Linux-Networking Digest #538, Volume #10 Thu, 18 Mar 99 10:13:43 EST
Contents:
Re: Redhat 5.1 - Can't ping outside box ("Paco")
Re: Setting up 2 nics under Linux RH 5.1 (MadAdmin)
Re: Help whit RealTek8019 ("Tom Holmes")
Re: IP-Masq: Can't Email Attachment from MAC OS Client (Samuel SOUK ALOUN)
Re: autofs +? smbmount (Harald Fuchs)
Help! Cant route WIN98 through Linux box! ("Michael Arnold")
Keypad protocol support after 2.0.35 ("Steven Kroesbergen")
Re: Help - "private" ip address conflict using ip masq. ("Leopold Toetsch")
Problem mailing with direct connect computers (mike)
Encrypted Passwords on a Linux Samba box?! ("Martin Cole")
Re: Help: Network card problems ("Chris Cantwell")
Re: Ethernet Card on 2.2.3 ("Chris Cantwell")
Re: Ethernet card IP setup in RedHat 5.2 ("Chris Cantwell")
Re: Help! Cant route WIN98 through Linux box! ("Sam Grainger")
Re: Setting up 2 nics under Linux RH 5.1 ([EMAIL PROTECTED])
Re: a free software dynamic Firewall gateway for MICO & alike ? (Owen Rees)
----------------------------------------------------------------------------
From: "Paco" <[EMAIL PROTECTED]>
Subject: Re: Redhat 5.1 - Can't ping outside box
Date: Thu, 18 Mar 1999 14:14:50 +0200
Check out your routing tables.
Type route print. You should see all the routes you have.
Make sure that you are on the same subnet as the other networked machines ie
if your linux box is 196.1.1.1 , you should have an IP address between
196.1.1.2 - 196.1.1.254
Also check to see through the ifconfig command if you have a dummy0 entry.
Remove it. It is fine if you have the eth0 and the LO entries.
Hope this helps
StMonBoy <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
>Okay, I recently set up a Redhat 5.1 box to play with. I have an NDC
>(Sohoware) network card in it. It uses a Matronix chip, so when I
installed
>Redhat I configured it to use the tulip drivers.
>
>All seems well but I can't ping to my other Win98 boxes on the network. I
can
>ping to my Redhat box itself.
>
>I went to Redhat's site and searched the Faq-o-matic thing and pulled up
the
>"Can't ping outside" FAQ.
>
>Running ifconfig -a I see that my network card is indeed configured.
>
>Then I ran netstat -r. The FAQ said that if the "default route doesn't
show up
>or has the wrong IP address, you may need to make changes to your setup."
>
>Okay, the default route doesn't show up for me. So what "changes" to I
make to
>my setup?
>
>Thanks!
>
>Steven Mon
------------------------------
From: [EMAIL PROTECTED] (MadAdmin)
Crossposted-To: comp.os.linux.questions,comp.os.linux.setup,alt.2600
Subject: Re: Setting up 2 nics under Linux RH 5.1
Date: Thu, 18 Mar 1999 13:15:29 GMT
Reply-To: LART
On Wed, 17 Mar 1999 23:23:52 GMT, the new guy <[EMAIL PROTECTED]>
wrote:
>I'm trying to install t nics under Red Hat linux 5.1
>One is a 3c509 (pci) and the second is a 3c503 (isa0
>
>Now, the 3c506 was detected and installed fine when I installed RH5.1
>and is ETH0.
>
>I tried to set up the 3c503 as ETH1 by adding the following entries
>in my conf.moduels file
>
>alias 3c509
>alias 3c503
>* snd card stuff*
>options 3c503 io=0x350 irq=9
>
>However nothing happened.
>
>These are the instructions I followed out of the ethernet-howto file.
>
>Could someone clue me in as to what I've done wrong.
>
>Thanx
>
>BTW If possible, please send response to [EMAIL PROTECTED]
>The New Guy
I've never had good luck with a NIC running on IRQ9. I generally use
3coms on IRQ10 or 5 if possible. I wouldn't put ANYTHING on IRQ9
except PCI cards that don't mind sharing interrrupts since that IRQ is
unique. IRQ9 cascades off IRQ2 unless they've finally gotten around to
doing interrupt controllers that don't do that.
~~~~~~~~~~~~~~~
The sig.....
"Bother," said Pooh.
"Eeyore, ready two photon torpedoes
and lock phasers on the Heffalump.
Piglet, meet me in Transporter Room Three."
------------------------------
From: "Tom Holmes" <[EMAIL PROTECTED]>
Subject: Re: Help whit RealTek8019
Date: Wed, 17 Mar 1999 23:21:22 -0500
I am a newbie Linux user myself, and I went through this same problem, so I
hope I can help.
1) you can get the new 'C' source code for the realtek pci network card at:
http://cesdis.gsfc.nasa.gov/linux/drivers/rtl8139.html
You can also go to the realtek web site and they will have the uncompiled
source code also.
2) You need to some get the source code onto your linux box. I put the file
on a MS-DOS formatted floppy, mounted the floppy, and then copied the file
to my directory.
3) Once I got the file to the right working directory;
/usr/src/linux/drivers you should see all the other source code drivers for
all the other network cards. Once here you can re-compile the kernel (hence
doing a 'make modules' command) and that will re-compile all the drivers
here.
Or you can re-compile the one file 'rtl8139.c' When the file is compiled
you will have an rtl8139.o file which belongs in the
/lib/modules/2.0.36-0.7/net directory. You will see all the other object
modules here.
4) use LinuxConf and setup the networking. You can set the eth0 interface
to use the rtl8139 module. You'll also be able to setup the ip address,
netmask, and gateway, etc.
5) Save the configuration, and if there are no errors when this is saved,
reboot the system.
These are the steps I took, and it took me a week of research to figure all
this. I got help from some very nice people, and I pass this information
along to you.
I realize these are not easy steps and this is more advanced work for a
Linux admin. Some people said 'just buy a nic that is supported.' But,
then I wouldn't have learned as much as I did.
Hope this helps.
Tom
Daniel Cederhammar wrote in message <[EMAIL PROTECTED]>...
>I have Debian Linux v2.0 on my computer but can�t get my network card to
>work. It�s a RealTek8019 or NE2000 compatible. I have tested to use NE2000
>compatible in Linux but it don�t work. HELP!!!!!!!!!!
>
>
------------------------------
From: Samuel SOUK ALOUN <[EMAIL PROTECTED]>
Subject: Re: IP-Masq: Can't Email Attachment from MAC OS Client
Date: Thu, 18 Mar 1999 12:51:20 +0100
Stephen Lubbs wrote:
> Hi folks,
> I have a small net with a Linux box providing IP-masquerading services
> for my internal net to a PPP link to the outside world. PPP is managed
> by diald. I have a very simple set of rules for ipfwadm that accepts all
> traffic (once I get everything working then I'll add security). I an
> NT4.0, a Win95, and a MAC OS 8.1 (Open Transport) client. Everything
> works quite well except for the MAC client which is unable to send
> e-mail with any kind of attachment. When I try to send an attachment the
> send hangs. I have tried this with both Netscape 4.5 and Eudora Pro 4.x.
>
> Could this be an MRU/MTU problem?? Any ideas??
>
I had the same problem with on a Mac-6100. It was due to a misfunction on
the motherboard. Apple had made an announcement about it, and the only
solution was to replace it !!!!
Search the Apple site and check if your model is ok.
Sam.
------------------------------
From: Harald Fuchs <[EMAIL PROTECTED]>
Crossposted-To: linux.samba
Subject: Re: autofs +? smbmount
Date: 18 Mar 1999 14:53:48 +0100
In article <[EMAIL PROTECTED]>,
Brian Mila <[EMAIL PROTECTED]> writes:
>> i like the autofs stuff because it can mount my dos partitions,cdrom
>> and nfs server.
>>
>> In my netwotk are some win9x clients.
>>
>> I can mount their disk manuallyby commands like
>> smbmount '//win9xpx/C' /mountdir/win9xc -P secret
>>
>> now i wanted to know how can i use the autofs to do these for me?
> Yes you can. Look in the examples/autofs directory of the samba
> source. If you don't have the source code, go get it from
> http://samba.anu.edu.au/samba/. There is a script called mount.smb
> which you need to install (which is not hard, look at the doc in the
> same directory). This script intercepts mount calls and passes them
> with the correct parms to smbmount. Your example below would now be:
> 'mount -t smb /win9xpx/C /mountdir/win9xc'.
It works for me without that. Just put
SHARE -fstype=smbfs,login=USER,passwd=PASS ://WINHOST/SHARE
into your /etc/auto.whatever file.
------------------------------
From: "Michael Arnold" <[EMAIL PROTECTED]>
Subject: Help! Cant route WIN98 through Linux box!
Date: Thu, 18 Mar 1999 07:56:21 -0600
Can someone out there help me? I cant seem to get my Windows 98 box to route
through my Linux box via pppd. I have been to every howto page with
absolutely no luck! My ppp is working fine! I can surf, send and receive
mail etc.. I just can't get the Windows box to go outside the Linux box! I
have my 98 box authenticating and logging into the Linux.That is all working
fine an dandy. I am stuck! Any help would be great!
Thanx................Mike
------------------------------
From: "Steven Kroesbergen" <[EMAIL PROTECTED]>
Crossposted-To: de.alt.comm.isdn4linux,nl.comp.isdn
Subject: Keypad protocol support after 2.0.35
Date: Thu, 18 Mar 1999 15:01:26 +0100
Hi all,
I've got a patch that supports the Dutch keypad protocol, however this patch
does not work after 2.0.35
Does anyone know how I can use the keypad protocol after 2.0.35 (I'm running
2.0.36 and 2.2.3)
Thanks,
Steven
steven at knd dot nl
------------------------------
From: "Leopold Toetsch" <[EMAIL PROTECTED]>
Subject: Re: Help - "private" ip address conflict using ip masq.
Date: Thu, 18 Mar 1999 10:26:13 +0100
Hi,
David M. Reed wrote in message <7cp7so$[EMAIL PROTECTED]>...
>I have ip masq. set up and working on Red Hat 5.2, with a TCI @Home cable
=CUT
>192.168.1.1 on one NIC for internal, and a dhcp assigned address on a 2nd
=CUT
> ipfwadm -F -a accept -m -S 192.168.0.0/24 -D 0.0.0.0/0
192.168.0.0/24 a typo or shouldn't you masq 192.168.1.0/24 ???
leo
------------------------------
From: mike <[EMAIL PROTECTED]>
Subject: Problem mailing with direct connect computers
Date: Thu, 18 Mar 1999 09:16:57 -0500
I am running RedHat 5.1 on both computers. One is a pentium 166
and the other is an AMD 486 66.
I set up two computers with a serial null modem cable and
have ppp runnning on both. I used the command line:
pppd -detach crtscts lock 192.1.1.17:192.1.1.18 /dev/cua3 &
I reversed the IP's on the other computer.
I list the IP's and made up host names for each in /etc/hosts
I can telnet and ftp between both computers and have also
mounted directories remotely, but I can't send e-mail between
each computer. I am using Pine. I can e-mail to different users
within the same computer, but not between them. Talk won't work
either.
What are the issues as to why I can't do the e-mail?
Some of my thougts were:
Possible need to set up a DNS. If so exactly how do I do it?
Is there a configuration file for the mail function to tell it
that the other computer exists?
Is it possible that one of the program functions has a bug like
named and needs a patch or rpm update?
Thanks
Mike
------------------------------
From: "Martin Cole" <[EMAIL PROTECTED]>
Subject: Encrypted Passwords on a Linux Samba box?!
Date: Thu, 18 Mar 1999 14:17:17 -0000
Can anyone help me with setting up the encrypted passwords facility on my
linux box. I have taken a look in my smb.conf, and turned on the encrypted
passwords facility, and I understand that I require a smbpasswd file, but
I'm unsure where to go from here.
I'm running redhat 5.2
Martin
[EMAIL PROTECTED]
------------------------------
From: "Chris Cantwell" <[EMAIL PROTECTED]>
Subject: Re: Help: Network card problems
Date: Thu, 18 Mar 1999 09:07:45 -0500
Use the Intel Softset2 utility under DOS to setup up the card. (If you don't
have this, download it from www.intel.com) Use an appropriate unused IO
addr and IRQ (i.e., io=0x300, irq=10). Disable Plug and Play, and Flash
EEPROM Programming. You can keep the defaults for the rest. Save the
settings with F10.
The EtherExpress Pro/10+ (ISA) card uses the eepro driver. Reboot to Linux
and modify the conf.modules to reflect the new settings:
alias eth0 eepro
options eepro io=0x300 irq=10
Note: the 2.2 kernel eepro driver gets the irq from the NIC's EEPROM and
doesn't need the irq= setting
Warning: "insmod eepro" without arguments will probably hang your machine.
The eepro driver _really_ needs an io address.
Make sure your /etc/sysconfig/network-scripts/ifcfg-eth0 reflects the
correct network settings (IP, subnet, default gateway)
try to manually start the network with:
# ifup eth0
check the eepro module is installed:
# lsmod
check the network configuration:
# ifconfig
ping localhost (loopback), your machine IP address, and another address on
the network.
If this all works, then reboot to make sure your NIC is recognized at boot.
BTW, read the NET-3-HOWTO and module HOWTO for more information.
Chris Cantwell
O'Loughlin wrote in message <7coka4$nd5$[EMAIL PROTECTED]>...
>I have a Intel EtherExpress PRO/10+ (PnP) network card running on a RH 5.1
>system. I can't get any of the intel modules to load. The card isn't
>recognized in the install ether, How do I get the bloody thing to work? or
>even what module do you use? I'm new to this linux OS and the HOWTO's
>haven't helped me.
>
>Help!
>
>Mail me at [EMAIL PROTECTED]
>
>Thanks
>Diarmaid O'Loughlin
>
>
------------------------------
From: "Chris Cantwell" <[EMAIL PROTECTED]>
Subject: Re: Ethernet Card on 2.2.3
Date: Thu, 18 Mar 1999 08:44:43 -0500
This is kerneld in 2.2. You need to disable IPX/Ax.25/Appletalk probing
with the following lines in conf.modules:
alias net-pf-3 off
alias net-pf-4 off
alias net-pf-5 off
Chris Cantwell
Joe Cloutier wrote in message ...
>I recently upgraded my kernel from 2.0.36 to 2.2.3. i had an Intel
>EtherExpress Pro/10+ running fine under 2.0.36. Now, that I have upgraded
>during bootup I get the message...
>
>monprobe cannot locate module net-pf-4
>monprobe cannot locate module net-pf-5
>monprobe cannot locate module net-pf-4
>...
>delaying initialization of eth0
>
>and it subsequently will not initialize my card. This was Redhat 5.2 with
a
>normal upgrade (not RPM). Any suggestions would be appreciated. I have
>exhausted all the HOW-TOs I could find.
>
>Joe Cloutier
>
>
------------------------------
From: "Chris Cantwell" <[EMAIL PROTECTED]>
Subject: Re: Ethernet card IP setup in RedHat 5.2
Date: Thu, 18 Mar 1999 09:14:08 -0500
Check the network card settings in Win95, and use those for your Linux
setup. Put these settings in /etc/conf.modules
alias eth0 <driver>
options <driver> io=0x000 irq=x
restart the network
# ifdown eth0
# ifup eth0
Check the network settings
# ifconfig
Chris Cantwell
Icon wrote in message ...
>Ethernet card IP setup in RedHat 5.2
>
>I have 2 linux boxes at home. I installed RedHat 5.2 off of 2 differnet
>CD-Roms. The first one works mint. The other one however is having problems
>getting onto the internet.
>
>The IP,DNS,subnet mask is OK. Default gateway is ok as well
>
>After it boots up the eth0 is initialized and ifconfig shows all correct
>info,,,except irq and DMA address which I am not sure what they are so I
>tried 10 210-300 and it didn't change anything.
>
>I can ping the static IP assigned to the card but nothing outside the box
>includibg the default gateway nor the other box that is humming in the same
>IP block connected to the same HUB.
>
>The NIC is OK because the box is dula-booted into 95 and 95 is flying on
the
>net with it.
>
>What else can I do????
>
>If you have any suggestions please post them.
>
>Thanks
>
>
>
------------------------------
From: "Sam Grainger" <[EMAIL PROTECTED]>
Subject: Re: Help! Cant route WIN98 through Linux box!
Date: Thu, 18 Mar 1999 09:20:12 -0600
Have you tried IP Masquerading? Check out the mini-how-to...
Sam
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.questions,comp.os.linux.setup,alt.2600
Subject: Re: Setting up 2 nics under Linux RH 5.1
Date: Thu, 18 Mar 1999 12:58:10 GMT
In article <[EMAIL PROTECTED]>,
the new guy <[EMAIL PROTECTED]> wrote:
> I'm trying to install t nics under Red Hat linux 5.1
> One is a 3c509 (pci) and the second is a 3c503 (isa0
>
> Now, the 3c506 was detected and installed fine when I installed RH5.1
> and is ETH0.
>
> I tried to set up the 3c503 as ETH1 by adding the following entries
> in my conf.moduels file
>
> alias 3c509
> alias 3c503
> * snd card stuff*
> options 3c503 io=0x350 irq=9
Double check this. Your io may actually be 0x300 instead of 0x350.
Naturally, double check your IRQ as well.
Kirk Lawson
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
Date: Thu, 18 Mar 1999 13:37:36 +0000
From: Owen Rees <[EMAIL PROTECTED]>
Crossposted-To: comp.object.corba
Subject: Re: a free software dynamic Firewall gateway for MICO & alike ?
Mike wrote:
>
> Have you checked with TIS Labs on whether the firewall toolkit might be
> extended to add an IIOP proxy? TIS Labs is heavily involved with CORBA
> and the firewall toolkit is still around, so this might be a
> possibility. I know that Network Associates (which bought TIS last
> year) is working on adding an IIOP proxy to Gauntlet, so the same proxy
> could be added to FWTK (since they're the same core technology).
>
> In terms of doing this yourself, I'm not sure why you think there needs
> to be a firewall based ORB. In terms of FWTK and VisiBroker, you'd need
> to proxy OAD requests (which are UDP requests on a known port). When
> the OAD responds with an IOR, it needs to translated so that the host
> field in the IIOP/IOR Profile is changed to be the outside IP address of
> the firewall. Then take the port and create a generic tcp/ip proxy
> (i.e. a plug) listening on that port for connections from the client
> which requested the IOR. The client talks to the firewall, the firewall
> changes the IPs (via standard transparency) and everyone is happy.
> Okay, so if the client calls clone() or another IOR is passed over the
> TCP connection, you might need to create another proxy. I'm no CORBA
> expert, so I'm probably missing some connection stuff, but you get the
> idea.
>
> Okay, so this is a bit inefficient. Really want a single IIOP proxy
> which listens on UDP port for location service requests then starts up a
> IIOP proxy as needed as a separate thread rather than as a separate
> process. Keeping all of them together should allow the firewall to
> handle the case where 2 CORBA servers give the same port to an external
> client. Then you can just change one of the ports given to the client.
>
> So, what am I missing? It has to be more difficult than this....
The tricky bit is enabling access to all the objects whose references
pass as parameters in either direction, without enabling access to any
object whose reference was not passed as a parameter. Being able to
revoke previously granted access is also a useful thing to have. It
helps to have a strategy for the initial set of objects that get
exposed. Having this work for objects implemented on arbitrary and
diferent ORBs is also useful. There are also other things to worry about
such as how you avoid having an ever increasing number of proxies for
objects that are created and destroyed in arbitrary locations.
>
> Mike
>
> Derek Viljoen wrote:
> >
> > I'm amazed that so many people are interested in this and I know of a
> > product under development, but knowing the company doing it, it may
> > NEVER see the light of day. The company is NEC, and the Japanese don't
> > enter "new" markets first. This is a shame because they've got an IIOP
> > proxy server that works seemlessly. They run it with both Orbix and
> > Visibroker, and were working out a deal with IBM to integrate it with
> > their firewall (I don't know the name). It's based on SOCKS which NEC
> > jointly developed with another company (I think Aventail?). NEC also
> > has a software VPN product based on the same technology. It blows me
> > away to read how many people want this stuff, and to know that NEC is
> > sitting on it!
> >
> > You should contact their marketing people!
If this is an implementation of the SOCKS-based part of the firewall
proposal recently adopted by OMG then it has its uses, but is not a
completely general solution.
> >
> > Derek Viljoen
> >
> > Alain Coetmeur wrote:
> > >
> > > anyway the more realistic solution would
> > > be to develop a IIOP gateway based on DII/DSI
> > > relaying, and a virtual implementation repository.
> > > depending on the address domain (inside, outside)
> > > of the client, the gateway would convert the IOR
> > > from "insider" to "insider" or reverse.
> > > I think it can be possible, but compute intensive,
> > > since we should unmarshall each IIOP message
> > > to find the IOR and convert them.
> > >
Implementing one of these is a good way to learn more than you ever
wanted to know about the state of various ORB implementations,
interoperability issues, and some of the less well explored corners of
CORBA. On the other hand, a gateway built like this is completely
transparent to the application components (apart from performance
issues), and works just fine with ORBs that are available today. (I.e.
you do not need any of the extra IOR or protocol stuff that is defined
in the OMG firewall proposal.)
--
Owen Rees, HP Laboratories, Bristol
Any statements in this message are personal opinions and do not
represent the views of HP.
------------------------------
** 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
******************************
