ax25 kernel problem sometimes accur

Hi

After few days of running 9a0tcp gateway, ax25 kernel at 9a0tcp get broken. When the problem occur, all connections are running well, so I think that ax25_list is not broken because ax25_find_socket and ax25_find_cb functions are doing good work, but running netstat or cat /proc/net/ax25 will do "segementation fault". This makes big problems to some of my applications.

The problem occur on kernels 2.4.18, 2.4.19 and 2.4.20, but I notice it only on 9a0tcp gateway, even am running it on several machines. Maybe, because 9a0tcp have few times more traffic.

I analyzed the kernel ax25 source, and I tried to find what can make ax25_get_info function crash. I don't know much about how kernel and sockets exactly work, but want to know if someone have the same problem, and maybe one of kernel guru will catch a problem.

But, what I suspect, in af_ax25.c we define a destructor for our sockets:

         sk->destruct = ax25_free_sock;

I guess that somebody/sometimes call this destructor when destroying a socket. I think that after that, socket cannot be considered as valid, but as control block still exists, shouldn't we set sk in control block to NULL ? ax25_get_info will still find that control block with sk set to destroyed socket, isn't it ?
I added to the end of ax25_free_sock function this line:

        sk->protinfo.ax25->sk=NULL;

and I am waiting to see if this will help.
Is there any expert that can tell me whats wrong here.

Many thanks,

73 de Tihomir Heidelberg, [email protected]

That would be great. I used to see this time to time on two different systems with significant traffic load, namely that of HA3PMF and WB0TAX. But HA5OB/Zoli also reported to experience it on his pc.

Once the system turnes to this state, modules like ax25 and netrom can not be removed any more and usually it dies with a kernel crash in 1 or 2 days. Unless I notice in time and restart the computer but I hate doing so. Always feel like somebody is watching, then he might believe I'm using windows ;-)))

73... Sanyi

Dibuz Sandor  | email: [email protected]
Pecs, Hungary | AX.25: [email protected] | UIN: 81545608

I am aware of several UK stations that are having lockup problems with systems running 2.4.19 kernels (using at least 3 different Linux Distributions). At the moment I know of just one system running 2.4.19 kernel that has been up for over 35 days.

The systems in question seem to run Ok for anything up to 12-14 days, until they stop. Sometimes they stop with a Kernel Panic (Flashing keyboard lamps), other times (at least in my case today) the system is just non-responsive either to external traffic (pings over LAN etc) or to Keyboard Input.

As far as we have determined to date the crashes all follow a pattern where AX25 / NETROM connections start to be refused and sometime there-after the system totally stops functioning.

Stewart (G0LGS)
[email protected]

Hi Sanyi,

thanks for reply, I applyed a fix explained in my post to linux-hams and I am waiting to see if this bug will occur again....

please, are you interested for testing this also ? it would be also good if ha5ob is interested... it would help me to find if it helps or not, because it is hard to find a host with high traffic that have this problem :)

to try it you just need to edit /usr/src/linux/net/ax25/af_ax25.c and in function ax25_free_sock just add this line at the end:

        sk->protinfo.ax25->sk=NULL;
so, the ax25_free_sock function should look:     
static void ax25_free_sock(struct sock *sk)
{
        ax25_free_cb(sk->protinfo.ax25);
        sk->protinfo.ax25->sk=NULL;
}

of course, then you should do make modules;make modules_install if you use ax25 in module or make install if you have it compiled in kernel.

this should not do anything harmfull in kernel, I am quite sure :)

by my opinion, when kernel destroy an socket, there is still a pointer in ax25 control block that points to that destroyed socket and everytime it access it it crash.

so, will see if this will help...

73 de Tihomir, [email protected]

Possibly this is the same bug I have had for ages?
Running 2.4.19 right now.
Last week end, I had a netrom/ax.25 link gone mad, with about 10 packets sent in bursts indefiniteley, the netrom queue was some 32 k. I tried axctl -kill but it segfaulted, then reboot which failed (kernel panic). Will try this change soon.
Thanks
Kjell, sm7gvf

Well, it could crash your system. sk->protinfo.ax25 is kfree'ed in the line before it, so you are accessing invalid data...

Your suggestion wants to be equalivent to sk = NULL;, so you should at least change it to use a valid statement. Though I wonder if you meant sk->protinfo.ax25 = NULL;

Best Regards,
ve1mct

Hi,

yap, you are right, this is not good. if anyone add that line, please remove it, it can make problems.. but interesting is that the kernel does not crash, hm...
that kfree is to short function to notice :)
I thought it does not free it beacuse it does not remove it from ax25_list. But, after some debuging I see that control block is removed before from ax25_list, and then ax25_free_sock is called to clean the rest.

but, why we clean the ax25 connection in two steps ? I looked in other protocols source and very few of them use socket destruct... why ax25_release do not clean all ?

the problem I am trying to solve is why ax25_get_info get crash... the one reason may be that we have dead socket in control block of one of our connection (because other connections are still working), and that can occur if from any reason destructor is called without previous calling ax25_release. Maybe there are other reasons, but just want to check if this one is or isn't our problem.

how bad is idea to clean all for both destruct and ax25_release in case it is not cleaned from the other one ? to me this is not nice, but....

so, will play with it and add some debugging info and wait to see if I can catch anything usefull.

73 de Tihomir Heidelberg, [email protected]

As is the 2.4 AX.25 code is totally racy. Basically it only has a chance to survive if the user is nice ...

73 de DO1GRB op Ralf

--
Loc. JN47BS / CQ 14 / ITU 28 / DOK A21

This reminds me... What is the situation with your fixes? I understood the kernel is now in feature freeze. Will the next release kernel have a bit more un-racy AX.25 stack?

Tomi Manninen           Internet:  [email protected]
OH2BNS                  AX.25:     [email protected]
KP20ME04                Amprnet:   [email protected]

I consider everything I'm currently working on as pure bug fix. As such they're not affected by the feature freeze.

73 de DO1GRB op Ralf

--
Loc. JN47BS / CQ 14 / ITU 28 / DOK A21

Agreed.

Anyway, what is the situation now? Are there still lots of known races unfixed in your tree?

Tomi Manninen           Internet:  [email protected]
OH2BNS                  AX.25:     [email protected]
KP20ME04                Amprnet:   [email protected]

The whole job is much bigger than you may assume. I've removed SOCKOPS_WRAP which so far was forcing single threaded execution of most of the AX.25 code. I did that because for 2.5 the removal of SOCKOPS_WRAP the entire kernel was on the agenda. This didn't happen yet but I'm removed it's use from AX.25 & Co. anyway. As the result the entire locking work isn't just plugging a few holes which would have been bad enough, it's completly redoing the locking stuff from scratch.

73 de DO1GRB op Ralf

--
Loc. JN47BS / CQ 14 / ITU 28 / DOK A21

Ok, I understand. Anyway some kind of progress report would be nice. Even something like: "don't hold your breath", if that is appropriate... :-)

I have gotten surprisingly many private queries about this lately, so to recap what I have understood so far: the 2.4 code is bad, full of races, works only if treated nicely. Your SOCKOPS_WRAP removal obviously makes things even worse until you redo the locking stuff. When that is done we have a perfect ax.25 stack again. Right? ;-)

Tomi Manninen           Internet:  [email protected]
OH2BNS                  AX.25:     [email protected]
KP20ME04                Amprnet:   [email protected]

Holding breath is definately something that could negativly impact your health ;-)

In some sense the improments that are necessary to make the AX.25 stack "future proof" are ridiculous. AX.25 is about the worst protocol design I've ran over in my life, so making a multi-threaded implementation of it is an amusing thought. But the other legacy protocols like Appletalk are going through the same pain ...

73 de DO1GRB op Ralf

--
Loc. JN47BS / CQ 14 / ITU 28 / DOK A21

Gents,

That's not encouraging. Is it really that bad?

Asking becuase TCP/IP over ax.25 is the one thing that can rescue packet, I think. For instance, the fact I can get my provate email in and out when ADSL and cable are down is a real benefit. The fact I can driva all around town and pick up mail with a radio connection, free of charge, is also very beneficial. I can have nmy servers send mail via radio with problem reports when prob;lems occur. Emergency use. The list of useful 1200 bps applications goes on!

But that asusmes the protocol is solid. My 2.0 kernel running old ax.25 stuff has been up 4 years now without dropping a packet. Are we saying nowadays it's worse? Should I can my 2.4.20 kernel with ax.25?

Sorry for dropping in in the middle!

Michael VA3MVW

I think he was refering to ax.25 connections, tcp/ip is generally done in UI frames, for those ax25 is not much different from e.g. ethernet.

On a side note:
What is the state of netrom? I have had only ax25 races with 2.4, netrom seems to be stable (or atleast stable enough to let ax25 crash first :) I am curious to what will need to change to netrom because I would have to update my inp3 patch to it...

Jeroen

Firenet just went through some of this along the east coast, where one gentleman was at his activated EOC, getting weather reports from firenet using ham tcp/ip when all other access was down. These were the only weather reports they were able to get at the time during the ice storms they had back there. He requested a special firenet feed for a 150 mile radius centered on the EOC location in order to keep the number of packets small.

Curt Mills, WE7U                    [email protected]
Senior Methods Engineer/SysAdmin
"Lotto:    A tax on people who are bad at math!"
"Windows:  Microsoft's tax on computer illiterates!" -- WE7U
"The world DOES revolve around me:  I picked the coordinate system!"

I got it working, new 2.4.20 kernel, different PC, all OK. But a couple of questions remain.

(And a side note: On this machine too, /dev/ttyS0 appeared dead. But at least it has two, and /dev/ttyS1 is just fine, so that is what I use. Not an ax25 problem btw, but odd.)

Anyway. I did this:

1. Installed RedHat 8
Custom install.

2. updated Redhat using up2date

3. Recompiled 2.4.20 kernel
Config ax25 on
Serial port kiss on
(SCSI support off (or it would not makedeps))

4. Installed from tarballs:
libax25
ax25tools
ax25apps
...according to AX25-HOWTO

5. edited two /etc/ax25 files:
ax25ports
rc.ax25

The first has:

   # /etc/ax25/axports
   r1  VA3MVW  9600  255  2  145.59MHz TCP/IP port

The second one has this, and I call it at the end of rc.local:

#rc.ax25
/usr/sbin/kissattach /dev/ttyS1 r1 44.135.90.48

   # direct to myself via loopback:
   /sbin/route add 44.135.90.48 lo

# Direct to area gateway port.ve3mch:
/sbin/route add 44.135.85.30 ax0

# Let all other 44 (ampr) routes go via there:
/sbin/route add -net 44.0.0.0 gw 44.135.85.30 netmask 255.0.0.0

# Direct to me at home:
/sbin/route add 44.135.90.46 ax0

/usr/sbin/mheardd

So now I have the following two questions.

1. I have set up TCP/IP OK, and I can ping/send mail/etc; but what about native ax25? Do I need to do anything? How can I check?

2. My ROUTE table shows TWO routes to 44./class A subnet! My previous ax25 box did not do this and the route commands ar ethe same. The route commands in the rc.ax25 script above give me this:

[root@jeeves ax25]# route
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use
Iface
port.ve3mch.amp * 255.255.255.255 UH 0 0 0
ax0

va3mvw.ampr.org *               255.255.255.255 UH    0      0        0 lo
va3oak.ampr.org *               255.255.255.255 UH    0      0        0 
ax0
216.138.207.128 *               255.255.255.224 U     0      0        0 
eth0
127.0.0.0       *               255.0.0.0       U     0      0        0 lo
44.0.0.0        port.ve3mch.amp 255.0.0.0       UG    0      0        0 

ax0
44.0.0.0 * 255.0.0.0 U 0 0 0
ax0
default 216.138.207.129 0.0.0.0 UG 0 0 0
eth0

What is that second 44.0.0.0 doing there?? ANy ideas?

Cheers
Michael VA3MVW

Hello all,

Wonder if anyone saw my previous message?

In any case, another question. I want to be able to telnet into one machine from another using radio. Encryption is forbidden for amateur use so this means clear passwords. Not good. (Even though I will not use root, of course).

BUT, compression _is_ allowed. How can I get the tcp/ip content to be compressed, so it is at least not easily readable?

Thanks,
Michael

PS plus, how can I get telnet to accept connections only on its radio port? This is a generalLinux question but if anyone happens to know the answer?

Compression is still hiding the data and I would say breaks the rules. Better solution: use SSH. You can have public key authentication (no cleartext passwords) without any encryption or compression.

telnetd usually starts from inetd, but inetd doesn't seem to have a way to only listen on certain interfaces. You could use TCP wrappers (ie /etc/hosts.{allow,deny}) to limit the IP addresses which are allowed to connect to ones which you know will be on the radio interface.

Hamish

Hamish Moffatt VK3SB <[email protected]> <[email protected]>

Hello,

You can firewall off port 23 on your internet interface... telnet won't be disabled, but the packets will never reach the daemon (service) for processing. Turn off port 23 on all LAN based connections, and keep it open on the ax interfaces.

This means you need to run either ipchains or iptables. I can help you with either if you wish.

Christian

Is encrypting even a password permitted outside of control of space stations?

73 de Maggie K3XS

-----/___.   _)   Margaret Stephanie Leber    / "The art of progress  /
----/(, /|  /| http://voicenet.com/~maggie   / consists of preserving/
---/   / | / |  _   _   _    `  _AOPA 925383/ order amid change and /
--/ ) /  |/  |_(_(_(_/_(_/__(__(/_  FN20hd / change amid order."   /
-/ (_/   '  K3XS  .-/ .-/    ARRL 39280   /___ --A.N.Whitehead ___/
/____ICQ 7161096_(_/_(_/__AMSAT 32844____/ <[email protected]>

Make sure to carefully set up SSH to do encrypted authentication but clear-text everything else. By default it encrypts everything going across the pipe.

Curt Mills, WE7U                    [email protected]
Senior Methods Engineer/SysAdmin
"Lotto:    A tax on people who are bad at math!"
"Windows:  Microsoft's tax on computer illiterates!" -- WE7U
"The world DOES revolve around me:  I picked the coordinate system!"

As I understand it, it's allowed in the U.S. for authentication purposes (passwords), but not for hiding the rest of the text. Can't speak for other countries rules.

Compression is also allowed as long as it's a well accepted and documented protocol that's in use. In other words, it's purpose is for getting more data through a smaller pipe rather than obscuring the meaning of the data. One should be able to grab the data and de-compress it, assuming they have the technical know-how to do so.

As with anything, you'll still get arguments both ways on both authentication and compression issues. See what the TCP/IP guys are doing on RF. They use encrypted authentication at times, and commonly accepted compression as well.

Curt Mills, WE7U                    [email protected]
Senior Methods Engineer/SysAdmin
"Lotto:    A tax on people who are bad at math!"
"Windows:  Microsoft's tax on computer illiterates!" -- WE7U
"The world DOES revolve around me:  I picked the coordinate system!"

Compression is not breaking the rules and hasn't been for a long time. NOS, JNOS, FBB and others all use compression for forwarding mail and it's perfectly legal. We've been doing this now for more than a decade.

SSH does do encryption and optionally can do compression as well. From the SSH man page " It (ssh) is intended to replace rlogin and rsh, and provide secure encrypted communications between two untrusted hosts over an insecure network."

Passwords for the purpose of remote control (repeaters, remote packet systems etc.) have been legally used for a long time. With the advent of Linux lots of us have been using SSH to control and administer remote packet sites. While the fact that SSH also encrypts the rest of your communication besides the password might be considered by some to be a violation, most of us consider that traffic to also be part of 'control' and in all the years we've been doing this the FCC has never disabused any of us of that idea. :-)

Ken, N7IPB
http://wetnet.net

How about using OPIE: (One time use passwords with server / client and no need for encryption).

http://packages.debian.org/stable/admin/opie-server.html
http://packages.debian.org/stable/admin/opie-client.html
http://packages.debian.org/unstable/libs/libpam-opie.html

Kelly Black
KB0GBJ

And I wrote and sent with a typical US centric view without thinking that Hamish is not in the US. Take what I said with the appropriate slant. :-)
Ken, N7IPB

Well, I am in Canada as well - but we are broadly the same as US. Encryption forbidden, compression allowed.

Mike VA3MVW

Good discussion about encryption, please keep it going :)

Meanwhile another question, and I do apologise for all the traffic.

Every time I ping another node it fails the first time, then works the 2nd time.

E.g. from va3mvw.ampr.org I ping va3oak.ampr.org. Both are on my local radio frequency and I have direct routes from one to the other and vice-versa.

The first time, instead of a ping (ICMP echo request), my node send a packet "from VA3MVW to QST". That results in no replies, so in the shell I see "no route to host".

If I ping again immediately, it DOES send the echo request, and gets back an echo ack from the other side, and all is OK< the ping works.

What beginner's mistake am I making? DNS sounds likely but no, I am pinging IP's, not names. So what else am I overlooking?

Thanks for your indulgence again, and thanks for any help!

Michael VA3MVW

SSH at least used to have a cipher option "none" that disables all encryption. It might require recompiling as the option is typically disabled to protect people who don't understand what it means...

I have used it many times combined with RSA authentication and compression and I always felt safe about not breaking the rules here in Finland. Of course I might have been wrong all the time... :)

Tomi Manninen           Internet:  [email protected]
OH2BNS                  AX.25:     [email protected]
KP20ME04                Amprnet:   [email protected]

These are ARP packets (Address Resolution Protocol, mapping IP-numbers to hardware addresses ie. callsigns).

Chances are your ARP timings are too aggressive, they usually are by default. This causes it to time out and report "no route" to you.

It would seem that the other system does indeed reply to the ARP query while you are not watching anymore but not in time for the first ping to succeed. The second time it's cached so everything works.

Two options. First, set up a static ARP entry. This is practical if your neighbors are very few and static ie. don't change things like their callsign-ssid's very often. The syntax is something like this:

arp -H ax25 -i <ifname> -s <ipaddr> <callsign>

Second option is to tweak the timings. I prefer this and have this in my startup script:

# ARP timings
#
PROCPATH=/proc/sys/net/ipv4/neigh/${IFNAME}

 echo 5          > ${PROCPATH}/mcast_solicit		# 5 retries
 echo 500        > ${PROCPATH}/retrans_time		# 5 sec intervals
 echo 1800       > ${PROCPATH}/base_reachable_time

For further info, see /usr/src/linux/Documentation/filesystems/proc.txt under "Network Neighbor handling".

Tomi Manninen           Internet:  [email protected]
OH2BNS                  AX.25:     [email protected]
KP20ME04                Amprnet:   [email protected]

With SSH set up for a private and public key pair, you aren't even exchanging encrypted passwords.

The remote machine has your public key, and you have your private key. The remote machine gives you something to encrypt using the private key and send back to it (a challenge). If you don't have the private key it is impossible to meet its challenge. If somebody else had the private key they could provide the proper response too.

Hamish

Hamish Moffatt VK3SB <[email protected]> <[email protected]>

OK, my mistake. However I was under the impression that the original poster was actually looking for a way to obscure his communications, which the regulations don't allow.

That can be disabled.

Hamish

Hamish Moffatt VK3SB <[email protected]> <[email protected]>

I believe that OPIE is really what you are looking for. I know this was mentioned before but seriously look at it. One Time Passwords. I have this enabled for login (console), ftp, telnet and ssh. I am using S/Key myself but you can find opie where the previous poster mentioned for their distro and you can find it in the folder 'extra/ham/login' in mine. =]

http://distro.ibiblio.org/pub/Linux/distributions/slackware/slackware-8.1/extra/ham/login/

Setup opie, make a login that *must* use it, and open up telnet over the air. No encryption goes over the airwaves. Anyone that watches you log in will not be able to follow in your footsteps.

Jack

All,

Bear with me one more time with beginner's questions...

I see I have to be root to use the "listen" app. Since listen has 755 permissions, why can I not run it as a normal user?

Also I see some apps have changed since the (excellent) AX25-Howto was written (e.g. ttylinkd is new). Where's the best place to find new (if there are any!) ham radio Linux software?

Many thx all
73 Michael VA3MVW

Ass listen 'sniffs' the network devices, it need to run as root. I know - this is not really precise, so someone correct me.

Take a look at http://radio.linux.org.au

73, Luc, LX2GT

Luc,

OK, makes sense (it is a promiscuous mode interface!). But what did I do on my old 2.0 kernel system (my 486, which has been up forever!) to allow it there?

Thanks for the link too.

This is still fun.

Michael

Any user can run the program (file permissions), but only root is allowed to perform the needed network operations (kernel behavior).

Make sure you're looking at this version of the HOWTO:

    http://www.linux.org/docs/ldp/howto/AX25-HOWTO/

There's a much older one still in circulation.

Lots of software isn't in the HOWTO, since it focuses on the basic transport. The hamsoft database at radio.linux.org.au list a lot of stuff, though its links are often out of date. Google is your friend, both for finding unlisted stuff, and for re-locating the lost hamstuff items. If you're looking for something specific, asking here would be a good thing to try also.

Dennis Boone
KB8ZQZ

Earlier, listen was by default installed with the SUID bit set and so any user could run it with root privileges. As this is considered somewhat dangerous (a bug in listen could allow a local root compromise), ax25-apps was changed not to do this. So now it is up to each sysop to enable it if needed. You can do this as root:

chown root /usr/bin/listen # if not already...
chmod u+s /usr/bin/listen

Tomi Manninen           Internet:  [email protected]
OH2BNS                  AX.25:     [email protected]
KP20ME04                Amprnet:   [email protected]

APRS Software: http://www.xastir.org

Run the development versions. They're very stable. "Stable" version is much too old now. "Development" versions come out about every two weeks. CVS version changes almost every day. New enough for you?

Curt Mills, WE7U                    [email protected]
Senior Methods Engineer/SysAdmin
"Lotto:    A tax on people who are bad at math!"
"Windows:  Microsoft's tax on computer illiterates!" -- WE7U
"The world DOES revolve around me:  I picked the coordinate system!"

the cleartext passwort for becoming root may be ommited with this littel tool, on
http://x-berg.in-berlin.de/cgi-bin/viewcvs.cgi/ampr/root/
(downloading the tarball (link on the bottom of the page is suggested)).

it does md5-password challenge in the standard way as bbs password protection wroks.
this is some kind of one-way-password mechanism, and it does not matter that anyone reads your answer string.

be aware that your ax25 connections with your PID=text session or your ax25.ip.tcp telnet session _may_ be "overtaken" by someone else.

73,
- thomas dl9sau

Hello all,

I want a simple message system on my packet boxes. Either ax.25 based or TCP/IP based.

I do not need heavy all-encompassing OS-like systems - all I want is a simple BBS-style system. Any recomemndations for what works well for you all, and is simple to install and maintain?

Background: I already have the system itself and ax.25 and TCP/IP over ax.25 working. Just need a BBS on top!

Michael

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I'm sure there's real software out there for just this purpose, but you could always roll your own. In fact, I don't think it would be that hard to make a simple message system. If you treat the ax.25 device as a TTY, then you can bind agetty to the device, and make the BBS software be the shell when they login.

Cheers

- --
Michael Melanson VE3MTM
[email protected]
73 33
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE+AUQhxg/UNuqCTUsRAmVnAJ9rYTdJYLCdx8eYKuGOmmehsJatswCgi2RV UCXPbyvDnmcYA7RANSlbM3w=
=rKzO
-----END PGP SIGNATURE-----

Hi all,

OK, I have another all- new machine ready to go online. Redhat 8 with Custom kernel (2.4.20 with ax25 and kiss compiled in), two serial ports, routing for my two IPs, and so on. An as-new P133 with 64 MB and no cpu fan - perfect for many years of reliable radio use.

(My server 'hambox' has been up for years - 371 days since the last reboot for regular dust-cleaning.)

Now, I asked before but forget the answers. Forgive me for asking again:

a) I should know this, but how can I get telnet to only work on the radio ports and not on the ethernet port? I.e. telnet (unencrypted) only from the ax25 ports.

b) I use tcp/ip for mail etc. But apart from that I would like to set up a simple messaging system/bbs, to use when someone goes to this system using ax25. I do not need "do it all" systems, just a simple bbs (and it's a 1200 bps port, so a simple one with little text would be great). Suggestions, all?

Many thanks for your indulgence.

73
Michael VA3MVW

Use the 'route' command. Make the radio port the default interface. All IP traffic will, from then on, use that port.

Consider using 'ax25d' to start up the 'mail' program, allowing people to send and receive local E-mail.

HTH...

Wilbert, ZL2BSJ

Hello again all,

Mmm... perhaps, though maybe 'mail' is a bit TOO simple, as it requires a bit of user knowledge, and machine-level user accounts (AFAIK) for all users.

Anyone ever use the other BBS- (not smtp-) type message systems? I see all sorts here and there but have no idea which ones are simple and still work. I suppose I can try them all out, but that will no doubt require a lot of hours spent.

(I seem to be less than stellarly successful with this. E.g. I cannot even get _local_ "talk" working: enabled it in xinetd but still refuses connections. Let alone do it from ax25. ALso, how do I create some loopback link for testing?).

73, Merry Christmas all,

Michael

RH uses by default xinetd so this should be easy. The directory /etc/xinetd.d/ has all the config stuff. Creative use of the "bind" and "only_from" keywords should solve this for you.

Just as an example I have on our club BBS a setup where a telnet to one of amprnet IP's assigned to that box goes to LinuxNode and another to the FBB BBS. A telnet to the "commercial" IP address goes to a regular Linux login (well, would go if it were enabled -- of course only SSH is allowed over the Net). And these are all at the default telnet port (23).

Other than a full FBB BBS, not really. Hessu started writing something to this extent but I think that was never finished.

Tomi Manninen           Internet:  [email protected]
OH2BNS                  AX.25:     [email protected]
KP20ME04                Amprnet:   [email protected]

Tomi,

[Telnet?]

OK, will look for those keywords in the readmes etc.

[BBS?]

OK. WIll try that then. If any hints for me (compiling on 2.4.20 kernel), let me know - otherwise I'll just try. (I already see that it is NOT a simple ./configure, make, make install!!)

73 and Merry Christmas

Michael
VA3MVW

Thank god for the Internet and for google!

Of course fbb failed the make.

...and now it makes fine.

Michael

OK, pity, but FBB is deleted. I have never seen a worse install in my life, of any software! Apart from the edits needed before it will even compile, there are no instructions (the README just says to 'make' and enjoy), and the scripts included give nothing but fatal errors. Plus the English is bad, but that is not the author's fault, and the licensing is odd (not GPL: it asks not to make changes)..

So I am glad it works for some of you, but for me this is such low standard software that I have deleted it.

Any ideas for other systems like this, then?

Michael

Hi All

using

and Tomi responded:

I havent used it myself, but awznode by Stefano Noferi IZ5AWZ seemed to have simple mail capability, passwords etc. It grew out of the normal "node" program so should be fairly familiar.

The reference I found was ftp://ftp.funet.fi/pub/ham/packet/linux/awznode/

73 and Merry Xmas to all

Steve, vk5asf

Hi all,

I am compiling the kernel and using sound card CMI8738, before that I need a confirmation for the kernel compile option.

According <http://tldp.org/HOWTO/AX25-HOWTO/x495.html>
6.1.6. Creating a user-mode Soundmodem device ..
Thomas Sailer has written a sound modem driver that runs in user-mode using the kernel sound drivers, so it should work with any sound card supported under Linux.
..
The driver is implemented as the user-mode program soundmodem. The graphical soundmodemconfig program allows configuring and testing the soundmodem driver. As well as kernel sound support you need the kernel AX.25 mkiss driver.
..

Should I do the following for the kernel compile option?:
SOUND
(*)y ( )m ( )n Sound card support
..
..
( )y (*)m ( )n C-Media PCI (CMI8338/8738) ..
..
(* )y ( )m ( )n Support CMI8738 based audio card ..
..
Or any other option is better?

Many thanks for any assistance.
Vivi

__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com

Hi,

----- Original Message -----
From: <[email protected]>

Hmm...
Why are you the first with this report ?

Merry XMAS
Cees - PA3AES

It has an 's <sysop_callsign>' command for sending mail to the sysop. The sysop gets to see it as E-mail. But it's one way only. There is no 'r' command for the user to read a reply. No 3rd party mail.

AWZnode is really an on-demand loading node (very good), not a PMS or BBS.

What I do here for a PMS is, to tell users to reconnect to my AWZnode and leave me an E-mail with 's', or drop me a line at the local BBS. LinPac then downloads it on the hour. Sorry, I can't be more helpful...

With regards to the 'talk' problem somebody mentioned, make sure that 'mesg y' is executed on the terminal you want to talk to. I have it in my .bash_profile

Wilbert, ZL2BSJ

I think you are wrong in several ways. The times I have set up fbb have been easy. Included is a scipt which you run, which creates all the needed directories and creates a default setup. I have rarely encountered a similar piece of complex software which is so easy to set up.

The english is certainly not bad and the licensing has been GPL for over a year now.

Low standard software? Can you explain to me why it is used all over the world by almost every bbs?

Yeah, build your own if you can't find what you want.

Joop

Joop,

[F6FBB not very useable]

But not instructions: they are only on the web site, and even there they are old and do not fit the software... took me an hour to see I had to run that script. How about a simple readme that mentions this? These are the very basics of software!

I must mention, I downloaded the tar.gz, not the rpm: I run a too-modern RedHat for that. Alas, there is no standard "./configure, make, make install", and the bbs when it does come up is mysterious, needs obscure text files edited, and displays Toulouse stuff by default...

But the readme urges users to not make changes on their own. Surely that's not GPL?

I would say, because there appears to be nothing else.

Ah, that's why I am asking here.

73 de Michael va3mvw

Joop,

Not to argue, but check http://ftp.f6fbb.org/fbbdoc/doccopyr.htm

This is definitely not GPL:

"This software is delivered at no cost and is free of all charges. It can be freely copied or installed for any HAM-RADIO application. Any other usage for commercial, professional, or industrial purposes is forbidden. Despite the free sharing nature of this product, a participation to the investment and to the mailing cost at the level of 100 FF or 20 US Dollars (or more) would be most welcome."

73,
Michael

You simply need working sound card support so what ever gets your "xmms" or whatever working should suffice.

Note however that not all soundcards have good enough drivers to be used with soundmodem. Soundmodem requires a bit more from the drivers than usual. You will have to test if your card works. Fortunately I think most cards/drivers work just fine.

Tomi Manninen           Internet:  [email protected]
OH2BNS                  AX.25:     [email protected]
KP20ME04                Amprnet:   [email protected]

FBB is not the easiest software to set up, I admit. Most of the documentation at http://www.f6fbb.org/fbbdoc/doc.htm still applies to the latest version but the Linux specific stuff is not well documented.

You should read all the doc files from the source package: README, INFO, FBBTREE and especially CHANGES!

It has been changed to real GPL some time ago. See the COPYING file included in the source. Unfortunately there are references to the old licence here and there...

Tomi Manninen           Internet:  [email protected]
OH2BNS                  AX.25:     [email protected]
KP20ME04                Amprnet:   [email protected]

----- Original Message -----
From: <[email protected]>

....

....

Wow !
The source-tree of the latest version I have here (7.04h) has all these things in it.

There is a file "COPYING" with the GPL license in it, there are several readme-files how you should compile the code, there is no "configure" needed, you should only check if the path of the ax25-utils/libraries differs from the standard path and change it if needed. Then you can type "make" and after that "make install" and the installation script will install everything in the right directory's and asks your station details. If you answer all questions, you will end up with a working installation. (and there is absolutely no need to remove include files from the source)

Looking in the sourcecode tree for a file called README or INFO only costs 1 minute of your time. (and is probably less then the time it took to write about "bad sofware")

If all of this is too difficult for you, please do not even try to install it. Probably the normal daily sysop work would also be too difficult ....

73
Cees

Hi All,

to

"node"

and Wilbert replied:

I downloaded a copy of it. From the README file:
"Other features:
- local users of the linux machine can be granted access to the mailreader and optionally to the shell. ...............(bits omitted) Users with the mail option enabled will receive messages like "New mail has arrived", mostly when the stated condition is met ;-)
- all users can send messages to a callsign associated in the kernel table with a linux username ("send" command)."

This may be a new feature in the v0.4-pre2 version I downloaded.

I think that is still true

It seems to be evolving into a PMS, going by what I've seen in the above.

regards

Steve

Cees,

Although I detect a tone of disrespect in your reply, which I think is uncalled for among amateurs, I will answer seriously.

To answer your point, no, the normal sysadmin work is not too difficult; although as the CTO of a tech company, I generally have other people do it. But if it WERE too difficult, I would assume you would be happy to help, as I would be happy to help too.

Now, I have the previous version, namely 7.03, because that is the last one I can find on the fbb site "last versions" page as source. As far as I can see, 7.04 is only available as rpm. And I am assuming
http://www.f6fbb.org/ is the authoritative source. If I am misreading the site or just not looking right, please tell me where I can find 7.04 source and I'll go download it!

73,
Michael VA3MVW

Cees,

To be exact, the f6fbb site has this on its "latest versions" page:

----------------------------------
LinFbb

LinFBB 7.03c daemon-version (Linux) source code (634 KB) LinFBB 7.03 daemon-version (Linux) Redhat 5.2 RPM (850 KB) LinFBB 7.04 all-versions (Linux) Redhat 6.1 RPM (879 KB) LinFBB 7.03 daemon-version (Linux) SuSE 7.0 RPM (352 KB) LinFBB 7.03 daemon-version (Linux) Slackware 7.0 package (353 KB) LinFBB 7.03 daemon-version (Linux) Debian package (333 KB) Windows client v1.12 for LinFBB 7.03 daemon-version (888 KB) 7pFBB 2.00 (7+) package for LinFBB 7.03.
MULTI 1.25 package for LinFBB.
C_FILTER package for LinFBB.
----------------------------------

..so that is why I thought the latest version source was 7.03 (and I have RedHAt 8, so a 6.1 rpm would almost certianly not work).

Cheers,
Michael

Yes, this (too) is not very well set up on the web page. The latest source distribution is at:

ftp://ftp.f6fbb.org/pub/f6fbb/distrib/linux/src/

Tomi Manninen           Internet:  [email protected]
OH2BNS                  AX.25:     [email protected]
KP20ME04                Amprnet:   [email protected]

Actually I don't think even this is necessary. It's left over from the time when libax25 was not available as a dynamic library with the headers installed in a standard location.

There is definitely a lot to be done in FBB to get these things right but then again having written a few software packages myself I can really say that documantation is definitely the most annoying, dull and even difficult part of it all... So I'm not going to say anything bad about Jean-Paul because of it... :)

Tomi Manninen           Internet:  [email protected]
OH2BNS                  AX.25:     [email protected]
KP20ME04                Amprnet:   [email protected]

Hi,

----- Original Message -----
From: <[email protected]>

That (ham-spirit) is why your first mail surprised me (and others I think)

If someone has problems with software, normally on this newsgroup he asks what is wrong, before he starts writing that it is the worst software he has seen in his life with missing installation and readme files, (and other remarks). But maybe I have misunderstood you and I apologise if so !

I still have the sourcecode of all versions since 7.02 upto 7.04h (25 versions) and they all have the readme/info/copying files in them, so probably you have a different installation archive. The www.f6fbb.org site is not updated all the time I think, and it is usually slow, but there are many mirrors where you can dowload the latest sourcecode. You could try ftp.funet.fi and navigate to the ax25 / bbs / fbb directory's. Also on Google you will find other download sources.

If you have any questions on installing FBB, you could try the XFBB mailing list too !
Most questions will be answered within hours.

73,

Cees

Op wo 25-12-2002, om 21:01 schreef [email protected]:

Michael,

I think you are referring to the old copyright notice.

Joop

Op wo 25-12-2002, om 15:54 schreef [email protected]:

To be honest, I have heard of at least one other ham who encountered the same problems. When I told him the only thing he had to do was run the script, his problems were solved.

Hmm, yes. A bit of fbb knowledge seems to be needed to get used to setting up fbb. I would say it's not software to begin with if you want a simple mailbox.

No.

Well fbb originally comes from a dos environment. This means you will find a dos-like environment both for setup and at runtime. But I guess it's kindof difficult to maintain a multi-platform mailbox and have it perform nicely on every platform. So, if you are looking for a unix/linux-like mailbox, you should try something else.

There are several mailboxes mentioned at
http://radio.linux.org.au/?sectpat=packet
of which dpbox is widely used in germany, but I have never tried it. I have tried the gw3tmh mailbox a few years back, but it is written for 2.0 kernels and needs to be ported to a recent kernel and ax25-lib/tools.

Morelia might also be interesting for you.

Regards,
Joop PA4TU

I guess my previous mesage regarding the use of AWZnode 0.4-pre2 was not too clear. There are two commands that are relevant:

The 'send' command allows node users (typically incoming packet radio users) to deposit a message. The target of the message is a callsign. That callsign must be registered in node.users, where it is mapped to an existing local Linux user. The message is then delivered to that user as E-mail (or it can be read from the AWZnode mailbox as per below)

The other command is the 'mail' command, which opens a mailbox, very much like the one produced by the Linux 'mail' command. Once again, the packet radio user must be registered in node.users for this command to work.

So, if a completely random ax25 user logs in, all they can do, is send a message to the sysop with 'send <sysop_callsign>, asking to create a Linux account for them, and to make an entry for them in node.users.

The latter maps their call to the Linux user, and allows them to use the AWZnode 'mail' command.

Whilst all of this works, a new caller cannot log in and leave a message for another unregistered ax25 user, which, I guess, is what Michael wanted his PMS/BBS to do. In any case: new ax25 users *can* leave him a message.

Aside from all this, if you are running a multi-port ax25 setup, AWZnode is a worthwhile thing to have. It doesn't gobble up any resources as it is started from ax25d, and gives ax25 users the sort of this they usually want: a cross-port gateway, mheard facilities, telnet, ping etc. I have had it running here for quite a while and it is very well-behaved.

Wilbert, ZL2BSJ

Has there been any update since 0.4-pre2? That came out quite a while ago (and I guess the "pre2" implies that a 0.4 version is imminent).

Bob, N7XY

Bob Nielsen, N7XY                          [email protected]
Bainbridge Island, WA  
IOTA NA-065, USI WA-028S 

Tomi, and others,

OK, useful advice. The first thing I am doing is going to

...to download version 7.04h source. I'll try with that and let you know if I am getting on any better. I have some hope, as there are quite a few versions between what I had and what the latest version there is.i

Michael

Last time I looked (a couple of months ago) 0.4-pre2 was the most recent version. I think it's almost 2 years old now, and seems to be perfectly OK.

Wilbert, ZL2BSJ

OK, so 7.04h is a bit better than 7.03 and I am making some progress. I have fbb working, replaced the dozens of "f6fbb's" and "Toulouse's" in the english.txt file (why not make them variables!), and I can now locally access and send/receive messages using xffbC. (Have not hooked up the radios yet until it is all working).

1. Can I get bulletins via the Internet instead of via radio? Not sure any remaining bbs's in my area to give me a feed.

2. Alternately, what do I do to make it a local message box only (ie no forwarding to ax25 bbs network)? Docs not very clear on this. E.g. what config changes do I make; how do I send bulletinsg just locally, etc.

Any other help gratefully acknowledged.

73 Mike VA3MVW

Hi, Mike:

I'm not sure I understand your questions, but I will try.

Yes.

I am baffled by this statement. I see no way for us to respond
to this statement. If you are not sure, we are probably even less sure. :-| You really ought to be sure, because if there are already pBBSs in your area and they are already getting bulletin feeds, then maybe another brand spanking new pBBS run by a brand spanking new sysop, may not be accepted with
loving arms. :-|

However, if you watch the messages go by on this list server,
you will see, from time to time, offers of bulletin feeds via internet (telnet).

Umm, if you set up no forwarding, it is a
"no forwarding to ax25 bbs network" pbbs. :-|

Uh, yea, I'm not very clear on this either.
If you are not going to forward, why have a 'full featured' pbbs? A 'built into the TNC PMS' does what you have mentioned. So, the only difference is the amount of message storage. :-|

As far as BBS configuration goes,
bulletins are sent locally the same way they are sent distantly.
The pBBS has no concept of distance or geography.

OK Chuck, here we go:

OK. Sorry to ask, but how? I assume you mean wait for the offers by telnet as you mention elsewhere: OK, that suits me fine, I'll wait.

Well, I used to run a bbs on our local VHF LAN frequency with a feed from another local station on UHF: but he has stopped doing packet. I am not sure anyone else still does. Background: I have been asked to run a BBS for local emergency purposes. But while I am at it, bulletins may be a nice addition. Since there is less and less packet I am pretty sure no local sysops will object. The reactions are either positive or "don't care about packet anymore".

I think packet radio is vastly underrated. The reason I am asking all these questions here about BBS's is that local message storage is a nice addition, but IMHO, tcp/ip is the real power. I mean:

- To get bits across my cell phone I pay $0.10 per kB or something. Amateur bits are free.

- When my home cable connection goes down, I get to wait 2-3 weeks for a repair van. The amateur infrastructure is under my own control. And reliable.

- In emergencies, cell and Internet access will go down. My ham packet station will not.

And keep in mind, I am trying to make a contribution! I have many PCs available, and I have static IP's (I can be a gateway), and I have many TNCs and radios available. And stable power, battery backups, good antennas, and as an ex telecomms engineer I understand reliability. (My packet station has now been up for 3 years with zero errors; last reboot 370 days ago at my instigation).

Yet when I ask here, no-one appears interested.

And begging everyone's forgiveness, I am not all that up on packet: my professional life is Internet and tcp/ip, so that I understand. This means I will sometimes ask silly questions and show my ignorance about ax25 and BBS's. Bear with me! ;)

I see - so by default fbbs does not forward? I am handicapped by the fact that all the docs assume you know all this stuff. I used to, 25 years ago:
I knew all bout 300 baud modems, Fidonet etc - but I have forgotten all that stuff and fbb docs are very unforgiving of people who have forgotten stuff they knew a quarter century ago! :-)

Aha, good question. I asked here for simple BBS's: TNC functionality would indeed have done: that is what I was looking for. But as I read it, everyone here said "use FBB, as everyone else does", and indeed there are no other alternatives, it appears. I am a pretty good reader, so point me to the docs and I am happy to read 'em.

(That explains my occasional frustration too. Point me to the docs, and then there are no docs, o0r as in FBB's case, they are hopelessly contradictory and outdated. Leaves guys like me in a bind)

But for TCP/IP I need to set the TNC to KISS, so that built-in functionality is not available...

Cheers, and thx for the help so far,

Michael VA3MVW

Hi, Mike:

I know much more about you now. I can understand better
why you ask the questions that you do.

I ran KA9Q's NET from 1988-1989,
(DOS)WA7MBL from 1989-1990,
(DOS) AA4RE from 1990-2001,
DOS FBB from 1999-2000,
WIN FBB from 2000-2001,
XFBB for awhile in 2001.
I used G8BPQ node code from 1990-2001.

I agree that the FBB docs do not seem to assume that this is
one's first installation. I first tried FBB in last half of 1999. My favorite pBBS was not being supported at that time and the 'y2k' bug was fast approaching. FBB was being supported at that time so I was trying to switch to FBB. The 'docs' seemed to assume that one was upgrading from the old v5.15 (IIRC) and the installer was familiar with the style and configuration. Alas, I was not.
:-(
I had been using G8BPQ node code for a decade and, thankfully,
I could configure FBB to use G8BQP node too. :-)

At first I used the DOS version.
I was startled to find that there were no provisions to 'monitor' what was happening on the ports. ! :-( Gosh, it was so nice to watch a forward session to see what was errant in a script file. FBB was like flying blind! :-(
Later when there were no RF paths in or out of the LAN, I
added the Windows version on another PC and picked up an internet telnet feed.

I did successfully install v7.04h within the last year.
However, I found no human activity and the only nodes I can hear are all single port nodes and they are all on the same frequency (145.01 MHz). I could find no packet on 222-225 MHz nor 440-450 MHz. :-(

What are the Canadian Amateur Radio Rules in regard to
the 802.11b devices? Here in the USA, channels 1 through 6 fall entirely within our 2390 - 2450 MHz band.

Imagine 11 Megabits per second! Quite an upgrade from
300 baud, eh? These 100mw transceivers are selling from 40 USD and up. With the addition of dish antennas (75 USD) they have proven reliable for LOS distances of up to 20 miles. Of course, LOS is the difficult part. :-|

Since you are fluent in TCP/IP and it is 'easy' to
encapsulate AX.25 in IP datagrams, imagine how easy it could be to link network nodes for 1200 baud users. ;-)

HTH, Chuck

Might I suggest Morelia written by Andy, g4xyw.

This is a program written in Python which can be fired off from [x]inetd and fronts all the familiar *nix utilities and presents them to the user in a *nos user interface which if not very powerful is at least well known. To quote from the web page at http://www.tvipug.org

....
Morelia is a front end for a Linux/Unix packet radio system and is modelled after the bbs which is available in the [JT]NOS packages. It is a cross between a NODE and a BBS, as it makes nntp news and mail available to the user as well as DNS, convers, operator chat and other sysop defined features.

Morelia does not really provide any of the actual servers/services but simply makes them available to either AX25 or TCP/IP users.
....

It can be found on the download page of the club web site and overview of its operation is under the club sponsored projects and the only requirements are a means of connecting to it (i.e. AX25 installed and running) and python V2.0 or later

Robin,

Excellent, sound slike just what I need. I am going to check it out (tomorrow; today's a work day!)

73

Hello all,

Here's the issue: any ideas welcome.

Box 1 (client) talks to box 2( bbs server), via ax25. When box2 sends a packet, box 1 (the client) hears it. But when I open by bbs, the first packet is 255 characters long, and is never heard, The next packets are shorter and are always heard. Like this:

This packet from va3oak to va3mvw is sent but never seen by va3mvw:

radio1: fm VA3OAK-1 to VA3MVW ctl I00^ pid=F0(Text) len 250 12:30:33

0000  [FBB-7.04h-AB1FHMRX$]MM-----------------------------------------
0040  -------MHello Michael, on 27-Dec at 12:30MWelcome to The VA3OAK
0080  Message System, located inMOakville, Ontario. For licensed amate
00C0  ur use only.M---------------------------------------------

This one follows and IS always seen:

radio1: fm VA3OAK-1 to VA3MVW ctl I01+ pid=F0(Text) len 47 12:30:33 0000 ---MMVA3OAK (B,LM,R<nr>,S<call>,I,O,X,!,#,?) >

Any ideas?

Michael

This is just a guess, but the lenght of the packet might not be the issue. It could be that the txdelay at VA3OAK is too short, thereby distroying the start of the packet. That would cause behaviour like this. (I'm assuming both packets are sent in one go.)

Tomi Manninen           Internet:  [email protected]
OH2BNS                  AX.25:     [email protected]
KP20ME04                Amprnet:   [email protected]

Tomi,

OK, I shoudl check this. Worth nothing that other transmissions (ax25) do work, and also, va3am was able to connect fine. So perhaps the problem is more likely to be on the va3mvw side. Not sure what to look for but that's where my efforts will now concentrate.
Thanks as always.

Michael

Ok,

That would seem like the txdelay at VA3OAK is enough for it's TX chain but unfortunately on a shared channel you will have to go by the weakest link. It might be that the TX-to-RX switching delay of the TRX or the synch delay of the TNC at VA3MVW might be too long.

Then agin, the problem might be something else just as well...

/Tomi

Tomi Manninen           Internet:  [email protected]
OH2BNS                  AX.25:     [email protected]
KP20ME04                Amprnet:   [email protected]

Dear all,

I will use packet radio only in a room

|pc1|--|sc1|--|ht1| <--/
                      /--> |ht2|--|sc2|--|pc2| 

The hardware and software are:
Personal Computer
Sound Card
HT
RH Linux 7.1 kernel upgraded 2.4.19 and radio packet support
libax25-0.0.9
ax25-apps-0.0.6
ax25-tools-0.0.8
node.0.3.0
soundmodem-0.7

First of all I am using stereo cable to connect 2 machines without ht and checking if the TCP/IP facilities works such as ping, ftp, telnet and ssh. |pc1|--|sc1|--------|sc2|--|pc2|

According to
<http://people.zeelandnet.nl/pa3gcu/installax25.htm> I have installed the ax.25 stuff and node.0.3.0 in pc1 and pc2, and all going well without error message in the directory </usr/local/src>. Than I installed the soundmodem-0.7 in the same directory.

my1com# ifconfig
sm0 Link encap: AMPR AX.25 HWaddr MY1COM

     inet addr: 44.128.1.5 Bcast: 44.128.1.255 Mask:
255.255.255.0
     ..

my2com# ifconfig
sm0 Link encap: AMPR AX.25 HWaddr MY2COM

     inet addr: 44.128.1.8 Bcast: 44.128.1.255 Mask:
255.255.255.0
     ..

Next I have connected sc1 and sc2 using stereo cables as follow:
MIC-sc1|----|SPK-sc2

SPK-sc1|----|MIC-sc2

But the "ping", "ftp", "ssh" and "telnet" commands don't work from pc1 to pc2 and virce virca. [my1com]$ ping my2com
PING my2com.ampr.org (44.128.1.8) from 44.128.1.5:
56(84) bytes of data
From my1com.ampr.org (44.128.1.5): Destination Host unreachable
From my1com.ampr.org (44.128.1.5): Destination Host unreachable
From my1com.ampr.org (44.128.1.5): Destination Host unreachable

--- my2com.ampr.org ping statistics ---
4 packets transmitted, 0 packet received, +3 errors, 100% packet loss

Please any idea what should I do to make "ping", "ftp", "ssh" and "telnet" commands workable and useable?

Thanks
Vivi

__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com

Is sm0 'UP'?? (next line, please)

If not: 'ifconfig sm0 up'

Next: set up the IP routing (man route). See below.

Good idea.
Run 'aumix' and make sure the line input and output are not muted. Make sure the volumes are not right down...

Are the sound modems actually transmitting audio? Check with as pair of headphones. If you blow your brains out, then it's working :-) Are you seeing anything on the interfaces? Try: 'listen -a -c'

You will have to tell your machine what port to send the frame on;

route add 44.128.1.8 sm0

You may also have to set an ARP entry (man arp)

Wilbert, ZL2BSJ

Hi, Vivi:

I'd like to see the output of
route -n

HTH, Chuck

Hello all,

When I use tcp/ip over ax25, I notice that packets are sent very frequently when I am on the keyboard. For some applications (e.g. using a BBS, which is almost character-oriented) that is suitable, but for some, like telnet, it is inappropriate: my telnet sessions are 'de facto' almost line-oriented, so it would be best if packets were sent after a CR *or* after a delay of, say, 500 ms. That would prevent a LOT of unwanted radio traffic. Right now, a 10-letter word can generate 5-10 packets, when one would be enough.

I am sure I am overlooking this, but can I set this? How?

I have of course read the HOWTO etc but no doubt skipped over this. Many thanks for pointing me in the right direction.

Michael va3mvw

In Telnet you can get to the command prompt (usually control-]) then use 'mode line' to switch to line mode. You may have to change some other options after that, such as local echo. There doesn't seem to be a way to force line mode before starting telnet.

Hamish

Hamish Moffatt VK3SB <[email protected]> <[email protected]>

Hi Michael,

An application that can solve your problem is axssh. You can find it at: http://hes.iki.fi/pub/ham/unix/utils/

Description:

Title:          Axssh
Version:        0.1
Entered-date:   
Description:    Linemode wrapper for SSH via Amateur Packet Radio links
Keywords:       ham, amateur, AX.25, TCP/IP
Author:         [email protected] (Joerg Reuter)
Maintained-by:  
Primary-site:   db0bm.automation.fh-aachen.de /pub/incoming/dl1bke
                axssh-0.1.tar.gz

Alternate-site:
Original-site:
Platforms: Linux, probably others as well. Required: libbsd, for
Linux Alpha "libutil", SSH package from ftp.funet.fi, VT100 emulation (Linux console, xterm...)
Copying-policy: GPL

HTH.

Henrique

Once your telnet client connects, hit the escape sequence (typically Control-] ). At the prompt, type 'mode line' to put your client in line mode.

I don't think it's mentioned anywhere. Perhaps a useful addition to the HOWTO?

Incidentally, I had a play with the telnetd server and line mode, some time ago, but found that the line mode option wasn't compiled in. (a pre-compiled binary...) I never got round to getting the sources, but I reckon it would be worth a try to get the daemon to default to line mode as well, for the reasons you mentioned. Anyway, have a look at 'man telnetd' and look for the -k option (from memory).

Hope this helps...

Wilbert, ZL2BSJ

It's been a long while since I have used it, but I recall that using splitscreen instead of telnet (specifying port 23) would give you line mode in addition to the split screen feature. Possibly this could also be done with colrconv, although I haven't tried it.

73,
Bob N7XY

I grabbed a copy of splitscreen and compiled it, but it doesn't appear to support doing this (nor does colrconv). Maybe there was some other split screen telnet client that I used, but I haven't been able to find it (it was several years ago and I don't remember the details).

Bob, N7XY

I'm pretty sure colrconv does support it. There won't be any terminal emulation capabilities that way though so I'm not sure how useful it is...

Tomi Manninen           Internet:  [email protected]
OH2BNS                  AX.25:     [email protected]
KP20ME04                Amprnet:   [email protected]

Yes, I think so ... because

my1com# ifconfig
eth0 ..
lo ..
sm0 Link encap: AMPR AX.25 HWaddr MY1COM

     inet addr: 44.128.1.5 Bcast: 44.128.1.255 Mask:
255.255.255.0
     ..

my2com# ifconfig
eth0 ..
lo ..
sm0 Link encap: AMPR AX.25 HWaddr MY2COM

     inet addr: 44.128.1.8 Bcast: 44.128.1.255 Mask:
255.255.255.0
     ..

The sound modem transmit audio/noise if I connect
spk_out to a speaker
[my1com]$ ping my2com

spk_sc1|-----|speaker| ...bep...bep...

[my1com] route -n
Destination Gateway Genmask Flags
Metric Ref Use Iface
44.128.1.0 0.0.0.0 255.255.255.0 U 0
0 0 sm0
192.168.10.0 0.0.0.0 255.255.255.0 U 0
0 0 eth0
127.0.0.0 0.0.0.0 255.0.0.0 U
0 0 0 lo

[my1com] route add 44.128.1.8 sm0
[my1com] route -n
Destination Gateway Genmask Flags
Metric Ref Use Iface
44.128.1.8 0.0.0.0 255.255.255.255 UH
0 0 0 sm0
44.128.1.0 0.0.0.0 255.255.255.0 U
0 0 0 sm0
192.168.10.0 0.0.0.0 255.255.255.0 U
0 0 0 eth0
127.0.0.0 0.0.0.0 255.0.0.0
U 0 0 0 lo

[my1com] cat /etc/ax25/axports
name callsign speed paclen window
description
sc my1com 1200 255 2
soundcard

[my1com] arp -n 44.128.1.8 -H ax25 -s MY2COM -i
sm0
[my1com] cat /proc/net/arp
IP add HW type Flags HW address
mask Device
44.128.1.8 0x3 0x6 MY2COM-0 *
sm0

[my1com]# ping my2com
PING my2com.ampr.org (44.128.1.8) from 44.128.1.5:
56(84) bytes of data
From my1com.ampr.org (44.128.1.5): Destination Host unreachable
..
[my1com]# listen -a
axconfig: port 1 not active
axconfig: port 2 not active
sc: fm my1com to my2com ctl UI pid=CC(IP) len 84
IP: len 84 44.128.1.5->44.128.1.8 ihl 20 ttl 64 DF prot ICMP
ICMP: type Echo Request id 22789 seq 3840 0000 -..>iD..................| "#$%&'()*+,-./01234567
sc: ....
....

Many thanks for all assistance, but 'ping' is still unuseable ;-(
What should I do please?

Thanks,
Vivi.

__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com

{\rtf1\ansi\ansicpg10001\deff0{\fonttbl{\f0\fnil\fcharset0 Times New Roman;}{\f1\fswiss\fcharset77 Helvetica;}{\f2\fmodern\fcharset77 CourierNewPSMT;}} \viewkind4\uc1\pard\tx560\tx1120\tx1680\tx2240\tx2800\tx3360\tx3920\tx4480\tx5040\tx5600\tx6160\tx6720\lang1033\f0\fs24 [my1com] route -n\par
Destination Gateway Genmask Flags Metric Ref Use Iface\par
44.128.1.0 0.0.0.0 255.255.255.0 U 0 0 0 sm0\par
192.168.10.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0\par
127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0 lo\f1\par
\par

\par
\f0 [my1com] route add 44.128.1.8 sm0\par
[my1com] route -n\par
Destination Gateway Genmask Flags Metric Ref Use Iface\par

44.128.1.8      0.0.0.0     255.255.255.255    UH     0         0      0     sm0\par
44.128.1.0      0.0.0.0     255.255.255.0        U        0         0      0     sm0\par

192.168.10.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0\par
127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0 lo\f1\par
\par

\par
\f0 [my1com] cat /etc/ax25/axports\par
name callsign speed paclen window description\par
sc my1com 1200 255 2 soundcard\par
\par
[my1com] arp -n 44.128.1.8 -H ax25 -s MY2COM -i sm0\par [my1com] cat /proc/net/arp\f1\par
\f0 IP add HW type Flags HW address mask Device\par
44.128.1.8 0x3 0x6 MY2COM-0 * sm0\f2\par
}

Actually either one will make a telnet connection, but neither provides terminal capabilities for telnet. As you said, not quite useful.

Splitscreen works with DXSpider, while colrconv does not.

Bob N7XY

Bob Nielsen, N7XY                          [email protected]
Bainbridge Island, WA  
IOTA NA-065, USI WA-028S 

Fix it, of course.

my1com is OK. It's transmitting. Now fix the other end, along the same lines.

Use 'listen -a' to see if the echo request arrives.

If not, fiddle with the mixer on the receiving end. Make sure the line input is not muted. Wind up the transmitted audio. Check your audio leads. Make sure the machine being pinged has a return route.

Wilbert, ZL2BSJ