as PAD for a modem bank ? Use the rotary command on the async port that will be your destination. ?I.E. ?rotary 99. ?This represents the last two positions in your 8 digit X25 address.? ? Then use ip host and/or x25 host commands to express the first 6 digits.? ? I can post an example later on.? ? John On Saturday, July 2, 2022, 1:51 AM, David Kuder <dkgrizzly@...> wrote:
|
|
Re: Cisco 2620 as PAD for a modem bank
john jones
Use the rotary command on the async port that will be your destination. ?I.E. ?rotary 99. ?This represents the last two positions in your 8 digit X25 address.?
toggle quoted message
Show quoted text
Then use ip host and/or x25 host commands to express the first 6 digits.? I can post an example later on.? John On Saturday, July 2, 2022, 1:51 AM, David Kuder <dkgrizzly@...> wrote:
|
|
Re: Linux and X.25
I am he that is working on the dynamic version of jh-xotd. I'm 99% sure I've missed something in making the changes to make it dynamic, and haven't had success testing it against my local cisco router. I also haven't yet figured out how to properly decode the X.25 packets from the tunnel interface to decide where to send them. All it does right now is bind to the XOT port and accept all incoming XOT traffic, shove it in the tunnel interface. Connections the opposite direction will fail when it can't find the outgoing connection for it.? I'll try to dig some more over the weekend as time allows.
My fork of jh-xotd:? -DK |
|
Cisco 2620 as PAD for a modem bank
So, like the other Dave, I've run into some unfortunate side effects of using a Cisco 26xx router. |
|
Re: Linux and X.25
On Jun 29, 2022, at 4:10 PM, Grant Taylor via groups.io <gtaylor@...> wrote:Yes jh-xotd creates it, based on the either it being the first available (=) or hard coded in the config file. c.And a config as: Where is this config added? jh-xotd?In the jh-xotd config file. There is another fork from the repo, that David Grizzly has been working on. That is Dynamic.This is somewhat difficult as it needs you to specify the remote destination, so that needs some work, as honestly we want this “Dynamic”. However it will work for testing. Here 10.0.0.2 is my Cisco.ACK Correct, it only needs to match what used in JH-XOTD config.The tun0 is important for the moment, for Pad_srv. However it kinda makes sense when you think that probably most links where Point to Point. Worst case I think you can use the Cisco as a border XOT router to this.How /important/ is it that it be the tunnel zero interface? Is it only a matter of matching the name that jh-xotd's open_x25.c is using? I have started going through the code to make it dynamic. However it's a horrible spaghetti mess, with possible buffer overrun everywhere. I have partially patched it, so you could set the interface in the config file, and then have JH-XOTD dynamically start pad_srv from its config file. However honestly this thing needs to be totally thrown away. I'm wondering about other things conflicting with the "tun0" interface name.pad_srv managed Getty. JH-XOTD purely takes in XOT packets, removes the X.25 payload and passes to the tun interface.3.Compiled pad_svr ()Does jh-xotd manage the getty? Or is my understanding backwards? I have been looking at the SIMH code, and it appears that the PDP-11 simulator supports the hardware for the DU-11 Sync interface, so it may well be possible to have that use the tun device. If so then you would still use JH-XOTD, but replace pad_srv with services on some kind of PDP-11 OS. Ok you now having comms...:-) |
|
Re: Linux and X.25
On 6/27/22 6:20 PM, Dave Brown wrote:
I have made a bunch of progress which might be helpful:I have some drive by comments / questions. There is a chance that they are answered in things I've not read yet. 1.Created a standalone X.25 module for Raspbian based off Kernel??5.15.32-v7+ (this was on a Raspberry Pi 2)Aside: This is a new and novel way to compile a module that I wasn't familiar with. 2.Compiled jh-xotd ()Am I correct in assuming that tun0 is the most common option to this script? What instantiates the tun0 interface? Is it the jh-xotd daemon? c.And a config as:Where is this config added? jh-xotd? This is somewhat difficult as it needs you to specify the remote destination, so that needs some work, as honestly we want this “Dynamic”. ?However it will work for testing. ?Here 10.0.0.2 is my Cisco.ACK The tun0 is important for the moment, for Pad_srv. However it kinda makes sense when you think that probably most links where Point to Point. Worst case I think you can use the Cisco as a border XOT router to this.How /important/ is it that it be the tunnel zero interface? Is it only a matter of matching the name that jh-xotd's open_x25.c is using? I'm wondering about other things conflicting with the "tun0" interface name. 3.Compiled pad_svr ()Does jh-xotd manage the getty? Or is my understanding backwards? Ok you now having comms...:-) -- Grant. . . . unix || die |
|
Re: Linux and X.25
开云体育I have made a bunch of progress which might be helpful: 1. Created a standalone X.25 module for Raspbian based off Kernel??5.15.32-v7+ (this was on a Raspberry Pi 2) a. The recipe for this is pretty easy, and really just a case of downloading the kernel source, and doing: cd /usr/src/linux/net/x25 CONFIG_X25=m make -C /lib/modules/$(uname -r)/build/ M=$(pwd) 2. Compiled jh-xotd () a. There is something wrong with the getopt in line 232. ?However if you change -1 to 255 it works b. You need a script file as: #!/bin/sh -e echo "Setting up $1" > /dev/console ip link set $1 up route --x25 add 0/0 $1 c. And a config as: tun0?10.0.0.2 /home/dave/jh-xotd/xotd-setup 256 This is somewhat difficult as it needs you to specify the remote destination, so that needs some work, as honestly we want this “Dynamic”. ?However it will work for testing. ?Here 10.0.0.2 is my Cisco. ?The tun0 is important for the moment, for Pad_srv. ?However it kinda makes sense when you think that probably most links where Point to Point. ? Worst case I think you can use the Cisco as a border XOT router to this. 3. Compiled pad_svr () a. Modify Makefile to put -lrt at the end pad_svr gcc line not at the start. ?Else you will get complaints about timer_X (e.g. timer_create) b. Modify "open_x25.c” look for a line as 'strcpy(subscription.device, “X”)’. ?Change the X to be the device you configured in the jh-xotd config c. Create a pad_svr config as: #This is config file for pad_svr program. # local_x121_address 1111 log_level 5 log_path /var/log/pad_svr.log ? Watchdog_timeout 10 Winsize_in 2 Pacsize_in 128 Forward_idle_timeout 1 x29_profile 1:0,2:0,3:0,4:1,13:4,15:0
Buffer_size 32000 Change the x121 address to be the address of this instance (i.e. whatever your put in the x29 route on the Cisco) 4. From a Cisco configure XOT route to point at IP of Linux box, with the address pad X <- Address from pad_svr If it all works you should be prompted to login. ?Ok you now having comms... If you tcpdump of tun0 you will be able to see the raw X.25 frames. ?Now that I can do all that I think it might be possible to make this work with RSX11M, if I can make a DEC synchronous interface join “tunX”. ?
|
|
Re: Linux and X.25
On 6/27/22 12:28 PM, Harald Welte wrote:
That paragraph alone shows you how mature that implementation is. Rebuilding the kernel for switching DTE/DCE role ;)I agree with the logic. However I think that it's extremely important to keep in mind when the documentation was written. A lot may have changed since then. There is also the fact that documentation tends to get out of sync with the kernel / application capability as the latter slowly moves forward. -- Grant. . . . unix || die |
|
Re: Linux and X.25
On 6/27/22 12:27 PM, Harald Welte wrote:
Hi Grant,Hi Harald, Sure. I was just assuming that people who playing with legacy tech would also likely want to interface legacy hardware.Fair enough. But "legacy hardware" can be a lot of things. As you say, it can be something with an X.21 interface. It can also be something with an RS-232 serial interface. The latter RS-232 serial interface can be had with contemporary equipment fairly easily; e.g. Raspberry Pi (et al.) with a TTL/RS-232 voltage level shifter and / or USB/RS-232 interface. There's also for example the FarSite WAN boards that expose X.21. Those should also show up as hldcX netdev and hence be used with the X.25 code. But of course there are always many different use cases.I suspect this would be in the realm of people that explicitly want an X.21 interface, probably because they have another piece of equipment that has a matching X.21 interface which they want to connect to. good luck :)Thank you. My personal take is that with at least some of those incomplete old kernel implementations for ancient protocols it's often easier to simply rewrite what's needed in userspace (I did so with frame relay about two years ago).I largely agree. Though I question how well a user space process can interface with specific hardware /without/ some sort of kernel involvement / blessing, even if it's only a stub driver to provide something minimal for the user space program to glom onto. At the speed of modern computers, there's not really any advantage of running it in the kernel. And if the mainline code has some constraints or bugs you need to work around, you have to rebuild the kernel etc. - a lot of effort.Fair enough. The kernel FR for example imposes some arbitrary MTU limits (1500?) while the FR specs permit for larger message sizes, and non-IP users actually use those in the real world (3GPP Gb interface over FR).1500 -- I presume -- bytes sounds like an Ethernet limitation applied to Frame Relay under the assumption that the FR is transporting Ethernet frames. -- Grant. . . . unix || die |
|
Re: Linux and X.25
On Sat, Jun 25, 2022 at 10:30:20AM -0600, Grant Taylor via groups.io wrote:
If you are running Linux/X.25 to another machine running Linux/X.25 withoutThat paragraph alone shows you how mature that implementation is. Rebuilding the kernel for switching DTE/DCE role ;) -- - Harald Welte <laforge@...> ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6) |
|
Re: Linux and X.25
Hi Grant,
On Sat, Jun 25, 2022 at 10:22:13AM -0600, Grant Taylor via groups.io wrote: Unfortunately, despite being related, I think that X.25 and AX.25 are about:) ok. I haven't ever looked at AX.25 in detail. Sure. I was just assuming that people who playing with legacy tech would alsoYou could followInteresting. likely want to interface legacy hardware. There's also for example the FarSite WAN boards that expose X.21. Those should also show up as hldcX netdev and hence be used with the X.25 code. But of course there are always many different use cases. good luck :)I've not seen any XOT code in Linux, so there doesn't seem to be aI have found xotd.c 0.04 from 1999-01-08 by Stephane Fillod. My personal take is that with at least some of those incomplete old kernel implementations for ancient protocols it's often easier to simply rewrite what's needed in userspace (I did so with frame relay about two years ago). At the speed of modern computers, there's not really any advantage of running it in the kernel. And if the mainline code has some constraints or bugs you need to work around, you have to rebuild the kernel etc. - a lot of effort. The kernel FR for example imposes some arbitrary MTU limits (1500?) while the FR specs permit for larger message sizes, and non-IP users actually use those in the real world (3GPP Gb interface over FR). Regards, Harald -- - Harald Welte <laforge@...> ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6) |
|
Re: Linux and X.25
On 6/25/22 10:30 AM, Grant Taylor via groups.io wrote:
If you are running Linux/X.25 to another machine running Linux/X.25 without an intermediate machine operating as a DCE, it will not work. What you need to do is build *one* of your machines with the LAPB layer operating as a DCE.? To do this edit the linux/include/net/lapb.h and change the line:This part makes me wonder about two things: 1) Copying and renaming the lapbether driver to be lapbether-dte / lapbether-dce so that I can load the kernel module that I want. 2) Configuring an X.25/LAPB/Ether switch system as the DCE for the LAN. I wonder how much, if any, of this I could get to work in a UML as opposed to a full / fat VM. -- Grant. . . . unix || die |
|
Re: Linux and X.25
Pre Script: Thanks Murphy. It's obvious that I would find the docs that I read last night /after/ I hit send on the message.
On 6/25/22 10:22 AM, Grant Taylor via groups.io wrote: I'm working to find the user space commands and utilities to bring the interface(s) up and how to use it (them).It seems as if the README file in the x25-utils-2.1.20 package has the most directly useful information. --8<-- To start up X.25 under Linux ---------------------------- First start up your LAPB interface. At present the only such interface is the LAPB over ethernet device. Firstly get your ethernet card up and running as normal using ifconfig. Next bring up the lapb0 device with: ifconfig lapb0 up Next set up your X.25 routes. I would guess at this stage you just want all X.25 packets to go out via via lapb0, so just do: x25route add 1/0 lpab0 To make sure its all OK do: cat /proc/net/x25* You will see alsorts of useless information ! To set up an X.25 TELNET link, start up telnetd on the remote machine with the command: ./telnetd <address> Where <address> is the X.121 address that you want telnetd to listen on. On the local machine use the command: ./telnet <locaddr> <remaddr> Where <locaddr> is the local X.121 address to be used for the outgoing connection, and <remaddr> is the X.121 address of the remote copy of telnetd. If you are running Linux/X.25 to another machine running Linux/X.25 without an intermediate machine operating as a DCE, it will not work. What you need to do is build *one* of your machines with the LAPB layer operating as a DCE. To do this edit the linux/include/net/lapb.h and change the line: #define LAPB_DEFAULT_MODE (LAPB_STANDARD | LAPB_SLP | LAPB_DTE) to: #define LAPB_DEFAULT_MODE (LAPB_STANDARD | LAPB_SLP | LAPB_DCE) and rebuild your kernel. I hope that this will be a run-time option eventually, but at the moment this is the only way to do it. -->8-- -- Grant. . . . unix || die |
|
Re: Linux and X.25
On 6/25/22 8:13 AM, Harald Welte wrote:
Hi Grant,Hi Harald, unortunately not. However, I know that a number of HAM radio operators are still using the AX.25 in Linux. It has not received a lot of maintenance in the last 15+ years or so, but there's some new grant to get that fixed:*nod*nod* Unfortunately, despite being related, I think that X.25 and AX.25 are about as compatible as IPv4 and IPv6. If even that. At least IPv6 can have IPv4 addresses mapped into it and converted. The net/x25 directory still exists. It looks like the application-side interface uses a sockets based API.I saw on the linux-x25 mailing list that there were updates to the Linux kerenel earlier this year. So there's active development work. I'm using this HDLC code (with frame relay on top, not X.25) even in production setups together with DAHDI on Digium and other cards. So at least up to that HDLC layer we know it works.ACK You could follow with the only difference that you have to use 'sethdlc hdlc0 x25' to tell the HDLC WAN code to put X.25 on top.Interesting. Though I suspect that's going to require specific hardware. Conversely X.25 over TCP or X.25 over Ethernet or X.25 over Async uses more generic hardware. :-) * drivers/net/wan/lapbether.cI was thinking that a lapb0, et al., device was created per interface as you suggested. I'm working to find the user space commands and utilities to bring the interface(s) up and how to use it (them). I've not seen any XOT code in Linux, so there doesn't seem to be a straight-forward way to hook those two up.I have found xotd.c 0.04 from 1999-01-08 by Stephane Fillod. It naturally doesn't want to compile as is on contemporary systems. I've not yet tried to compile on older commensurate systems. -- Grant. . . . unix || die |
|
Re: Linux and X.25
Hi Grant,
On Sat, Jun 25, 2022 at 12:18:57AM -0600, Grant Taylor via groups.io wrote: Does anyone have any experience with X.25 on Linux?unortunately not. However, I know that a number of HAM radio operators are still using the AX.25 in Linux. It has not received a lot of maintenance in the last 15+ years or so, but there's some new grant to get that fixed: The net/x25 directory still exists. It looks like the application-side interface uses a sockets based API. If you grep for ARPHRD_X25, you can find the only drivers that work with that code: * drivers/net/wan/hdlc_x25.c This is a backend for the generic HLDC / synchronous WAN device code. I'm using this HDLC code (with frame relay on top, not X.25) even in production setups together with DAHDI on Digium and other cards. So at least up to that HDLC layer we know it works. You could follow with the only difference that you have to use 'sethdlc hdlc0 x25' to tell the HDLC WAN code to put X.25 on top. * drivers/net/wan/lapbether.c This is the LAPB-over-Ethernet driver you already hinted. I've never used it, but looking at the code it seems that it automatically creates a "lapb*" interface for each ethernet interface in your system. I wondering if / hoping I can establish an XOT connection with ?strid'sI've not seen any XOT code in Linux, so there doesn't seem to be a straight-forward way to hook those two up. -- - Harald Welte <laforge@...> ============================================================================ "Privacy in residential applications is a desirable marketing option." (ETSI EN 300 175-7 Ch. A6) |
|
Linux and X.25
Hi,
Does anyone have any experience with X.25 on Linux? I'm doing a lot of reading about X.25 on Linux, particularly X.25 via LAPB over Ethernet and X.25 over asynchronous serial. This is involving a lot of archeology and really old archives of things form the late '90s / early '00s. I wondering if / hoping I can establish an XOT connection with ?strid's Telebahn and then switch (term?) to multiple things local to my LAN. -- Grant. . . . unix || die |
|
Re: First x25 call attempt
john jones
I think I have joy here !
toggle quoted message
Show quoted text
611_1(config)#do pad 75241001 Trying 75241001... *Mar 1 03:34:35.215: X.25 host name sent for DNS lookup is "410.752.x25.org" *Mar 1 03:34:35.259: DNS got X.25 host mapping for "410.752.x25.org" via network *Mar 1 03:34:35.359: [24.22.130.236,1998/192.168.0.100,40812]: XOT O P2 Call (22) 128 lci 1 *Mar 1 03:34:35.363: From (6): 687111 To (8): 75241001 *Mar 1 03:34:35.363: Facilities: (6) *Mar 1 03:34:35.363: Packet sizes: 1024 1024 *Mar 1 03:34:35.363: Window sizes: 15 15 *Mar 1 03:34:35.363: Call User Data (4): 0x01000000 (pad) I need to set up a point-to-point link where I have both ends to make sure but definitely getting close. John -----Original Message-----
From: [email protected] <[email protected]> On Behalf Of john jones via groups.io Sent: Thursday, June 23, 2022 8:20 PM To: [email protected] Subject: Re: [X.25] First x25 call attempt OK Thanks ! Just got back from dinner so I'll poke around a bit. John -----Original Message----- From: [email protected] <[email protected]> On Behalf Of ?strid smith Sent: Thursday, June 23, 2022 6:21 PM To: [email protected] Subject: Re: [X.25] First x25 call attempt hi john, i've turned on the PAD and the router that it talks to, for the next few hours. nothing's connected to it (yet?) but you are more than welcome to dial it and get the "alerting" response. -- ?strid smith (she/her) =<[ c y b e r ]>= antique telephone collectors association member #4870 On 2022-06-23 at 2:58 pm PDT, john jones via groups.io wrote: So I have my side mostly set up. |
|
Re: First x25 call attempt
john jones
OK Thanks !
toggle quoted message
Show quoted text
Just got back from dinner so I'll poke around a bit. John -----Original Message-----
From: [email protected] <[email protected]> On Behalf Of ?strid smith Sent: Thursday, June 23, 2022 6:21 PM To: [email protected] Subject: Re: [X.25] First x25 call attempt hi john, i've turned on the PAD and the router that it talks to, for the next few hours. nothing's connected to it (yet?) but you are more than welcome to dial it and get the "alerting" response. -- ?strid smith (she/her) =<[ c y b e r ]>= antique telephone collectors association member #4870 On 2022-06-23 at 2:58 pm PDT, john jones via groups.io wrote: So I have my side mostly set up. |