Linux-Networking Digest #116, Volume #12 Thu, 5 Aug 99 01:13:42 EDT
Contents:
Re: dhcpcd, RH/Mandrake 6.0, and @home ("Jeff")
Linux - @home cable modem service wierdness (Vincent)
Re: Is there an easy way to change user password from desktop (lorax)
Re: Adding a default route (lorax)
Re: HELP: How do I set up a caching DNS server? (lorax)
Re: 486 to linux box? ([EMAIL PROTECTED])
Re: Newbie needs help with DNS: Can't get host lookup (lorax)
Getting the penguinpowered script to run ("Gene Heskett")
Re: users can't mount cdrom (Daniel Forester)
Re: Qpopper authentication problems ([EMAIL PROTECTED])
Re: Linux driver for 3Com NIC: (Vidar Andresen)
How to Configure for Realtime Streaming (RTSP) (Richard Petty)
Re: Rtl8139.c drive not compiling (Vidar Andresen)
Re: 3c905b works at 100base? ("guru meditation")
Re: MODEM DIAL-IN PROBLEM ("Santiago Fisher")
Power Mac G3 dropping packets (Brian Rectanus)
Re: POP3 (Chris Mahmood)
Re: 486 to linux box? (Roger Plant)
Re: IPChains Help Needed (Greg)
Re: Help with RH6 & Routing Problem ("Santiago Fisher")
Re: ARP replies w/ >1 interface on same network ("Santiago Fisher")
----------------------------------------------------------------------------
From: "Jeff" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.redhat,alt.os.linux.mandrake
Subject: Re: dhcpcd, RH/Mandrake 6.0, and @home
Date: Wed, 4 Aug 1999 22:01:57 -0400
On a somewhat related note, I was using a RH5.0 build with ipfwadm for a
masq firewall that worked perfectly using modem/pppd and added dhcp from
build 0.65?. I added another NIC and brought up shaw@home. Got my number and
all was fine except...msn was the only HTTP??? All other services worked,
FTP was fast but intermittent. I've since installed 6.0 and I'm new to
ipchains. Looks like it's the same command structure as ipfwadm???
Having trouble with the ewrk3.o file. Worked for the same io/irq with 5.0???
(That is my fish to fry....:) )
Anyone got a ipchains script to bring up two nics with @Home? I'll have time
this weekend to get two nics up in 6.0 but I would like some shortcuts for
the firewall script. The rest can be delt with after I can use 0.0.0.0/0
HTTP
stressed <[EMAIL PROTECTED]> wrote in message
news:HgOp3.33100$[EMAIL PROTECTED]...
> I had freaky results using anything at all to configure for @Home during
> boot. This was with RH 5.2 and with 6.0.
>
> If anything I "play" with comes close to a base I/O or tampers with any
> resources near a NIC, the NIC init can fail and driver modules and related
> parameters get all screwed up. Not to mention timeouts. But then again,
I'm
> using very old equipment.
>
> I recently added a vortex card (3c59x.o) and I load all of my NIC modules
> manually through "/etc/rc.d/rc.local". Nothing gets loaded until the
system
> comes up "clean". Here's what I have:
>
> ##At bottom of /etc/rc.d/rc.local
> modprobe 3c59x (loads driver module for 3c590 card)
> dhcpcd -h DNHOSTNAME eth0 (configures 3c590 gateway device a la @Home
DHCP)
> modprobe 3c509 irq=10 (loads module for 3c509 card)
> ifconfig eth1 192.168.1.1 netmask 255.255.255.0 up (configures LAN card)
> route add -net 192.168.1.0 netmask 255.255.255.0 (adds local network to
> route table for v5.2)
>
> Whenever I loaded things this way it was reliable.
>
> When I chose to let the kernel try at boot - it was unreliable at best.
> Then, when I tried to undo the "kernel's way", it got crazy. For instance,
> if I went to X and started kernel config, then removed "eth" modules,
they'd
> reappear at next boot. This is because they were still referenced in
> "/etc/conf.modules". If the order of the cards wasn't right or the order
of
> configuration wasn't right - look out. And if initialization failed for
say
> eth0, then Linux would just move on to configure the next card as eth0,
> which should have actually been eth1, see what I mean? The manual way I
> listed above seems best for @Home.
>
> Also, I had to get a later dhcpcd for RH 6.0. I was using 0.70 and RH 6.0
> needs a later version. I read about a problem where pump and/or dhcpcd
> (can't remember which) couldn't get complete IP information from NT DHCP
> servers. The manual way above leaves pump out of the picture and dhcpcd
does
> the trick nicely.
>
> For the record, my "PC" modem light only goes on when the card is actually
> active. If dhcpcd fails, the light goes out on the modem and the card's
> "dead". With a Windows environment, the light stays on on the modem no
> matter what - as long as the TP cable is attached to both the modem and
the
> computer and the card has power. This can be misleading.
>
> I have a short paper I started to write but never finished awhile back. I
> don't know how accurate it really is, but it helped a lot of @Home users
get
> online with RHL 5.2. It's at
> http://members.aol.com/papadorker/cable-home.htm if you want to check it
> out.
>
> Hope this helps
>
> [EMAIL PROTECTED]
>
>
>
>
>
>
>
>
>
>
>
>
>
>
------------------------------
From: [EMAIL PROTECTED] (Vincent)
Subject: Linux - @home cable modem service wierdness
Date: Wed, 4 Aug 1999 23:04:13 -0400
This one's strange. When connected to the cable modem directly from
Win98 or WinNT machines, the cable modem ISP @home's mail server is
accessible. I can also connect to erol's and Netcom's mail servers.
But, when I connect to the internet from either Windows machine through a
gateway Linux machine running RedHat 5.2, I cannot connect to @home's
mail server. Netcom's and erol's mail servers are still accessible.
Also, the browser's home page is different, even though no changes were
made to it.
the @home tech support person told me that its mail server was down,
what I was experience was just a coincidence. However, needless to say,
when I removed the Linux machine, and connected directly to the cable
modem again, our original home page loaded, and we were able to get to
@home's mail server again.
I'm using ipfwadm rules to enable ip_masquerading.
Anybody have an explanation for this? Thanks.
------------------------------
From: lorax <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.help,comp.os.linux.setup
Subject: Re: Is there an easy way to change user password from desktop
Date: 5 Aug 1999 02:24:25 GMT
In comp.os.linux.networking Tim Gibson <[EMAIL PROTECTED]> wrote:
: I have a RH 5.2 (soon to be Mandrake 6.0) box and about 15 Win98 clients
: that use it for Internet, Samba, etc. They all log into it using their
: Linux username and passwd. What I would like is an easy way to have them
: change their passwd from their own machine without having to log into Linux
: thru telnet. They afraid of the command line and don't know what to do once
: that $ is staring at them. Anyway, is there a utility or a script I could
: run from a webpage that would allow them to use something other than
: 'telnet' and 'passwd' to change their passwd. Thanks.
I would suggest written a perl script to do through a browser, using .htaccess
files to verify users.
------------------------------
From: lorax <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Re: Adding a default route
Date: 5 Aug 1999 02:20:31 GMT
In comp.os.linux.networking Nitin G. <[EMAIL PROTECTED]> wrote:
: I am running a RedHat 6.0 system using dhcp. I cannot seem to get the
: default route configured. I can do a:
: route add default dev eth0
: to add the refault route. I wanted to have this added automatically at
: bootup. which file to I update to accomodate this?
: Thanks,
make sure these lines are in /etc/sysconfig/network
GATEWAYDEV=eth0
GATEWAY=x.x.x.x
------------------------------
From: lorax <[EMAIL PROTECTED]>
Subject: Re: HELP: How do I set up a caching DNS server?
Date: 5 Aug 1999 02:40:11 GMT
Alexander Atkin <[EMAIL PROTECTED]> wrote:
: I have a small LAN with a Linux router allowing it to access the
: internet through my V.90 modem.
: How could I set up a cachine DNS server to help speed up my access as my
: ISP has a very slow DNS server sometimes and so im sure it would double
: the speed of my access if the router knew the IP addresses of my
: favourite sites.
: I could put them in manually into hosts but that would be rather a drag
: as I have quite a few favourite sites and some of them have changed ISPs
: in the past so I dont want to have to keep reconfiguring it.
: Thanks,
: Alex.
Considering you have bind installed,
/etc/rc.d/init.d/named.init start
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: 486 to linux box?
Date: Thu, 05 Aug 1999 03:23:28 GMT
First, yes there is a way to connect a serial cable...but it's not the
easiest way, and you will not be able to do what you're asking. Plus,
in order to use a serial connection in Windows 3.1, you must be running
DOS 6.0 or later.
The recommended way is to buy a couple of NIC cards (roughly about $30
each). Then take a crossover cable and connect the two computers. You
will need to run/install the server app for Linux. Getting Windows 3.1
to see the interface will be difficult.
If you're not familiar with Novell Netware, NT 3.0, or installing
drivers for 3.1, I recommend upgrading. Besides...Windows 3.1 is NOT
Y2K compliant (I don't thing DOS is either).
Dave
In article <[EMAIL PROTECTED]>,
Markus <[EMAIL PROTECTED]> wrote:
> I am wondering what is the easiest way to connect an old 486 (33MHz)
> running windows 3.1 to a pentium pc running Redhat 6.0 so that I will
be
> able to browse the internet on the 486 using the pentium as the
gateway
> to the net. I still want to be able to browse the internet with the
> pentium. My pentium has a catv connection to the net.
> Is there some cable that I can connect the 2 pcs by, and then some
free
> software?
>
> Cheers,
> Markus
>
--
Dave
Niagara Falls
[EMAIL PROTECTED]
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: lorax <[EMAIL PROTECTED]>
Subject: Re: Newbie needs help with DNS: Can't get host lookup
Date: 5 Aug 1999 02:42:45 GMT
[EMAIL PROTECTED] wrote:
: I can't get my DNS to resolve on the host file first. I have host.conf
: set with hosts first and files are first in nsswich.conf. Restarted the
: daemon and everything and it still won't do.
: Help!
: [EMAIL PROTECTED]
Make sure you are loading the zone in /etc/named.conf and /etc/named.boot
------------------------------
Date: 04 Aug 99 22:45:02 -0500
From: "Gene Heskett" <[EMAIL PROTECTED]>
Subject: Getting the penguinpowered script to run
Unrot13 this;
Reply to: <[EMAIL PROTECTED]>
Lo all;
www.penguinpowered.com has a neat little service, a remote DNS if you
will, where you run this script when you logon, and your machine can
then be found by entering an address of mymachine.penguinpowered.com
until it goes back offline.
Neat as can be, provided you can get it to run automaticly. I do have
it working for the reboot condition since the machine is online via a
dialup full time.
Unforch, if it gets knocked offline for some reason, and the PERSIST=1
is set in the config file so it redials, the address set in the PenPwrd
DNS is then bogus and needs to be updated.
Into what script do I put that one liner that runs the perl script that
does this, so that its done as soon as a valid IP-Remote has been
obtained on the (re)dial?
Cheers, Gene
--
Gene Heskett, CET, UHK |Amiga A2k Zeus040 50 megs fast/2 megs chip
Ch. Eng. @ WDTV-5 |A2091,GuruRom,1g Seagate,CDROM,Multiface III
|Buddha + 4 gig WDC drive, 525 meg tape
|Stylus Pro, EnPrint, Picasso-II, 17" vga
RC5-Moo! 690kkeys/sec isn't much, but it all helps
email gene underscore heskett at iolinc dot net
--
------------------------------
From: Daniel Forester <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup,comp.os.linux.misc
Subject: Re: users can't mount cdrom
Date: 5 Aug 1999 03:39:13 GMT
In comp.os.linux.misc Jeff Greer <[EMAIL PROTECTED]> wrote:
: I like Linux, but setting up some things is a complete waste of
: time. I am not interested in knowing how to make my cdrom
: mountable by guests. There should be a standard script to handle
: this.
: --
: Jeff Greer
: B.S. computer science, University of MO - Rolla
Yeah, so this isn't meant to be a flame or anything, but if you think
there "really should be a standard script... blah blah blah", and, "I'm
not interested in knowing how to make my CDROM mountable by guests...."
This strikes me as odd. ESPECIALLY as a CS major, shouldn't you care a
LITTLE about this stuff? Don't they teach the process, and not the final
answer?
--
Daniel E. Forester
Georgia Institute of Technology
http://www.prism.gatech.edu/~gte061f/
A waist is a terrible thing to mind.
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Qpopper authentication problems
Date: Thu, 05 Aug 1999 03:43:43 GMT
I'm having the same problem. I tried enabling SHADOW password support as
you suggested below, but still get the same "ERR: unable to open
/etc/pop.auth database file" message. Did it work for you?
In article <7mcu72$8ks$[EMAIL PROTECTED]>,
mike <[EMAIL PROTECTED]> wrote:
> I found on the newsgroups the command for enabling SHADOW password
support
> (that was the problem...):
>
> "./configure --enable-specialauth"
>
> mike
>
> dan wrote:
> >
> > I am setting up a Red Hat 6.0 to run as a POP3 mail server.
Everything
> > works fine, I can telnet into port 110 and I get the welcome
message.
> > But, when I try and connect from a netscape messenger on a Win98
> > workstation I get an error saying that the password is incorrect. I
> > know it isn't because I can log on directly to the box with the
login
> > and password. Any ideas would be appreciated.
> >
> > Daniel Good
> >
> >
> > If you love someone set them free, if they come back, set them on
fire!
> > ********************************************
http://www.datatel-mn.com/
> >
> >
> > Sent via Deja.com http://www.deja.com/
> > Share what you know. Learn what you don't.
>
> ------------------ Posted via SearchLinux ------------------
> http://www.searchlinux.com
>
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: [EMAIL PROTECTED] (Vidar Andresen)
Crossposted-To: comp.os.linux.hardware,comp.os.linux.misc
Subject: Re: Linux driver for 3Com NIC:
Date: Thu, 05 Aug 1999 04:00:35 GMT
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (Jonathan Wilson) wrote:
>Are you sure these files work... As I have installed both on my Linux
>server and it has problems. It keeps getting an IRQ 0 for my NIC???
Any help in http://cesdis.gsfc.nasa.gov/linux/misc/irq-conflict.html ?
[...]
The solutions
If the card is reporting IRQ0 or IRQ255, that indicates the
card has not been assigned an interrupt. There are two
likely fixes: either the BIOS does not have enough IRQ
lines available for PCI devices (e.g. all are assigned to
"legacy" ISA devices instead of "PnP"), or the BIOS has a
"PnP OS" setting that must be disabled. Yes, this is
confusing: the "PnP OS" setting is bad (it really means
"Windows OS?"), but the PnP IRQ assignment is good.
If the card is reporting a valid IRQ, but that IRQ is being
used by another device you have an interrupt conflict. The
easiest and generally best solution is to put the
conflicting device on another IRQ line. This can only be
done through the PCI BIOS setup. Unlike ISA cards, PCI
cards have no way of setting their own IRQ. That is done
at boot time by the PCI BIOS, and the BIOS reports its
selection
[...]
And moore. Look for machine/motherboard trouble there and in
http://cesdis.gsfc.nasa.gov/linux/drivers/vortex.html
>It's really odd. As I've compile the module a couple of times in
>different as described in the file and always get the same result.
>
>Has anyone got it to work... and if so could you please tell me how,
>or send me the *.c program you compiled.
Look into the 3c59x.c for your nic. Something like
static char *version =
"3c59x.c:v0.99Kb 5/7/99 Donald Becker
http://cesdis.gsfc.nasa.gov/linux/drivers/vortex.html\n";
and down ....
{"3cSOHO100-TX Hurricane", 0x10B7, 0x7646, 0xffff,
PCI_USES_IO|PCI_USES_MASTER, IS_CYCLONE, 128, vortex_probe1},
(It would not help you if i compiled the driver. I guess that part is
done well if no errormessage. So i guess it is a pci-irq-matter.)
>>>Does anyone know where I can get the Linux driver for 3Com's
>>>OfficeConnet 3CSOHO100-TX Fast Ethernet NIC ?
Do 3com have several '3CSOHO100-TX'?
Mvh Vidar Andresen
------------------------------
From: [EMAIL PROTECTED] (Richard Petty)
Subject: How to Configure for Realtime Streaming (RTSP)
Date: Thu, 05 Aug 1999 03:39:25 GMT
My Linux box is acting as a router for my home LAN.
I want to allow my clients workstations to have access to Apple QuickTime
4 streaming.
I got this info from Apple:
http://www.apple.com/quicktime/resources/qt4/us/proxy/proxy.html
Recommendations for Firewall Administrators
QuickTime follows the conventions of the RTP and RTSP
internet standards to stream media over the web. To
enable QuickTime 4 to work properly inside your firewall,
please follow the following IETF recommendation:
Open port 554 for RTSP/TCP data.
Open ports 6970 through 6999 (inclusive) for RTP/UDP data.
For more information on the RTP and RTSP internet standards:
RTSP: http://info.internet.isi.edu:80/in-notes/rfc/files/rfc2326.txt
RTP: http://info.internet.isi.edu:80/in-notes/rfc/files/rfc1889.txt
I'm having trouble translating this info into something useful.
I've examined the /etc/services file and studied it's form, but I'm still
don't think I can get it right. The HOWTO's haven't helped, either.
--Richard
--
Spam deterent: Remove the "bogus" part for a correct address.
------------------------------
From: [EMAIL PROTECTED] (Vidar Andresen)
Subject: Re: Rtl8139.c drive not compiling
Date: Thu, 05 Aug 1999 04:00:40 GMT
In article <7o2acf$4rk$[EMAIL PROTECTED]>,
"John N" <[EMAIL PROTECTED]> wrote:
>
>I have a SMC1211tx network interface card, in my experimental Linux machine
>(Running Manhattan, RH 5.2) no being recognized by RH.
>
>I have downloaded the latest version of the driver from Donald Becker's NASA
>site and copied as rtl8139.c to /lib/modules/preferred/net. When I tried to
>compile it with the directives in the final part of the final I get a ton of
>errors and it does not produce the required .o (module) file.
>
>Any ideas?
http://cesdis.gsfc.nasa.gov/linux/misc/modules.html ?
Mvh Vidar Andresen
------------------------------
From: "guru meditation" <[EMAIL PROTECTED]>
Subject: Re: 3c905b works at 100base?
Date: Wed, 4 Aug 1999 22:08:02 -0500
thats odd...I have a 3c Etherlink XL (3c905b) which uses the standard kernel
module..no patches. When I installed it, Linux regonized it as a vortex
card, and everything was fine. Infact it found both of them and set them
up. I use the back of the NIC to tell what speed I'm at (the 10/100
lights). All I can suggest is using kernel 2.2.x
------------------------------
Date: Wed, 4 Aug 1999 20:45:26 -0700
From: "Santiago Fisher" <[EMAIL PROTECTED]>
Subject: Re: MODEM DIAL-IN PROBLEM
I got uugetty to work fine with autoanswer (which is what you're trying to
do). (I got stuck later on, trying to use the ringback feature, with no luck
so far...)
The only different thing you're doing from what I did, and I believe it
could well be the cause for your modem not answering, is that you are
treating your modem as a dumb terminal *instead* of a modem. Try changing
your inittab line to:
S1:456:respawn:/sbin/uugetty -d /etc/conf.uugetty.ttyS1 ttyS1 F14400 vt100
I believe dumb terminals behave as DTEs, whereas modems are DCEs. You use a
straight serial cable to connect to an external modem, and a _null-modem_
cable to connect a terminal to a serial port.
Then you could try pushing serial speed up by using the Examples gettydefs
file included in the /usr/doc/... which supports higher speeds (115200
should work).
You can also get a wealth of information by turning on full debugging with
DEBUG=777
I got a 14.4k modem working with a F57600 entry. See my Aug.3 post titled
"uugetty - can't get RINGBACK to work" for the detail of my config files.
Hope this helps. Regards,
Santiago
Steve Nieman wrote in message <7o6s56$[EMAIL PROTECTED]>...
>Hi ...
>
>I have a Zoom Communications 56K External Modem.
>I have configured a dial-out PPP connection in
>Red Hat 6.0 using this modem without a problem.
>
>I now need to set up a dial in capability for
>my off-site colleagues and that has proven to
>be a real problem. I've tried a bunch of things
>but I cannot get the modem to pick up. It detects
>the ring but will not attempt to connect. All I
>get is a click and a pause and then ringing resumes.
>I've tried a lot of protocols and baud rates but
>nothing is working.
>
>I've copied the format of various relevant files.
>
>Help !!!!!
>
[...]
># inittab
>
># Run uugetty on serial port
>S1:456:respawn:/sbin/uugetty -d /etc/conf.uugetty.ttyS1 ttyS1 DT14400 vt100
------------------------------
From: Brian Rectanus <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.powerpc
Subject: Power Mac G3 dropping packets
Date: Wed, 04 Aug 1999 11:56:42 -0400
I have a Power Mac G3 desktop. I have linuxppc 1999 and debian
installed. Both systems drop packets like mad on eth0. It is not
noticable via telnet or ssh or even web surfing, but transferring large
files is a pain. When using FTP things just lock up until a timeout,
then all is fine. I thought it was my 10baseT line for a while, but I
switched lines off a machine I know to work well, but the same thing
happens. I have tried to upgrade kernels also. I have visited 2.2.7,
2.2.9 and now on 2.2.10.
I assume it is a configuration problem with my network card. Perhaps it
is running in duplex mode or 100baseT when it shouldn't? Has anyone
else noticed this sort of problem? Is there a way to configure the card
parameters? If so, what are the available parameters?
-Brian
--
Brian Rectanus
[EMAIL PROTECTED]
------------------------------
From: Chris Mahmood <[EMAIL PROTECTED]>
Subject: Re: POP3
Date: 04 Aug 1999 18:44:34 -0700
Anthony Valentine <[EMAIL PROTECTED]> writes:
> If you haven't already, uncomment the pop3 (and pop2 if you use pop2) line(s)
> in the /etc/netd.conf. Then refresh inetd with:
>
> killall -HUP inetd
>
> then try to telnet to port 110, and you should get something like:
> +OK POP3 mis3.sbs.com v7.59 server ready
you may also want to use /etc/hosts.allow and /etc/hosts.deny to
restrict who can use POP, etc. as it's notoriously insecure.
-ckm
------------------------------
From: [EMAIL PROTECTED] (Roger Plant)
Subject: Re: 486 to linux box?
Date: Thu, 05 Aug 1999 00:44:53 GMT
On Wed, 04 Aug 1999 14:28:45 +0200, Markus <[EMAIL PROTECTED]> wrote:
>Not sure if somebody will read this...but I read NET-3-HOWTO, however, I
>feel like I need a little more basic help...
>I have 1 3Com NIC in the 486 and 2 in my pentium. I also have the RJ45
>cable in between, so it seems like I have all hw I need to get the two
>computers to interwork. But now is where my bother begins. What are the
>apps that I need to allow the 486 (win 3.11) to browse the net using the
>pentium as the provider of the information from the net?
>
>Cheers,
>Markus
>
I have had some success with microsoft's 32 bit TCP/IP stack, for WFW
3.11. It is free from Microsoft's site.
It gives you network internet functionality (winsock), and also allows
you to connect to smb shares (Samba) on your Linux box.
Regards
Roger
===========================================================
Roger Plant :-) Email: [EMAIL PROTECTED]
===========================================================
------------------------------
From: Greg <[EMAIL PROTECTED]>
Subject: Re: IPChains Help Needed
Date: Wed, 04 Aug 1999 21:27:39 -0700
Monte Phillips wrote:
> If you are accessing the net via a modem you will need to point
> ipchains to it.
>
> ipchains -A forward -i ppp0 -j MASQ
>
> "Greg" <[EMAIL PROTECTED]> wrote:
> >I have RedHat 6.0 on which I have set up the ipforwarding / masqerading
> >using
> >ipchains. I have loaded the IRC module but I can not DCC anyone from my
> >windows 95 box which is behind the firewall. I have enabled:
> >
> >/sbin/ipchains -F forward
> >/sbin/ipchains -P forward DENY
> >/sbin/ipchains -A forward -i eth0 -j MASQ
> >echo 1 > /proc/sys/net/ipv4/ip_forward
> >
> >I have loaded the ip_masq_irc.o module.
I think the problem lays more in the fact
that DCC acknowlagement is directed on a
different port than the ones that are being used as
masquerading. This kind of connection is not really masq. It is a
request from the remote machine to connect to me.
It goes something like that:
I send a request saing: Connect to me on
123.123.123.123 ip and 2930 port
(could be a range of ports) then the other
machine is sending the actual
request to my ip (linux box). The question is
how do I get the request
transfered to the windows box and how do I
establish the connection.
thanks
Greg
--
Greg Furmanek
[EMAIL PROTECTED]
=============================================
Long Island '79 Charnobyl '86 Windows '98
------------------------------
Date: Wed, 4 Aug 1999 20:45:29 -0700
From: "Santiago Fisher" <[EMAIL PROTECTED]>
Subject: Re: Help with RH6 & Routing Problem
Seems that the NT box successfully "informed" the router about the existance
of the internal network (using whatever routing protocol you enabled), and
that for some reason you're not being able to do the same thing with routed
on your RH6 box.
I would guess that packets from the internal network *are* reaching the
router, but the router doesn't have a route for your internal network. You
could confirm this with a tcpdump on eth0.
Why don't you try adding a static route in your router, for network
203.38.98.192 netmask 255.255.255.192, with a next-hop of 203.38.93.182 ?
If that works, you can then go on and see why routed is not working as you
need (or maybe you can just leave things like that!).
Hope it helps.
Santiago
Shane Chrisp wrote in message ...
>I am trying to setup a firewall using RH6 and I am running into a problem
>where the IP Forwarding is getting to the external ethernet card but then
>not being forwarded by the card to the router. My network is as follows.
>
>Router 203.38.93.181 - 255.255.255.252
> |
> |
>Linux external eth0 203.38.93.182 - 255.255.255.252
> |
> |
>Linux Internal eth1 203.38.98.194 - 255.255.255.192
> |
>All other hosts on this same segment
>
>
>My problem is that when i set the routing up in RH6 i can ping from say
>203.38.98.195 to 203.38.93.182 but not to 203.38.93.181. The linux box does
>not appear to be forwarding the request to the router. The routing tables
>are setup correctly as i confirmed them by temporarily placing an NT box
>where the linux box is and configuring the routing and IP Forwarding. It
all
>workied well. But with the RH6 box with IP forwarding enabled and running
>routed i am unable to get it working.
>
>Does anyone have any ideas?
------------------------------
Date: Wed, 4 Aug 1999 20:45:18 -0700
From: "Santiago Fisher" <[EMAIL PROTECTED]>
Subject: Re: ARP replies w/ >1 interface on same network
Chuck,
I don't see why _all_ of the client-PC interfaces should respond to the
server's ARP-request, being that only one of them should have the IP address
for which a MAC address is being requested... (after all, you are using a
different IP address in each of the client PC's interfaces, aren't you ?).
I agree with you in that all 16 interfaces will _hear_ the ARP-request, but
only one of them should answer with its hw address.
Anyways, I think "hard-wiring" your ARP cache's entries on the server side
might help. Use the "arp -s IP_address MAC_address" if the server runs Linux
(I believe you'll have to add the lines to one of the rc scripts if you want
them to survive through reboots) and add all of the client's interfaces.
Whatever problem you were experiencing, I believe this should guarantee that
"return" traffic will be always picked up by the correct bridge.
Hope this helps.
Regards,
Santiago
Chuck Musser wrote in message <[EMAIL PROTECTED]>...
>Hi,
>
>I've having some trouble with using a Linux machine to test a bunch of
>Ethernet-to-ATM bridge devices.
>
>Here's my setup:
>
>1.) One client PC with Linux 2.2.10 and 4 Adaptec Quartet 4-port
>Ethernet boards, each port connected to one bridge.
>2.) The bridges all connect to a single class B network
>3.) A multihomed server with 16 addresses on the class B net.
>4.) Host routes on the client that route traffic bound for one of the
>server's IP address through a specific Ethernet interface. So an FTP
>session with 172.18.1.1 always goes through eth1, for example.
>
>Here's what happens:
>
>Traffic bound for a given IP address on the server always makes it
>through the correct interface. Return traffic, however, can come through
>__any__ of the interfaces. After a bit of snooping with tcpdump, I
>discovered that when the server sends out ARP requests, the request is
>"heard" by all the interfaces (because they are all in the same
>broadcast domain), and they all respond with their own MAC address.
>There is probably nothing wrong with this behavior in the normal case.
>However, my goal is to use the machine to simulate 16 individual
>computers. If the return traffic drifts between interfaces, I can't
>measure throughput and reliabilty with any accuracy.
>
>I'm looking at two ways around this: 1.) Some sort of tricky subnetting
>or 2.) modifying the kernel's ARP behavior to be more selective. Anyone
>have any ideas about which is the best way to proceed?
>
>Thanks,
>
>Chuck
------------------------------
** 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
******************************
