Linux-Networking Digest #86, Volume #11 Sat, 8 May 99 19:13:39 EDT
Contents:
Re: ipfwadm and RedHat 6.0 (Paul Nevin)
Re: ppp compression problem (Clifford Kite)
Re: Problem connnecting (another one... all the details) (Clifford Kite)
Re: Best Free X Windows Server for Win95/98 Box on Samba/Linux Network? ("Curt")
Re: Linux ppp server and win98 dialup client (James Carlson)
ipchain question. Need help. (PhilT)
Re: Connect Linux box with Win95 using TCP/IP (mist)
Re: isp's fair use policy ("Curt")
Re: Is it possible for aol to run under linux? ("David C. Brown")
Trouble with ethernet card (Seshu Parvataneni)
Linux as a Router? ("Kevin Adam")
Re: Olicom Token-Ring drivers for 2.2.6 kernel..Help please (Pascal Fleer)
Re: fetchmail works -- but so does sendmail (Paul Black)
Modem sharing with Samba?? (Arash)
Re: LinuxPing -> NTping DEC PCI tcp-ip problems (Matt)
Re: LinuxPing -> NTping DEC PCI tcp-ip problems (Matt)
This should not be so hard... (Ferdinand Tempel)
----------------------------------------------------------------------------
From: Paul Nevin <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Re: ipfwadm and RedHat 6.0
Date: Sun, 09 May 1999 04:25:44 +0800
>
> I just installed RedHat 6 which does not have ipfwadm anymore. Apparently
> there is ipfwadm-wrapper and ipchains modules to do the same thing. Has
> anyone managed to configure the new modules to do what many of us do with
> ipfwadm.
>
> Here is my internal ipfwadm configuration to share a single modem with my
> intranet..
>
> extract from my /etc/rc.d/rc.local
> ############################
> echo "1" > /proc/sys/net/ipv4/ip_forward
> /sbin/depmod -a
> /sbin/modprobe ip_masq_ftp.o
> /sbin/modprobe ip_masq_raudio.o
> # deny all ips except specified ones
> /sbin/ipfwadm -F -p deny
> # allow only these ips
> /sbin/ipfwadm -F -a m -S 192.168.10.2/24 -D 0.0.0.0/0 #bob
> /sbin/ipfwadm -F -a m -S 192.168.10.3/24 -D 0.0.0.0/0 #jane
> /sbin/ipfwadm -F -a m -S 192.168.10.4/24 -D 0.0.0.0/0 #mike
> /sbin/ipfwadm -F -a m -S 192.168.10.5/24 -D 0.0.0.0/0 #kym
> /sbin/ipfwadm -F -a m -S 192.168.10.6/24 -D 0.0.0.0/0 #paul
> # add the route
> /sbin/ifconfig eth0 192.168.10.1
> /sbin/route add -net 192.168.10.0
> echo "IP forwading completed"
> ###############################
>
> My internal network is running on 192.168.10.x with this linux box/modem on
> 192.168.10.1.
>
> Any help appreciated.
To all those that helped me .. I send my wholehearted appreciation. For those
looking for an answer to this problem I have included my corrected ipchains
configuration for my network. Refer to the above settings to see how it was set
up with ipfwadm.. not a lot changes!
###########IPCHAINS modification###############
echo "1" > /proc/sys/net/ipv4/ip_forward
/sbin/depmod -a
/sbin/modprobe ip_masq_ftp.o
/sbin/modprobe ip_masq_raudio.o
#Deny all by default
/sbin/ipchains -P forward DENY
#Allow only these IPs
/sbin/ipchains -A forward -s 192.168.10.2/24 -d 0.0.0.0/0 -j MASQ
/sbin/ipchains -A forward -s 192.168.10.3/24 -d 0.0.0.0/0 -j MASQ
/sbin/ipchains -A forward -s 192.168.10.4/24 -d 0.0.0.0/0 -j MASQ
/sbin/ipchains -A forward -s 192.168.10.5/24 -d 0.0.0.0/0 -j MASQ
/sbin/ipchains -A forward -s 192.168.10.6/24 -d 0.0.0.0/0 -j MASQ
#Add the route
/sbin/ifconfig eth0 192.168.10.1
/sbin/route add -net 192.168.10.0 eth0
###############END OF IPCHAINS##################
Tada..
Thanks again Linux community for your help.
------------------------------
From: kite@NoSpam.%inetport.com (Clifford Kite)
Crossposted-To: comp.protocols.ppp
Subject: Re: ppp compression problem
Date: 8 May 1999 09:43:30 -0500
Brian Modra ([EMAIL PROTECTED]) wrote:
: I'm having trouble negotiating ppp with an ISP. My Linux system is
: trying to be the client.
: Here's the relevant part of the debuging info:
: sent [CCP ConfReq id=0x1 <deflate 15> <deflate(old#) 15> <bsd v1 15>]
: rcvd [IPCP ConfReq id=0x1 <compress VJ 0f 01> <addr 195.197.0.195>]
: sent [IPCP ConfAck id=0x1 <compress VJ 0f 01> <addr 195.197.0.195>]
: .... so at this point the local IP address has been negotiated
: successfuly ...
: .... then the ISP sent a compression request which the local ppd rejected
: rcvd [CCP ConfReq id=0x1 < 11 06 00 01 01 03>]
: sent [CCP ConfRej id=0x1 < 11 06 00 01 01 03>]
: rcvd [IPCP ConfNak id=0x1 <addr 195.197.4.251>]
: .... it seems that the ISP then said "well in that case I don't like your
: address either ...
The ISP requested STAC compression, pppd can't do it because the algorithm
is proprietary. Normal occurrence. The IPCP ConfNak just says the IP
address pppd requested (0.0.0.0 ?) isn't acceptable and suggests that
pppd try 195.197.4.251 instead.
: .... and so the local ppd requested the address again
: sent [IPCP ConfReq id=0x2 <addr 195.197.4.251> <compress VJ 0f 01>]
: rcvd [CCP ConfRej id=0x1 <deflate 15> <deflate(old#) 15> <bsd v1 15>]
: sent [CCP ConfReq id=0x2]
: rcvd [IPCP ConfAck id=0x2 <addr 195.197.4.251> <compress VJ 0f 01>]
So pppd requests the suggested address and the ISP approves it. Pppd
sends an "empty" CCP request just before the IPCP ConfAck is received.
: .... but it seems that the ISP got really stubborn.
: sent [IPCP TermReq id=0x3 "Interface configuration failed"]
: ...............................
: The problem seems to be the
: rcvd [CCP ConfReq id=0x1 < 11 06 00 01 01 03>]
: .... which was rejected by my local pppd.
It shouldn't be. Since the ISP and pppd have no compression algorithm
that the other can use you can set the pppd option "noccp" and reject
CCP compression altogether. PPP implementations are often broken in one
way or another and this might work for an ISP that is confused by the
"empty" CCP request that pppd sent.
Are you sure you've authenticated properly. If you do login/password
then it may not be enough, many ISPs require PAP or CHAP authentication
and it's possible that a login/password would allow PPP negotiations to
get just so far and no further. If the ISP does, say, PAP, then try
dropping the chat login/password and see whether PAP will work.
: What should I do?
: I'm using ppp 2.3.3 on kernel 2.1.109 with kernel modules activated.
: The ppp modules and compression modules are loaded in the correct order.
Agghhhh. Well, the previous suggestions are reasonable even though
they are incorrect! It serves me right for not reading the entire
post more carefully.
You haven't read the linux/Documentation/Changes file that comes with the
kernel source. From kernel 2.1.100 up you need at least ppp-2.3.5 and
you should get ppp-2.3.7 to avoid certain nasty problems with ppp-2.3.5 .
It's at cs.anu.edu.au/pub/software/ppp/ .
--
Clifford Kite <kite@inet%port.com> Not a guru. (tm)
/* I gave up on politics when no matter who I voted for, I regretted it.
* -- Pepper...and Salt, WSJ */
------------------------------
From: kite@NoSpam.%inetport.com (Clifford Kite)
Subject: Re: Problem connnecting (another one... all the details)
Date: 8 May 1999 11:32:50 -0500
[EMAIL PROTECTED] wrote:
: I can't configure my internet connection under linux.
: Here's the error message from /var/log/messages
: localhost chat[310]: 64000^M May 8 12:13:29 localhost chat[310]: Wellcome
: to ITU TIES Internet Services^M May 8 12:13:29 localhost chat[310]: ^M May
: 8 12:14:13 localhost chat[310]: alarm May 8 12:14:13 localhost chat[310]:
: Failed May 8 12:14:13 localhost pppd[306]: Connect script failed May 8
Your chat script expects "ogin:" and doesn't get it. Maybe the ISP is
using a prompt like "Username:" and maybe it doesn't do login/password
but requires PAP or CHAP authentication.
--
Clifford Kite <kite@inet%port.com> Not a guru. (tm)
/* Better is the enemy of good enough. */
------------------------------
Reply-To: "Curt" <[EMAIL PROTECTED]>
From: "Curt" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.x,alt.os.linux,comp.os.linux.misc,comp.os.linux.setup
Subject: Re: Best Free X Windows Server for Win95/98 Box on Samba/Linux Network?
Date: Sat, 8 May 1999 07:54:59 -0500
Use the display number you got/gave when you started vncserver. Usually
host:1
An Bui <[EMAIL PROTECTED]> wrote in message
news:7h18fg$df6$[EMAIL PROTECTED]...
> hi all
> I just download the VNC. i try the Viewer and they ask for host:display
> for host i type in my host like my IP but display what number should i put
> down. can someplease give me a brief explain for this. I am a very new
> with linux.
>
> olivier
>
------------------------------
From: James Carlson <[EMAIL PROTECTED]>
Crossposted-To: comp.protocols.ppp
Subject: Re: Linux ppp server and win98 dialup client
Date: 04 May 1999 07:15:10 -0400
Ronald Hovens <[EMAIL PROTECTED]> writes:
> What doesn't work: when I use win98 dialup everything SEEMS to work: the
> connection is established, ppp is succesfully started and stays up, but
> I cannot ping nor telnet nor browse my linux box over the phone line. I
> use no firewall, and all these actions work fine on my Lan.
Did you enable IP forwarding? Try:
echo 1 > /proc/sys/net/ipv4/ip_forward
> What is wrong? Many thanks in advance!
> Info: (all according to the ppp-howto)
> I use a class C network (192.168.0.x)
> I use fixed ip numbers (192.168.0.1 for the linux ppp server,
> 192.168.0.10 for the laptop)
> I use an alias for ppp: alias ppp="exec /usr/sbin/pppd -detach"
>
> My /etc/ppp/options/file looks like this:
> proxyarp
Unless that 192.168.0.10 address is in the same subnet as your
Ethernet interface for your local LAN, you do NOT want to have this
option.
--
James Carlson, Software Architect <[EMAIL PROTECTED]>
IronBridge Networks / 55 Hayden Avenue 71.246W Vox: +1 781 372 8132
Lexington MA 02421-7996 / USA 42.423N Fax: +1 781 372 8090
"PPP Design and Debugging" --- http://people.ne.mediaone.net/carlson/ppp
------------------------------
From: [EMAIL PROTECTED] (PhilT)
Subject: ipchain question. Need help.
Date: Tue, 04 May 1999 11:41:02 GMT
Reply-To: [EMAIL PROTECTED]
I'm trying to restrict people from doing a telnet into my internal
network (192.168.1.0/24).
I have a Linux box doing Masquerading using ipchains (2.2.5 kernel)
and auto dialling using 'diald'. My ISP gives out only dynamic IP to
my 'ppp0'. How am I going to get the IP Address to setup my ipchain
rules?
I know it has to do with TCP packets with the SYN flag set for port
23. How do I set this up? I kinda read the IPChains-HOWTO, but being
the bright fellow that I am, I could not get all that it is trying to
explain.
My requirement was to stop people from doing a telnet from 'ppp0' into
any of the machines in the internal network.
I have tried the following, but it does not work
ipchains -A input -p tcp -y -s 0.0.0.0/0 23 -j DENY
Would any kind soul, please enlighten me?
Thanks in advance.
=====
PhilT
[EMAIL PROTECTED]
------------------------------
From: mist <[EMAIL PROTECTED]>
Subject: Re: Connect Linux box with Win95 using TCP/IP
Date: Sat, 8 May 1999 20:39:02 +0100
Reply-To: mist <new$[EMAIL PROTECTED]>
[EMAIL PROTECTED] scribed to us that -
>Hi Mist and Curt. Thanks for your advice so far but I still havent had much
>luck. Here are the exact details of what I tried:
>
>Red Hat Linux release 5.2 (Apollo)
>Kernel 2.0.36 on an i586
>
>localhost login: root Password: Last login: Sat May 8 14:15:10 on tty2
>localhost.localdomain:/root# broadcast 192.168.0.255 up < (ifconfig eth0
>192.168.0.1 netmask 255.255.255.0 broadcast 192.168.0.255 up)
??
>localhost.localdomain:/root# route add -net 192.168.0.0
>localhost.localdomain:/root# netstat -nr Kernel IP routing table
<snip - difficult to read in that formatting but I think it looked ok>
Something I found in another post -
Try using running "tcpdump -i eth0 -n ip proto 1" while you are pinging
the other machine to see if icmp echo request packets are even making it
out of the local machine's interface.
If you see echo requests, then the local machine is ok but the remote
one is broken.
If you don't see echo requests, then it's the local machine that's
broken.
The actual solution to that problem, was, I believe that the wrong IRQ
was being used for the NIC. This was fixed by specifying IRQ 10 as a
legacy ISA card in the bios.
I take it you know that both NICs are ok?
>My linux box isnt currently connected to the net, so I dont need to worry
>about IP Masquerading. The only thing I want to do is get both machines to
>talk.
>
<snip>
Yup, the reason that that was mentioned was because the windows gateway
was set wrongly.
--
Mist.
------------------------------
Reply-To: "Curt" <[EMAIL PROTECTED]>
From: "Curt" <[EMAIL PROTECTED]>
Subject: Re: isp's fair use policy
Date: Sat, 8 May 1999 08:06:09 -0500
I'm not sure what you're trying ot accomplish here, but:
I guess you could setup your linux box to handle dial in. Then, you coulds
run a script that starts your connection to your ISP after a short time.
#!/bin/bash
sleep 60
ifup ppp0
Is it really that surprising that your ISP didn't plan on setting up a modem
and phone line just for you, for around $20/month??? I imagine they'd be
glad to sell you a dedicated ISDN connection for around $150/month. At
least those are the going rates in this area.
ben <[EMAIL PROTECTED]> wrote in message
news:7h0b4t$[EMAIL PROTECTED]...
> Hi,
>
> I would like to be able to access my linux box at home over the
> internet. The problem is that my ISP has a "fair use" policy, so they
> don't like it when I leave my computer dialed in all day. Has anyone
> setup or know how to setup my linux box so I can call the line that the
> modem is connected to via a voice line, have it ring but not answer,
> then wait until the line is free and dial my ISP? Or can anyone think
> of a better solution?
>
> Any help would be greatly appreciated. I don't think my ISP likes me
> being logged in for over 20 hours a day.
>
> Thanks,
> Ben
>
------------------------------
From: "David C. Brown" <[EMAIL PROTECTED]>
Subject: Re: Is it possible for aol to run under linux?
Date: Sat, 08 May 1999 20:51:30 GMT
heh I never heard Linux and AOL in the same sentence heh
SMinor wrote:
>
> nt
------------------------------
From: Seshu Parvataneni <[EMAIL PROTECTED]>
Subject: Trouble with ethernet card
Date: Sat, 08 May 1999 14:45:14 -0700
Hi,
We've recently installed Linux Version 5.2 and everything seems to
be just fine. Thanks to the developer's of Linux. I am having problems
while detecting the ethernet card's hardware address. At some point
linux detected it but then afterwards it did not. Since I am new to any
of the installlation process I do not know what's happening. The output
during the two (successful and unsuccessful) detections were as follows
:-
/**Successful attempt **/
May 6 14:54:41 kernel: eth0: 3Com 3c905B Cyclone 100baseTx at 0x6900,
00:10:4b:68:fd:3c, IRQ 9
May 6 14:54:41 kernel: 8K byte-wide RAM 5:3 Rx:Tx split,
autoselect/NWay Autonegotiation interface.
May 6 14:54:41 kernel: Enabling bus-master transmits and whole-frame
receives.
/**Unsuccessful attempt **/
May 8 14:06:04 kernel: eth0: 3Com 3c905B Cyclone 100baseTx at 0x6900,
ff:ff:ff:ff:ff:ff, IRQ 9
May 8 14:06:04 kernel: 1024K word-wide RAM 3:5 Rx:Tx split,
autoselect/NWay Autonegotiation interface.
May 8 14:06:04 kernel: Enabling bus-master transmits and early
receives.
These messages were copied from the /var/log/messages file.
Thanks in advance
Seshu Parvataneni
------------------------------
From: "Kevin Adam" <[EMAIL PROTECTED]>
Subject: Linux as a Router?
Date: Sat, 8 May 1999 16:21:34 +0100
I am a systems administrator who has been trying to get Linux into our
organisation for some time. Until now I have not had a good enough "excuse"
to do this. However, I now have a cunning plan! Here is the problem:
We have 2 offices connected by a British Telecom 64 Kb kilostream leased
line. The far end office uses token ring and we use ethernet. At the moment
these are connected by 2 IBM PS2 machines running the IBM token ring bridge
program. I don't have the budget to replace the PS2's (non Y2K) with Cisco's
etc. I know I can use 2 windows NT machines to perform the routing, but they
will need rebooting every 10 minutes and they cost nearly as much as the
Cisco's anyway.
So:
1. Can I use Linux for this task?
2. What software & hw will I need?
3. Will I be able to install it? (no real Linux experience so far)
Prize for the best response is a smug smile and the warm feeling in the
knowledge that you helped stop 2 more NT boxes entering the world ;-)
Thanks,
Kevin Adam.
------------------------------
From: Pascal Fleer <[EMAIL PROTECTED]>
Subject: Re: Olicom Token-Ring drivers for 2.2.6 kernel..Help please
Date: Sat, 08 May 1999 18:08:45 +0200
Reply-To: [EMAIL PROTECTED]
Hi Daniel,
Olicom has a driver for kernel 2.1.129 ! And you can use it for 2.2.x Kernels.
Ther compile and run without problems.
Be careful when you use special block devices (like the compaq smart/2 raid
controllers). We got it running but it was pain.
The reward is that runs fast.
Pascal
Daniel Ulfe Paredes wrote:
> Hello. I Have a Olicom Token-Ring PCI card working fine in kerner 2.0.35.
> Olicom Token-Ring support in this kernel was provided by download in Olicom
> website making a
> Patch in kernel source and "make" related utilities. I want to use this card
> whit
> 2.2.6 kernell but in
> make config i still haven't an option to compile Olicom TokenRing Adapter.
> There is posible to compile this drivers like a module in 2.0.35 version and
> after use this
> in 2.2.6 kernel?. If this is not possible.... How can I patch source kernel
> in order to incorporate
> Olicom support???
>
> BTW: I'm not have much experience using linux kernel modules option
>
> Regards in advance...
>
> Daniel Ulfe from Spain
------------------------------
From: Paul Black <[EMAIL PROTECTED]>
Crossposted-To: uk.comp.os.linux,demon.ip.support.unix
Subject: Re: fetchmail works -- but so does sendmail
Date: Sat, 08 May 1999 16:58:58 +0100
[EMAIL PROTECTED] (Simon Child) wrote:
>
> On Sat, 08 May 99 10:32:28 GMT, Phil Hunt <[EMAIL PROTECTED]> wrote:
>
> >I've tried not running sendmail, but this is no good because it causes
> >fetchmail to stop working too (fetchmail tries to deliver its mail to
> >sendmail). There's nothing in the Demon section of the fetchmail FAQ,
> >either.
>
> Instead of starting sendmail from /etc/rc3.d/init.d or wherever Suse
> does it, start it in /etc/inetd.conf with a -bs switch and it will
> receive mail from fetchmail but not from Demon.
That is not sufficient to prevent Demon connecting to your SMTP port.
Some form of access control is needed (TCP wrappers or firewall).
Paul
------------------------------
From: [EMAIL PROTECTED] (Arash)
Crossposted-To: comp.protocols.smb,linux.samba
Subject: Modem sharing with Samba??
Date: Sat, 08 May 1999 22:10:31 GMT
Hi
can i connect my modem to my linux box and share it with my
MS-Windows PCs so that they can use the modem to connect to eg. an
ISP? Can i use Samba for that?
Do i need special software on my windows pcs? Can i use call back?
Thanx,
Arash :-)
------------------------------
Date: Sat, 08 May 1999 23:07:49 +0100
From: Matt <[EMAIL PROTECTED]>
Subject: Re: LinuxPing -> NTping DEC PCI tcp-ip problems
mist, cheers I will try the..
> route add -net 192.168.10.0
would this mean that the linux box would then have the IP address
of 192.168.10.0 I thought xxx.xxx.xxx.0 IP configs should not be
advised or does'nt it matter ?
You mentioned the following..
> from the linux box, perhaps specifying the eth0 interface as well.
Being quite new to the Ethernet network and setups how would one
specify the eth0 interface is there a command line I should use ?
Many thanks
Matt
mist wrote:
> Matt <[EMAIL PROTECTED]> scribed to us that -
> >Hi, I am having great problems with the configuration of Linux and NT
> >I have the network stats below.
> >
> >I am using two DEC PCI cards one has a tcp-ip address of 192.168.10.1
> >(NT)
> >the other 192.168.10.2 (Linux).
> >
>
> <snip>
>
> >
> >route -n
> >
> >
> >Kernel IP routing table
> >Destination Gateway Genmask Flags Metric Ref Use Iface
> >192.168.10.2 0.0.0.0 255.255.255.255 UH 0 0 0 dummy0
>
> That is surely wrong? The dummy route takes traffic for the supposed IP
> address of the eth0 NIC? (In preference to the NIC itself.)
>
> >212.211.0.254 0.0.0.0 255.255.255.255 UH 0 0 0 ppp0
> >192.168.10.0 0.0.0.0 255.255.255.0 U 0 0 1 eth0
> >127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 2 lo
> >0.0.0.0 212.211.0.254 0.0.0.0 UG 0 0 4 ppp0
> >
> >ifconfig
> >
> >lo Link encap:Local Loopback
> > inet addr:127.0.0.1 Bcast:127.255.255.255 Mask:255.0.0.0
> > UP BROADCAST LOOPBACK RUNNING MTU:3584 Metric:1
> > RX packets:38 errors:0 dropped:0 overruns:0
> > TX packets:38 errors:0 dropped:0 overruns:0
> >
> >dummy0 Link encap:10Mbps Ethernet HWaddr 00:00:00:00:00:00
> > inet addr:192.168.10.2 Bcast:192.168.10.255 Mask:255.255.255.0
>
> dummy0 has the same address as your NIC. What happens without this
> interface?
>
> > UP BROADCAST RUNNING NOARP MULTICAST MTU:1500 Metric:1
> > RX packets:0 errors:0 dropped:0 overruns:0
> > TX packets:0 errors:0 dropped:0 overruns:0
> >
> >eth0 Link encap:10Mbps Ethernet HWaddr 00:40:C7:99:0D:F4
> > inet addr:192.168.10.2 Bcast:192.168.10.255 Mask:255.255.255.0
> > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> > RX packets:0 errors:6 dropped:0 overruns:0
> > TX packets:0 errors:5 dropped:0 overruns:0
> > Interrupt:11 Base address:0xe800
> >
>
> <Snip>
>
> >
> >Ping my own box (itself)
> >
> >PING 192.168.10.2 (192.168.10.2): 56 data bytes
> >64 bytes from 192.168.10.2: icmp_seq=0 ttl=64 time=0.1 ms
> >64 bytes from 192.168.10.2: icmp_seq=1 ttl=64 time=0.1 ms
> >64 bytes from 192.168.10.2: icmp_seq=2 ttl=64 time=0.1 ms
> >
>
> Well, yes, it will work because it's going straight to the dummy
> interface and back.
>
> >Ping the other box (winNT)
> >
> >--- 192.168.10.2 ping statistics ---
> >3 packets transmitted, 3 packets received, 0% packet loss
> >round-trip min/avg/max = 0.1/0.1/0.1 ms
>
> Have you done
>
> route add -net 192.168.10.0
>
> from the linux box, perhaps specifying the eth0 interface as well.
>
> <snip>
>
> [NGs trimmed]
>
> --
> Mist.
------------------------------
Date: Sat, 08 May 1999 23:09:38 +0100
From: Matt <[EMAIL PROTECTED]>
Subject: Re: LinuxPing -> NTping DEC PCI tcp-ip problems
mist, cheers I will try the..
> route add -net 192.168.10.0
would this mean that the linux box would then have the IP address
of 192.168.10.0 I thought xxx.xxx.xxx.0 IP configs should not be
advised or does'nt it matter ?
You mentioned the following..
> from the linux box, perhaps specifying the eth0 interface as well.
Being quite new to the Ethernet network and setups how would one
specify the eth0 interface is there a command line I should use ?
Many thanks
Matt
mist wrote:
> Matt <[EMAIL PROTECTED]> scribed to us that -
> >Hi, I am having great problems with the configuration of Linux and NT
> >I have the network stats below.
> >
> >I am using two DEC PCI cards one has a tcp-ip address of 192.168.10.1
> >(NT)
> >the other 192.168.10.2 (Linux).
> >
>
> <snip>
>
> >
> >route -n
> >
> >
> >Kernel IP routing table
> >Destination Gateway Genmask Flags Metric Ref Use Iface
> >192.168.10.2 0.0.0.0 255.255.255.255 UH 0 0 0 dummy0
>
> That is surely wrong? The dummy route takes traffic for the supposed IP
> address of the eth0 NIC? (In preference to the NIC itself.)
>
> >212.211.0.254 0.0.0.0 255.255.255.255 UH 0 0 0 ppp0
> >192.168.10.0 0.0.0.0 255.255.255.0 U 0 0 1 eth0
> >127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 2 lo
> >0.0.0.0 212.211.0.254 0.0.0.0 UG 0 0 4 ppp0
> >
> >ifconfig
> >
> >lo Link encap:Local Loopback
> > inet addr:127.0.0.1 Bcast:127.255.255.255 Mask:255.0.0.0
> > UP BROADCAST LOOPBACK RUNNING MTU:3584 Metric:1
> > RX packets:38 errors:0 dropped:0 overruns:0
> > TX packets:38 errors:0 dropped:0 overruns:0
> >
> >dummy0 Link encap:10Mbps Ethernet HWaddr 00:00:00:00:00:00
> > inet addr:192.168.10.2 Bcast:192.168.10.255 Mask:255.255.255.0
>
> dummy0 has the same address as your NIC. What happens without this
> interface?
>
> > UP BROADCAST RUNNING NOARP MULTICAST MTU:1500 Metric:1
> > RX packets:0 errors:0 dropped:0 overruns:0
> > TX packets:0 errors:0 dropped:0 overruns:0
> >
> >eth0 Link encap:10Mbps Ethernet HWaddr 00:40:C7:99:0D:F4
> > inet addr:192.168.10.2 Bcast:192.168.10.255 Mask:255.255.255.0
> > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> > RX packets:0 errors:6 dropped:0 overruns:0
> > TX packets:0 errors:5 dropped:0 overruns:0
> > Interrupt:11 Base address:0xe800
> >
>
> <Snip>
>
> >
> >Ping my own box (itself)
> >
> >PING 192.168.10.2 (192.168.10.2): 56 data bytes
> >64 bytes from 192.168.10.2: icmp_seq=0 ttl=64 time=0.1 ms
> >64 bytes from 192.168.10.2: icmp_seq=1 ttl=64 time=0.1 ms
> >64 bytes from 192.168.10.2: icmp_seq=2 ttl=64 time=0.1 ms
> >
>
> Well, yes, it will work because it's going straight to the dummy
> interface and back.
>
> >Ping the other box (winNT)
> >
> >--- 192.168.10.2 ping statistics ---
> >3 packets transmitted, 3 packets received, 0% packet loss
> >round-trip min/avg/max = 0.1/0.1/0.1 ms
>
> Have you done
>
> route add -net 192.168.10.0
>
> from the linux box, perhaps specifying the eth0 interface as well.
>
> <snip>
>
> [NGs trimmed]
>
> --
> Mist.
------------------------------
From: Ferdinand Tempel <[EMAIL PROTECTED]>
Subject: This should not be so hard...
Date: Sat, 08 May 1999 22:11:55 +0000
Hi,
I think I'm gonna need some expert help:
I'm trying to set up a 486DX100 machine as a firewall/webserver.
For this I have three NIC's, one Realtec ne2000 clone with integrated
hub (a real nice card, for the internal network) and two regular ne2000
clones (Realtec and Winbond), from which one is going to be tied to the
outside connection. The last one is unused, hence not build in, but just
available.
On the mainboard there are three free PCI slots.
Easy enough I just put the two realtecs into a slot and tried to get the
thing going.
Linux (redhat 5.1, freshly installed) recognises the two cards as the
right ones and assignes IRQ 0 and 15 to them (bios on autodetect).
Not possible (0 doesn't seem to exsist, 15 is reserved for an ide
controller), so I set the IRQ's manually to a free one (10 and 11 for
starters) in the bios.
Reboot, and again the NIC's are identified OK, and even with the right
IRQ.
Now here comes the great thing:
At least from one of them I get the error:
"Tx timed out, interrupt lost?..."
This can be from either card, depending on which one is found first.
I've tried switching slots, switching cards, and all the free IRQ's in
all combinations.
The result is that I think only IRQ 9 seems to work on one card, but all
the other free IRQ's (5,10,11,14) fail on the other one (either Winbond
or Realtec) with the error above but even the "working" IRQ (9) still
doesn't make the machine pingable on the network (the hub is assigned to
IRQ 9).
I've tried this before on a cyrixP166+, and it worked: NIC's: Realtec
with hub, and Winbond, IRQ's: 10 and 11.
I hope this has been a clear description of the problem.
Now for the question:
- Am I doing something wrong? (probably)
- Are my PCI slots damaged? (could be)
- Is there something wrong with the NIC's? (not likely)
Please enlighten me...
Relevant output of dmesg, /proc/ioports, /proc/interrupt follows:
/proc/interrupt
0: 19672 timer
1: 224 keyboard
2: 0 cascade
8: 1 + rtc
9: 0 eth0
10: 0 eth1
13: 1 math error
14: 21514 + ideo
/proc/ioports
<snip>
d000-d01f : eth1
d100-d11f : eth0
dmesg
<snip>
ne2k-pci.c:v0.99L 2/7/98 D. Becker/P. Gortmaker
http://cesdis.gsfc.nasa.gov/linux/drivers/ne2k-pci.html
ne2k-pci.c: PCI NE2000 clone 'RealTek RTL-8029' at I/O 0xd100, IRQ 9.
eth0: PCI NE2000 found at 0xd100, IRQ 9, 00:20:18:57:D2:CF.
ne2k-pci.c: PCI NE2000 clone 'RealTek RTL-8029' at I/O 0xd000, IRQ 10.
eth1: PCI NE2000 found at 0xd000, IRQ 10, 00:20:18:38:47:05.
eth1: Tx timed out, lost interrupt? TSR=0x3, ISR=0x97, t=1000.
eth1: Tx timed out, lost interrupt? TSR=0x2, ISR=0x3, t=329.
<snip, this goes on for a few times>
--
Regards,
FoT
------------------------------
** 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
******************************
