Linux-Networking Digest #998, Volume #10 Fri, 30 Apr 99 21:13:36 EDT
Contents:
Samba "dfree command" (kyhm)
Re: Removing IPX module from kernel (Steven Howe)
Help! Linux on the Loose (Rick Kennett)
Re: Multiple IP Addresses ("Curt")
Re: @home/linux/network problem (Tim Garlick)
[Q]: ppp with kernel-2.2.x (Daishi Harada)
Dialup PPP Server Problem (Adam)
Re: Sharing Tape Drive Between Linux and Win9x via Ethernet: Possible? (Johannes
Niess)
Re: DNS (Juergen Heinzl)
Re: AS400 tools and connectivity? (Benedikt Hochstrasser)
Re: Samba (Richard Torkar)
Re: Question: Download Priority on one PPP line ?? (James Youngman)
Re: Netatalk: Two big problems: Contention and Symbolic Links (David Foster)
Re: getRight like software ([EMAIL PROTECTED])
Re: [Q]: ppp with kernel-2.2.x (Daishi Harada)
RedHat 5.1 firewall to Internet through cable-modem. 2 subnets on the lan. (Michael
T)
Re: Firewall with one NIC? ([EMAIL PROTECTED])
Re: Win98 Telenet Problem (Ted Liu)
Re: Dialup PPP Server Problem ("Greg")
----------------------------------------------------------------------------
From: kyhm <[EMAIL PROTECTED]>
Subject: Samba "dfree command"
Date: Fri, 30 Apr 1999 14:38:52 -0700
Does anybody know if samba really supports the "dfree command" param
in smb.conf in current (2.0.3 etc) versions? I've written a custom
dfree script and samba _says_ it's using it, but the numbers are the
same.
I've independently verified that samba is reading the parameter and
that the script is executable, etc. and works. I just continue to get
wrong numbers.
I've now spent the last hour going through the source (smbd/dfree.c,
smbd/quota.c, etc) and none of these functions ever seem to reference
the "dfree command" value. Could somebody suggest where to look? I'm
guessing samba does a popen() to run the script, but I can't find it in
the "proper" source files...
Any help greatly appreciated, while I've still got some hair left. ;-)
--
Morgan Hughes
C programmer and highly caffeinated mammal
[EMAIL PROTECTED]
------------------------------
From: Steven Howe <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Re: Removing IPX module from kernel
Date: Fri, 30 Apr 1999 22:29:40 +0000
==============7A6D087BE813E349401C2886
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
[EMAIL PROTECTED] wrote:
> Does anyone know how to remove the IPX module from the kernel without all
> hell breaking lose? The kernel seems fine normally, but when I change the ONE
> item: IPX and remove the "M" when I "make menuconfig", I get this slew of
> errors when I run a depmod -a:
>
> /lib/modules/2.0.36/misc/appletalk.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/fmv18x.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/eql.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/8390.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/eth16i.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/ni65.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/ni52.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/ewrk3.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/depca.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/eepro.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/eexpress.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/at1700.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/lance.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/dummy.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/plip.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/e2100.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/hp-plus.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/hp.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/ne.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/hp100.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/new_tunnel.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/shaper.o: unresolved symbol(s)
> /lib/modules/2.0.36/net/ppp.o: unresolved symbol(s)
> /lib/modules/2.0.36/ipv4/ip_alias.o: unresolved symbol(s)
> /lib/modules/2.0.36/ipv4/ipip.o: unresolved symbol(s)
> /lib/modules/2.0.36/ipv4/rarp.o: unresolved symbol(s)
>
> Why the hell is it doing this?
>
> -----------== Posted via Deja News, The Discussion Network ==----------
> http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
being more verbose then most..
echo of the modules list has compiled with ipx That's why they are bitching.
After building
you kernel did you, remove the old modules, and install the new ones?
make bzImage
make modules
pushd /lib/modules
mv 2.0.36 2.0.36.old
popd
make modules_install
cp System.map /boot/System.map-2.0.36
ln -s /boot/System.map-2.036 /boot/System.map
/sbin/depmod -a 2.0.36 -m /boot/System.map-2.0.36
This should clear up the messages. Turning ipx off in /etc/conf.modules is also
a good
idea (review /usr/src/linux/Documentation/modules.txt)
--
Steven Howe
[EMAIL PROTECTED]
==============7A6D087BE813E349401C2886
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit
<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
[EMAIL PROTECTED] wrote:
<blockquote TYPE=CITE>Does anyone know how to remove the IPX module from
the kernel without all
<br>hell breaking lose? The kernel seems fine normally, but when I change
the ONE
<br>item: IPX and remove the "M" when I "make menuconfig", I get this slew
of
<br>errors when I run a depmod -a:
<p>/lib/modules/2.0.36/misc/appletalk.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/fmv18x.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/eql.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/8390.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/eth16i.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/ni65.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/ni52.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/ewrk3.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/depca.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/eepro.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/eexpress.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/at1700.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/lance.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/dummy.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/plip.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/e2100.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/hp-plus.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/hp.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/ne.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/hp100.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/new_tunnel.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/shaper.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/net/ppp.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/ipv4/ip_alias.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/ipv4/ipip.o: unresolved symbol(s)
<br>/lib/modules/2.0.36/ipv4/rarp.o: unresolved symbol(s)
<p>Why the hell is it doing this?
<p>-----------== Posted via Deja News, The Discussion Network ==----------
<br><a
href="http://www.dejanews.com/">http://www.dejanews.com/</a>
Search, Read, Discuss, or Start Your Own</blockquote>
being more verbose then most..
<br>echo of the modules list has compiled with ipx That's why they are
bitching. After building
<br>you kernel did you, remove the old modules, and install the new ones?
<p>make bzImage
<br>make modules
<br>pushd /lib/modules
<br>mv 2.0.36 2.0.36.old
<br>popd
<br>make modules_install
<br>cp System.map /boot/System.map-2.0.36
<br>ln -s /boot/System.map-2.036 /boot/System.map
<br>/sbin/depmod -a 2.0.36 -m /boot/System.map-2.0.36
<p>This should clear up the messages. Turning ipx off in /etc/conf.modules
is also a good
<br>idea (review /usr/src/linux/Documentation/modules.txt)
<br>
<pre>--
Steven Howe
[EMAIL PROTECTED]</pre>
</html>
==============7A6D087BE813E349401C2886==
------------------------------
From: Rick Kennett <[EMAIL PROTECTED]>
Subject: Help! Linux on the Loose
Date: Fri, 30 Apr 1999 15:02:40 -0400
Hi all,
Well we have cast off the Microsoft shackles and fired up a Linux mail
server. We feel real good about this but we are experiencing a problem
and have yet to find an answer. Perhaps somone here might help. The
issue is that on our network, most desktops, when looking to retrieve
mail, connect to the Linux box, send the login info, and get the mail
lickety split. There are however a small number of desktops that quickly
make the connection but take forever to send login information. We
believe that we have the machines configured exactly the same way and
are puzzled as to whether this is a Linux issue or a networking issue.
Hence this post to linux.networking. Any clues as to what might be going
wrong?
Thanks Much
Rick
------------------------------
Reply-To: "Curt" <[EMAIL PROTECTED]>
From: "Curt" <[EMAIL PROTECTED]>
Subject: Re: Multiple IP Addresses
Date: Fri, 30 Apr 1999 17:54:27 -0500
add the following to your /etc/rc.d/rc.local file
/sbin/ifconfig eth0:0 aaa.bbb.ccc.ddd
/sbin/route add -host aaa.bbb.ccc.ddd dev eth0:0
(with the new IP of course)
dan <[EMAIL PROTECTED]> wrote in message
news:oB7W2.6613$[EMAIL PROTECTED]...
> Hi all
> I'm using Linux 5.1 and trying to get multiple IP addresses on it for
> the sake of a webserver...... does anyone know the file I should be
editing?
> Or whats the best way to go about this???
> Thanks
> Dan
> [EMAIL PROTECTED]
>
>
------------------------------
From: Tim Garlick <[EMAIL PROTECTED]>
Subject: Re: @home/linux/network problem
Date: Fri, 30 Apr 1999 14:13:48 -0500
Okay, well the problem turned out to be the Netgear enet card. When I
removed it, the network started working just fine. I'm not sure why it
was interfering. I now have to go figure out how to build a new kernel
that will use the second card as well.
-Tim Garlick.
------------------------------
From: Daishi Harada <[EMAIL PROTECTED]>
Subject: [Q]: ppp with kernel-2.2.x
Date: 30 Apr 1999 16:01:15 -0700
Hi,
I am having the following problem with using ppp
together with kernel 2.2.7 (this problem also
existed with 2.2.5):
The connection comes up fine, and I can do things
such as nslookup, ping, traceroute to the rest of
the world, but I can't ssh, ftp, http. My understanding
from others who have posted similar problems is that
this means that udp is passing through but tcp isn't.
This appears to be a problem specific to the combination
of ppp with 2.2.x, since the combinations of ppp w/ 2.0.x
and ethernet w/ 2.2.x work without any problems.
It was suggested to me that perhaps tcp header compression
could be the culprit, and I've added 'noccp' to my
pppoptions. This makes the connect transaction which
occurs between me and the remote machine (at least as
seen through /var/log/syslog) appear the same between
ppp w/2.0.x and ppp w/2.2.x. However, this hasn't
fixed the problem.
I am in the process of trying out various other flags
for pppd, but at this point I'm mostly shooting in the
dark. Might there be other issues that I haven't
considered?
Any help/advice would be appreciated,
Daishi
ps. I'm using a debian system somewhere in between
slink and potato.
------------------------------
From: [EMAIL PROTECTED] (Adam)
Subject: Dialup PPP Server Problem
Date: Fri, 30 Apr 1999 23:08:49 GMT
I have a RH5.2 machine acting as a dial in server. The server is part
of a local network that has access to the internet. From the server, I
can see all the other machines on the local network, and, access the
internet.
I configured this Linux box to act as a dial in server for Windows
95/98/NT machines. Here is a list of what I have done so far:
1. Configured the Kernel to support PPP & IP_Forwarding
2. echo "1" > /proc/sys/net/ipv4/ip_forward
3. Setup mgetty and edit /etc/inittab with:
S0:345:respawn:/sbin/mgetty ttyS0 -D /dev/ttyS0
4. Issued kill -1 1. The modem inits and will accept calls
5. chmod u+s /usr/sbin/pppd
6. /etc/bashrc: alias ppp="exec /usr/sbin/pppd -detach"
7. /etc/ppp/options:
lock
-detach
proxyarp
asyncmap 0
modem
debug
kdebug 7
crtscts
proxyarp
mtu 1500
mru 1500
8. /etc/ppp/options.ttyS0
login
proxyarp
192.168.140.3:192.168.140.12
netmask 255.255.255.192
ms-dns 207.239.163.30
ms-dns 207.239.163.195
idle 1800
9. /etc/mgetty+sendfax/login.config
/AutoPPP/ - a_ppp /usr/sbin/pppd auth -chap +pap login proxyarp
kdebug 7 debug
Additional configuration stuff...
10. The Linux server is on the same subnet as the IP addresses
assigned to the dial in user.
11. I have checked IP_Forwarding and it appears to be working because
I can ping the Linux server, and any other computer on the network
with the Linux server... just nothing beyond the local network (ie.
can't reach the internet). All machines on the local network
(including the server) CAN see the internet.
12. The logs have no errors listed
13. I had some confusion about setting defaultroute in my
/ppp/options. I am assuming defaultroute is a setting used on the
client side and not the server side. Anyway, I've tried it with and
without defaultroute in the options file and it made no difference.
To sum it all up...
Everything works exactly as it should... except... I can't see
anything beyond the server's local network.
I'm now wondering if it has something to do with a default gateway
setting for the ppp dial in user. I've not seen anything listed in
the HOWTO's about setting an options like this.
Also, I'm wondering if I need to somehow bind the ppp connections to
the ethernet connection the server uses.
If anyone can help I would really appreciate it.
Thanks
Adam
------------------------------
From: [EMAIL PROTECTED] (Johannes Niess)
Subject: Re: Sharing Tape Drive Between Linux and Win9x via Ethernet: Possible?
Date: Sat, 01 May 1999 00:39:52 GMT
[EMAIL PROTECTED] (Lewis) wrote:
>Folks:
> A few questions if I can:
>1. Does anybody know if it is possible to share a tape drive between a
>Linux box (I'll probably use Redhat v5.2) and a Win9x box, and use the
>tape drive to back up both machines? The machines will be connected
>via ethernet. Could Samba be used for this purpose?
>2. Where would be the best place to place the tape drive? In the
>Linux box, or the Win9x box?
>3. Is there any great benefit of having a SCSI tape drive over an
>internal IDE tape drive for the support?
>(Obviously, the IDE is cheaper, and is supported. I'd really like to
>get a parallel port external tape drive, but I see that is not
>supported by Linux. Or is it?)
>I'm going to be putting Linux on an old Pentium 166MMX, once we get a
>newer computer. I'm presently specifying parts for the new computer,
>hence the questions about tape drives.
>Thanks in advance for all the help.
>Roger
Check out www.amanda.org for a great GPL backup solution.
------------------------------
From: [EMAIL PROTECTED] (Juergen Heinzl)
Subject: Re: DNS
Date: Fri, 30 Apr 1999 23:07:25 GMT
In article <7gc26u$hhq$[EMAIL PROTECTED]>, Gilbert wrote:
>Hi:
>
>Until 2 days ago, our DNS Server was working well. Now when I try to start
>it I got the following message: can't open '/etc/named.conf'
>I never had this file before. I have named.boot
>
>any idea ?
bind-8.x uses /etc/named.conf. See http://www.isc.org/ for more
and a very good online documentation for bind.
Minor note, on a glibc-2.1.x system bind-8.2 cannot be compiled
without some minor changes.
Setting up the named.conf should not be too complicated but
see above, get the documentation at least, it is quite good.
Cheers,
Juergen
--
\ Real name : J�rgen Heinzl \ no flames /
\ EMail Private : [EMAIL PROTECTED] \ send money instead /
------------------------------
From: [EMAIL PROTECTED] (Benedikt Hochstrasser)
Subject: Re: AS400 tools and connectivity?
Date: Fri, 30 Apr 1999 22:49:12 GMT
Check with IBM - they've recently jumped on the Linux bandwagon.
There's also some TN5250 emulations around.
Keep in mind that your AS first has to run on TCP/IP...
Ben
In article <[EMAIL PROTECTED]>,
email:[EMAIL PROTECTED] (James Tyrrell) wrote:
> I was wondering if therre were any Linux programs or tools for
> connectivity with AS400 platforms?
>
> Thanks.
>
> Jim.
Regards, Ben
(mind the bogus address [EMAIL PROTECTED] <spammers ahoy>; my real maildomain
is
hotmail.com and my userid there is b_hochstrasser)
------------------------------
From: Richard Torkar <[EMAIL PROTECTED]>
Subject: Re: Samba
Date: Sun, 25 Apr 1999 12:56:35 +0200
v4cal wrote:
>
> Is there a graphic version of samba
>
Look at the link below in my signature for where to get LinNeighborhood.
And when you configure samba you can always use SWAT that comes with
samba.
Read the manuals.
Richard Torkar
--
http://milkyway.thn.htu.se/~ds98rito/
Hoping the problem magically goes away
by ignoring it is the "microsoft approach to programming"
and should never be allowed.
(Linus Torvalds)
============================================================
PGP Key ID / PGP Key Fingerprint:
D40BA0AD / C7 5D A3 B7 1A 28 7E CE E6 41 82 AE E6 EC 20 D1
------------------------------
From: James Youngman <[EMAIL PROTECTED]>
Subject: Re: Question: Download Priority on one PPP line ??
Date: 25 Apr 1999 11:05:14 +0100
[EMAIL PROTECTED] (francois) writes:
> Hello world !
>
> Problem:
> When I'm downloading http or ftp files and when I want to surf during
> this time, it's very slow
> e.g. ppp daemon did not dynamically adjust priority
It's not a priority thing, it's just that the incoming packets are
large so you often have to wait over a second for the next packet.
Reduce your MRU value.
--
ACTUALLY reachable as @free-lunch.demon.(whitehouse)co.uk:james+usenet
------------------------------
From: [EMAIL PROTECTED] (David Foster)
Crossposted-To: comp.protocols.appletalk
Subject: Re: Netatalk: Two big problems: Contention and Symbolic Links
Date: 30 Apr 1999 19:49:40 -0400
In <7gceis$fs6$[EMAIL PROTECTED]> [EMAIL PROTECTED] (Rod Smith)
writes:
>Now, let me ask this: Have you tried doing this with a Mac server rather
>than a Linux server? If the same thing happens, then it's likely a file
>locking issue in At Ease itself, not in the server software. If not, then
>netatalk is probably handling file locking differently than MacOS, thus
>resulting in the problem.
IIRC, netatalk doesn't do file locking (or rather, doesn't implement
the LockBytes stuff). It's possible this is where the problem is.
Still, At Ease is a fairly popular package for the Mac, and I've never
heard any discussion of using it with netatalk. Does anyone out
there do this? (I'm considering this exact setup at one site within
the next few months, so I'd love to hear any comments).
--
--
David Foster | As a philosophy of nothingness, it has a
[EMAIL PROTECTED] | certain attraction to people of our time
finger for PGP key | -- Fides et Ratio
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: alt.uu.comp.os.linux.questions
Subject: Re: getRight like software
Date: Fri, 30 Apr 1999 23:58:43 GMT
Hi oneself,
Try Caitoo at http://devel-home.kde.org/kget/
It's supposed to be like Gozilla and Getright.
I haven't tried it myself, as I'm still setting up my Linux PC
Bill
[EMAIL PROTECTED]
On Sat, 01 May 1999 01:16:11 +0300, oneself <[EMAIL PROTECTED]> wrote:
>Hi!
>
>I'm looking for a download manager (like getRight, and goZilla)
>which supports resuming and other features.
>Is there something like that anywhere ?
>
>
> thanks.
>
------------------------------
From: Daishi Harada <[EMAIL PROTECTED]>
Subject: Re: [Q]: ppp with kernel-2.2.x
Date: Fri, 30 Apr 1999 17:03:45 -0700
I wrote:
> It was suggested to me that perhaps tcp header compression
> could be the culprit, and I've added 'noccp' to my
> pppoptions. This makes the connect transaction which
> occurs between me and the remote machine (at least as
> seen through /var/log/syslog) appear the same between
> ppp w/2.0.x and ppp w/2.2.x. However, this hasn't
> fixed the problem.
I didn't think carefully enough about what the suggestion
actually was; adding 'novj' as an option to pppd solved the
problem.
OTOH, it still doesn't make sense to me why this broke.
Is it plausible that it's a 2.2.x bug (in the VJ code)?
------------------------------
From: [EMAIL PROTECTED] (Michael T)
Subject: RedHat 5.1 firewall to Internet through cable-modem. 2 subnets on the lan.
Date: Sat, 01 May 1999 00:09:00 GMT
Reply-To: [EMAIL PROTECTED]
2 NT PCs, each on its own network accessing the internet through the
Linux box connected via cable-modem.
I have 3 NICs in the box all working fine. Can ping all the cards from
anywhere on the LAN. I can ping the Internet from the Linux box
Cannot ping from NT to internet!
However, http proxy (Squid) is working.
I have masquerading setup from both subnets to the Internet with
ipfwadm:
ipfwadm -F -a masquerade -P tcp -S 192.168.100.0/255.255.255.0 -D
0.0.0.0/0
same for udp and the other subnet
One NT box is IP 192.168.101.50
MASK 255.255.255.0
GW 192.168.101.1
TIA for any help
Michael
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Firewall with one NIC?
Date: Sun, 25 Apr 1999 10:54:36 GMT
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (Michael Schurr) wrote:
> Due to security-consideration I wanna install on my computer a
> firewall.
For one computer, there is no need for a firewall. Firewalls are only useful
when you have a *network* you're trying to protect. For a single machine you
can use tcpserver or some equivalent to achieve the desired security. I
believe there is an RPM for that [and other packages], but I don't know of
any places from which you can download them. I'm actually a FreeBSD-head
myself. You can download the source and compile yourself, if you like:
http://pobox.com/~djb/ucspi-tcp.html
> There only shall work http, ftp and mail (-sending and -receiving).
Unless you want to restrict access to the above ports to a limited set of IP
addresses, you can achieve your desired level of security via the following
steps:
comment out *everything* in /etc/inetd.conf
remove or chmod 0 all rc scripts that start up network services [1]
install sshd [and it's rc scripts]
install qmail [and it's rc scripts]
install apache [ditto]
reboot your machine
[1] - Use `netstat -a` to get some idea what's currently running
You'll need to use an ssh client to get into your machine remotely [may be a
problem for some, non-UNIX, users -- but not impossible], and you'll need to
use `scp` rather than `ftp`, but that will provide an even stronger level of
security than you are currently asking for. sendmail is known to have tons
of security holes, and new ones seem to be popping up every few months or so,
which is why I suggest qmail [although security is only _one_ of the benefits
qmail offers]. qmail also comes with instructions on how to install and
configure tcpserver.
> Thanks a lot...
HTH. HAND.
-Bill Clark
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: Ted Liu <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: Re: Win98 Telenet Problem
Date: Fri, 30 Apr 1999 20:09:52 -0400
I just found the answer to the problem. Win98 telnet client put a period
after the sever IP number (I don't know why win98 would do that) - in my
case, telnet trying connect to 192.168.1.1. not 192.168.1.1 So the
telnet client would hang. It is still a puzzle to me why if win95 machine
is on, telnet to the IP number with the period will succeed after all.
To John Xu: John, I am not exactly sure about your situation, but it is
worth of looking into the IP number the telnet client is trying to connect
to.
john xu wrote:
> I have the similar rpoblem with that you have.
> Using the same username and password and from any machine except a win98,
> I can telnet to this Linux box and logon. However, from this win98, I can
> ftp
> to one of our ISP machine (SunSparc) but can not log on to WINNT and the
> linux box. I tried add registry EnablePlainTextPassword and set value to
> 0x01
> but the problem can not be solved. I do think the problem is Win98
> specific but
> I could not find solution.
>
> If you got a way to fix it, let me know please.
> My email address is [EMAIL PROTECTED]
>
> Thanks
>
> [EMAIL PROTECTED] wrote:
>
> > I am having this weird problem and couldn't find an answer to it.
> >
> > A win98 machine in a small network (1 linux, 2 win95 and 1 win98)
> > couldn't not telnet into the linux server, unless one of the win95
> > machines were on at the same time. Otherwise, telnet would just hang
> > there.
> >
> > The linux server is running RedHat 5.2 and serves as a file server
> > (Samba) and a gateway/firewall to Internet (ip masquerading). All the
> > machines have their own private ip numbers.
> >
> > I don't think win98 telnet client is bad (or could be??) since it will
> > connect to the server if a win95 box is on.
> >
> > By the way, I have no telnet problems with the two win95 machines.
> >
> > Any clues? Thanks.
> > Ted Liu
------------------------------
From: "Greg" <[EMAIL PROTECTED]>
Subject: Re: Dialup PPP Server Problem
Date: Fri, 30 Apr 1999 20:12:17 -0400
Hello Adam,
Try setting the ip masquerading.
do you need the proxyarp in the options file ??
What is the options.ttys0 used for ??
the netmask of 255.255.255.0 should be ok.
Greg.
Adam wrote in message <[EMAIL PROTECTED]>...
>I have a RH5.2 machine acting as a dial in server. The server is part
>of a local network that has access to the internet. From the server, I
>can see all the other machines on the local network, and, access the
>internet.
>
>I configured this Linux box to act as a dial in server for Windows
>95/98/NT machines. Here is a list of what I have done so far:
>
>1. Configured the Kernel to support PPP & IP_Forwarding
>
>2. echo "1" > /proc/sys/net/ipv4/ip_forward
>
>3. Setup mgetty and edit /etc/inittab with:
> S0:345:respawn:/sbin/mgetty ttyS0 -D /dev/ttyS0
>
>4. Issued kill -1 1. The modem inits and will accept calls
>
>5. chmod u+s /usr/sbin/pppd
>
>6. /etc/bashrc: alias ppp="exec /usr/sbin/pppd -detach"
>
>7. /etc/ppp/options:
>
>lock
>-detach
>proxyarp
>asyncmap 0
>modem
>debug
>kdebug 7
>crtscts
>proxyarp
>mtu 1500
>mru 1500
>
>8. /etc/ppp/options.ttyS0
>
>login
>proxyarp
>192.168.140.3:192.168.140.12
>netmask 255.255.255.192
>ms-dns 207.239.163.30
>ms-dns 207.239.163.195
>idle 1800
>
>9. /etc/mgetty+sendfax/login.config
>
>/AutoPPP/ - a_ppp /usr/sbin/pppd auth -chap +pap login proxyarp
>kdebug 7 debug
>
>Additional configuration stuff...
>
>10. The Linux server is on the same subnet as the IP addresses
>assigned to the dial in user.
>
>11. I have checked IP_Forwarding and it appears to be working because
>I can ping the Linux server, and any other computer on the network
>with the Linux server... just nothing beyond the local network (ie.
>can't reach the internet). All machines on the local network
>(including the server) CAN see the internet.
>
>12. The logs have no errors listed
>
>13. I had some confusion about setting defaultroute in my
>/ppp/options. I am assuming defaultroute is a setting used on the
>client side and not the server side. Anyway, I've tried it with and
>without defaultroute in the options file and it made no difference.
>
>To sum it all up...
>
>Everything works exactly as it should... except... I can't see
>anything beyond the server's local network.
>
>I'm now wondering if it has something to do with a default gateway
>setting for the ppp dial in user. I've not seen anything listed in
>the HOWTO's about setting an options like this.
>
>Also, I'm wondering if I need to somehow bind the ppp connections to
>the ethernet connection the server uses.
>
>If anyone can help I would really appreciate it.
>
>Thanks
>
>Adam
------------------------------
** 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
******************************
