Linux-Networking Digest #107, Volume #12 Wed, 4 Aug 99 12:13:31 EDT
Contents:
Can't configure PCMCIA card. ("Jim Tench")
Network cards for both WindowsNT and Linux ([EMAIL PROTECTED])
I can't ping myself!!! (Flavio Poletti)
Re: Linux vs. Wingate
Re: TCP Data Loss? (Paul Chien)
More IPCHAINS question - Help ! (ST)
find dont work on NW volumes, plz help ([EMAIL PROTECTED])
Re: Alarm - can't connect using PPP ("Bill Smith")
Problems with mounted NW volumes. ([EMAIL PROTECTED])
Re: Dial-up backup to DSL ("Greg Boes")
Web server question (BH)
Re: pls help: umount2: Devices or resource busy (S.T. Wong)
Re: It is Linux stupid! (Thaddeus Speed)
Re: Modem hangup during PPP connect in RedHat 6.0 (Kelly A Sigmon)
secure uucp? (Cameron L. Spitzer)
----------------------------------------------------------------------------
From: "Jim Tench" <[EMAIL PROTECTED]>
Subject: Can't configure PCMCIA card.
Date: Wed, 4 Aug 1999 15:05:34 +0100
Help!!!...
I have a P200 MSB laptop (one of the kapok 6100 machines) with a Dlink 660
network card and a Xircom 56K modem. Neither of them appear to be working.
I've recompiled the kernel up to 2.2.0 and tried recompiling the PCMCIA
package, but still no joy. I've set the network up the same way I have in
the past, so I'm assuming I set up something wrong with the new kernel ->
modules. Anybody have a similar set-up and could advise/send sample config
files, so that I could take a look... I'm on the verge of downgrading my
kernel back to 2.0 as I never had any real problems with that.
This brings me onto a thorny subject. I love Linux and have been using it
for the last 4 years (since 0.99), back in the good old days of scratching
ours heads and calculators to get X to work on our 1 meg 386's. Thank god
that era's past. I would dearly love to see Linux make a bigger dent on the
home PC/work station market. I think the reason why it is not doing just
that is problems exactly like the one I am experiencing. I can install NT
and 95/8 on this laptop (and do regularly as I have several swappable
drives) and never have a single issue with any of the hardware, but I always
do with Linux. I think Linux is mature enough and has a big enough software
base to take on the might of the MS home PC's (it is already doing so in the
server market), but for this to happen the install process, needs to be
simpler and more robust. Don't know if any Linux activist are reading this,
but if so, how about it guys?.
Jim
------------------------------
From: [EMAIL PROTECTED]
Subject: Network cards for both WindowsNT and Linux
Date: Wed, 04 Aug 1999 13:59:13 GMT
I want to buy a low-end network card for both WindowsNT and Linux
(RH6.0). One local store Axion Tech seems to have good reputation. But
their tech support says none of their cards have drivers for Linux. I
just can't believe it. Can anyone help? Their NICs page is at
http://www.axiontech.com/cgi-local/hardware.asp?category=Network&subcat
egory=ADAPTERS
I'm particularly interested in their best sellers, i.e. "3COM:3C905BTX,
ETHERLINK XL, 10/100MBPS, PCI, TWISTEDOEM, PCI for $50" or
"D-LINK:DFE-530TX, 10/100 MBPS ETHERNET PCI ADAPTERPNP, NWAY AUTO
NEGOTIATION, FULL DUPLEX, LED STATUS for $19"
Is there one central place on the Web that lists all NICs supported by
Linux?
Thank you very much for help.
Yong
[EMAIL PROTECTED]
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: Flavio Poletti <[EMAIL PROTECTED]>
Subject: I can't ping myself!!!
Date: Wed, 04 Aug 1999 18:44:18 +0200
Reply-To: [EMAIL PROTECTED]
Hi all,
I have a weird problem to solve. I run on a machine whose IP number
(within the LAN) is 194.30.30.52. I assigned it some days ago.
Today, I tried to ping myself:
$ ping 194.30.30.52
but I'm not able to 'see' myself:
ping: sendto: Network is unreachable
Then, I tried to ping 127.0.0.1 - it sounds like I'm no more
127.0.0.1. Is it possible?
I've another linux machine, where all things go ok, and I tried to
look at the configurations file: they are almost the same, in the sense
that where the former has 'polettix', the latter contains 'flaviox'.
I've compared network configurations under the control-panel, too,
and I found an interesting difference going under the 'Interfaces' tab:
my 'lo' interface is misconfigured, that is that it is present, but
contains no IP address (where the 'correct linux machine' contains
127.0.0.1), and is not activated. I tried to click on it but it says it
is not possible to change lo's configuration.
The problem is not to ping myself - it sounds like wget does not
work at all, together with lynx, ecc. On the other side, Netscape goes
well.
How can I solve my problem? Does anyone know anything about it?
Good bye,
Flavio.
PS: I can ping other machines within the LAN. Moreover, I can ping
myself from the other machines.
------------------------------
From: [EMAIL PROTECTED] ()
Subject: Re: Linux vs. Wingate
Date: 4 Aug 1999 13:08:18 GMT
On Wed, 04 Aug 1999 11:54:13 +0200, Raymond Doetjes <[EMAIL PROTECTED]> wrote:
>Managing masq is nothing just 3 lines of code or when you use SuSE setting one
>line to yes and it works.
Perhaps so, but if you want demand-dialing things get a lot more
complicated. After playing with diald for weeks trying to get it
to work properly with a network of Windoze machines, I threw
in the towel.
I finally came to the realization that it was a more effective use
of my time to reboot our Wingate box when needed than to spend hour
after hour, day after day, week after week pouring over thousands of
pages of documentation and articles in an attempt to discover the
mystical incantations that would permit linux to do what Wingate or
Sygate do out of the box with a 15-minute install.
I have no doubt that linux is more stable and more efficient. I just
don't have the time to deal with it for this.
--
Roger Blake
(remove second "g" from address for email)
------------------------------
From: Paul Chien <[EMAIL PROTECTED]>
Subject: Re: TCP Data Loss?
Date: Wed, 04 Aug 1999 16:35:18 +0200
To those who are following this thread,
Alan Jones has helped to point out one bug in the usage of send(), but
the (data loss) problem still occurs.
All help appreciated.
Thanks
Paul
Alun Jones wrote:
>
> In article <[EMAIL PROTECTED]>, Paul Chien
<[EMAIL PROTECTED]>
> wrote:
> > My testing platform was between two Linux machines. So I will be
> > interested to know how the code runs on other platforms, eg between 2
> > Unix machines. But again, maybe there is a bug in my code.
>
> The most obvious note here is that you don't make any consideration in
your
> sending code that the buffer might be filled by part of your message -
i.e.
> that send() returns a value somewhat less than the size of the buffer
passed
> in - in which case, you will need to send the remaining data.
>
> Could this be your problem?
>
> Alun.
That was definitely a bug you pointed out, but it did not solve the
problem. I have changed the server sending code to:
/* Send message header */
slen_acc = 0;
while (slen_acc != MSGADD)
{
if ((slen_act =
send (sock, &msghead[slen_acc], MSGADD - slen_acc, 0)) <
0 )
{
printf ("ERR: statMsgSend\n");
return -1;
}
slen_acc += slen_act;
}
/* Send message data */
slen_acc = 0;
while (slen_acc != slen)
{
if ((slen_act =
send (sock, &data[tlen + slen_acc], slen - slen_acc, 0))
< 0)
{
printf ("ERR: statMsgSend\n");
return -1;
}
slen_acc += slen_act;
}
Thanks again. I will look if there are similar bugs on the receiving
side.
Paul
------------------------------
From: [EMAIL PROTECTED] (ST)
Subject: More IPCHAINS question - Help !
Date: Wed, 04 Aug 1999 14:52:14 GMT
The IPCHAINS HOWTO says:
If your machine is the only connection to an external network, then you
can tell the Linux kernel to reassemble all fragments which pass through
it, by compiling the kernel with IP: always defragment set to `Y'. This
sidesteps the issue neatly.
I can make it a yes by changing the contents of
/proc/sys/ip_always_defrag to be 1 vice 0 (on RH 6.0) and rebooting?
Okay I think I'm getting to the end of configuring my ipchains, but I
still need to decipher section 2.3 of the HOWTO. It says that for the
2.2 kernel I have to have set CONFIG_FIREWALL=y and CONFIG_IP_FIREWALL=y
set. I am assuming that I set these in the /proc/net/ip_fwchains files
and reboot to get them to stick??? Am on the right track here?
thanks
sean
------------------------------
From: [EMAIL PROTECTED]
Subject: find dont work on NW volumes, plz help
Date: Wed, 04 Aug 1999 14:23:32 GMT
Hi.
I use NW 4.11 as a fileserver where I put all our
docs. Linux Red Hat 6.0 is used as a webserver.
My problem is that I have built a bourne shell
script that search a mounted volume on the NW
server for files accessed last X hours. Those
docs will be processed and put into some of the
web documents.
The script works like this:
- I mount the NW volume with ncpmount
- The command "find $MYPATH -mmin -120 -print
[formatting removed]" runs.
(And I do my stuff)
- Script ends with unmounting NW volume
Now... If I have touch'd or edited a file under
$MYPATH(Mounted volume) from Linux the script
reports it, but if I edit the file directly from
win95's mounted disk(T: in this case) the find
command ignores the edited file.
How can that be???
All help or suggestions helps.
Thanks.
P Varga.
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: "Bill Smith" <[EMAIL PROTECTED]>
Subject: Re: Alarm - can't connect using PPP
Date: 4 Aug 1999 09:40:05 -0500
W.G. Unruh <[EMAIL PROTECTED]> wrote in message
news:wgu20.933635816@riemann...
> "Bill Smith" <[EMAIL PROTECTED]> writes:
> Well, it tells you. "Connect script failed" That means that your connect
> script failed. The connect script is what chat runs. It failed. To figure
out why you must also set up debugging, with a -v after the chat.
> If you posted your pppd command together with your chat script we might be
able to see why. Also post the debugging output from /var/log/messages.
Okay, here it is:
pppd command:
/usr/sbin/pppd /dev/ttyS2 57600 debug connect "/usr/sbin/chat -v '' ATZ
OK ATDT6580202 CONNECT '\d\c'" (as per
http://axion.physics.ubc.ca/ppp-linux.html)
output from /var/log/messages:
The output was different the second time - I'm not sure why.
1st time:
Aug 3 13:27:26 localhost kernel: PPP Dynamic channel allocation code
copyright
1995 Caldera, Inc.
Aug 3 13:27:26 localhost kernel: PPP line discipline registered.
Aug 3 13:27:26 localhost kernel: registered device ppp0
Aug 3 13:27:26 localhost pppd[340]: pppd 2.3.3 started by root, uid 0
Aug 3 13:27:27 localhost chat[341]: send (ATZ^M)
Aug 3 13:27:27 localhost chat[341]: expect (OK)
Aug 3 13:28:12 localhost chat[341]: alarm
Aug 3 13:28:12 localhost pppd[340]: Connect script failed
Aug 3 13:28:12 localhost chat[341]: Failed
Aug 3 13:28:13 localhost pppd[340]: Exit.
2nd time:
Aug 3 13:28:23 localhost pppd[343]: pppd 2.3.3 started by root, uid 0
Aug 3 13:28:24 localhost chat[344]: send (ATZ^M)
Aug 3 13:28:25 localhost chat[344]: expect (OK)
Aug 3 13:29:10 localhost chat[344]: alarm
Aug 3 13:29:10 localhost pppd[343]: Connect script failed
Aug 3 13:29:10 localhost chat[344]: Failed
Aug 3 13:29:11 localhost pppd[343]: Exit
I didn't specify a script in the pppd command. Should I have? Or is one
used by default?
> You could also read and follow
> axion.physics.ubc.ca/ppp-linux.html
I encountered these problems while following this guide. I did learn a few
things from it, however.
------------------------------
From: [EMAIL PROTECTED]
Subject: Problems with mounted NW volumes.
Date: Wed, 04 Aug 1999 14:26:21 GMT
Hi.
I use NW 4.11 as a fileserver where I put all our docs. Linux Red Hat
6.0 is used as a webserver.
My problem is that I have built a bourne shell script that search a
mounted volume on the NW server for files accessed last X hours. Those
docs will be processed and put into some of the web documents.
The script works like this:
- I mount the NW volume with ncpmount
- The command "find $MYPATH -mmin -120 -print [formatting removed]"
runs.
(And I do my stuff)
- Script ends with unmounting NW volume
Now... If I have touch'd or edited a file under $MYPATH(Mounted volume)
from Linux the script reports it, but if I edit the file directly from
win95's mounted disk(T: in this case) the find command ignores the
edited file.
How can that be???
All help or suggestions helps.
Thanks.
P Varga.
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: "Greg Boes" <[EMAIL PROTECTED]>
Subject: Re: Dial-up backup to DSL
Date: Wed, 4 Aug 1999 10:10:29 -0500
Anja Hansen wrote in message <7o7kpm$lm8$[EMAIL PROTECTED]>...
>Anyone have any ideas on the best way to have an automatic backup for DSL?
>
>The idea and setup:
> One Slackware Linux box performing IP Masq. for a bunch of Windows
>machine on a LAN. The Linux box would be connected to the Internet via
>DSL, but needs to check somehow to see if the connection is live. If it
>isn't, it needs to automatically dial up, send a page to the
>administrator, and generally keep on with it's business (nothing but a
>gateway).
>
>Any ideas? Thanks,
>-Anja
Check this out, I haven'e used it but ....
subject: nettest 0.9
added by: Rene Chaddock on Mar 02nd 1999, 13:11
license: GPL
category: Console/Networking
homepage: http://zorro.pangea.ca/~renec/nettest.php3
download: http://zorro.pangea.ca/~renec/nettest/nettest-0.9.tar.gz
changelog: http://zorro.pangea.ca/~renec/nettest/CHANGELOG
description:
Nettest is a program which monitors a network connection,
and takes some action (either email, audible notification,
syslog entries, or all of the above) if/when the connection
goes down.
changes:
Supports multiple connections with seperate parameters for each
connection, automatically forks into background, and a few
rcfile parameters have been changed.
|> http://freshmeat.net/news/1999/03/02/920398313.html
------------------------------
From: [EMAIL PROTECTED] (BH)
Crossposted-To: comp.os.linux.help
Subject: Web server question
Date: Wed, 04 Aug 1999 15:05:29 GMT
I'm hoping someone here can help me out with a serious problem. A
while back we were targeted by a malicious spammer who undertook a
massive mailbombing campaign and spam attack on our web messageboards.
We warned the guy several times, but eventually had to contact his
ISP...which got him nailed with a few hundred bucks worth of fines and
kicked from his ISP.
Well the guy is now back online and he's obviously pissed. The only
problem this time is that he's figured out public proxies and is
masking his true IP. Since we can't trace him, our only course of
action is to ban the proxies as he posts. I have actually had real
time battles with this guy blocking IP's as he does something, only to
have him rotate proxies and start over again 1 minute later.
So, does anyone:
A) Know how to trace someone using a proxy and
B) Know how to block all public proxy users from accessing a server.
Any help or suggestions would be GREATLY appreciated.
------------------------------
From: [EMAIL PROTECTED] (S.T. Wong)
Crossposted-To: comp.os.linux.misc
Subject: Re: pls help: umount2: Devices or resource busy
Date: 4 Aug 1999 14:25:10 GMT
Chris Mahmood ([EMAIL PROTECTED]) wrote:
: [EMAIL PROTECTED] (S.T. Wong) writes:
: > We upgraded our PIIs from RH 5.0 to RH 6.0 (kernel 2.2.10) recently. One of
: > them is NFS server that exports filesystems to the rest of RH PCs. However,
: > the NFS server hangs up on shutdown each time with following error :
: >
: > Unmounting file systems umount2: Devices or resource busy
: > umount: /dev/pts: device is busy
: Is the filesystem busy? Are you using hard or soft mounts?
: -ckm
The filesystem is not busy at all. I just disabled Unix98 pty file system in
kernel and comment out the line for /dev/pts. The error is gone now.
Thanks a lot.
Regards,
/ST
------------------------------
From: Thaddeus Speed <[EMAIL PROTECTED]>
Subject: Re: It is Linux stupid!
Date: Wed, 04 Aug 1999 09:05:28 -0600
John N wrote:
> Hello everybody,
>
> I've installed RH5.2 with all the trimmings, but there only one problem.
> I still cannot make my network component work.
>
> Even though I've mouth-to-mouth resucitation, aspirin, and some other stuff,
> I can not revive the dead ...
>
> This looks to me like the Blair Witch Project ... and I'm shaking ...
>
> The problem is this.
> I've downloaded Donald Becker's rtl8391 source code for my SMC1211tx nic,
> but since I'm a DOS/Windows-in-transition-to-a-better-world kind of guy, I'm
> stuck with with the source, cannot compile, finaly I do not know how to
> attach the final .o file to the kernel.
>
> Help please, before I commit harakiri with a razor blade.
>
> John Nunez <:)
>
> P.S. Still loving Linux ... :)
the instructions for compiling the driver are commented at the bottom of the
source. so just cat the source, and type in what it says. It has one entry for
multiple processors and one not, and then another one. Be sure to use the
right one.
after you have the .o file, move it to /lib/modules/[kernel version]/net.(this
is where it is in RH6, you better make sure.)
Then write back if you don't know how to do the conf.modules and
/etc/sysconfig/network-scripts/ifcfg-et0 thing
TIP: if you don't want to mess with the config files directly, use Control
Panel: kernel button for ailiasing the module, and networking for the rest.
This also has an easy "activate" button to save you from modprobe, etc.
Dan.
------------------------------
From: Kelly A Sigmon <[EMAIL PROTECTED]>
Subject: Re: Modem hangup during PPP connect in RedHat 6.0
Date: Wed, 04 Aug 1999 11:05:20 -0400
I found the problem. Turns out there was an extra option in the
/etc/ppp/options file that needed setting for CHAP to work correctly. I
just knew it was an authentication problem. Big sigh of relief.
Thanks for your help!
Kelly
<snip!>
------------------------------
From: [EMAIL PROTECTED] (Cameron L. Spitzer)
Subject: secure uucp?
Date: 4 Aug 1999 15:59:54 GMT
I've been running a uucp mail feed to a couple of users' machines.
It gets us around the problems of SMTP to a dynamic-IP host.
I'm now trying to get my users to switch from rsh/Telnet and rcp/FTP to
ssh and scp, so we won't have sniffable passwords.
But it seems uucp-over-TCP passwords are in the clear.
What's the right way to secure a uucp feed so it can't be sniffed
or wiretapped?
Cameron
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.networking) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Networking Digest
******************************
