Linux-Networking Digest #725, Volume #10 Fri, 2 Apr 99 23:13:40 EST
Contents:
two 3c590 nics do not route (Bernd Broermann)
Re: SAMBA/PASSWORDS (Basil Blume)
EQL only working one way ("Scott R. Knight")
IP forwarding situation ([EMAIL PROTECTED])
Re: Slow ethernet LAN driving me crazy!! (brian moore)
Re: NE 2000 ISA config problems (KNE2000LC works 98 not linux) (Ron Watkins)
HUB TOPOLOGY QUESTION ("mech")
NETZERO setup on Linux ([EMAIL PROTECTED])
Re: RedHat Lousy Support ([EMAIL PROTECTED])
Re: Odd line (to me) in messages file ("Brian E. Parker")
Re: winmodems (Graham Murray)
Re: Finding a process on a port (Graham Murray)
Re: STAC compression??? (Graham Murray)
Setting transceiver for eth card (Andreas Dilger)
Re: Different IP in one network (Lew Pitcher)
----------------------------------------------------------------------------
Date: Thu, 01 Apr 1999 20:35:03 +0000
From: Bernd Broermann <[EMAIL PROTECTED]>
Subject: two 3c590 nics do not route
Hi,
I have 2 vortex cards (3c590) cards in a linux box 2.0.36 as modules
Kernel is configured for routing.
LAN ex 49.1.1.215 49.0.0.0
linuxgateway eth0 49.1.1.9 eth1 10.0.0.1
ISDN internet 10.0.0.2 ISDN Router Brick XM
the linuxgateway cannot route from 49.1.1.215 to 10.0.0.2 when I use:
49.1.1.215:~# ping 10.0.0.2
100% lost
10.0.0.2:~# ping 49.1.1.215
100%lost
I only can ping both linuxgateway nics 49.1.1.9 and 10.0.0.1
internet access works fine, because there are only proxys in use.
But why did my router not route ???
I is because of the equal nic cards ?
thank you for help
bernd
============================================================================
Routingtables
linux-router#
Kernel IP routing table
Destination Gateway Genmask Flags MSS Window irtt
Iface
10.0.0.0 0.0.0.0 255.255.255.255 UH 1500 0 0
eth1
10.0.0.2 0.0.0.0 255.255.255.255 UH 1500 0 0
eth1
10.0.0.0 0.0.0.0 255.255.255.0 U 1500 0 0
eth1
127.0.0.0 0.0.0.0 255.0.0.0 U 3584 0 0
lo
49.0.0.0 0.0.0.0 255.0.0.0 U 1500 0 0
eth0
0.0.0.0 10.0.0.2 0.0.0.0 UG 1500 0 0
eth1
49.1.1.215:~ # netstat -nr
Kernel IP routing table
Destination Gateway Genmask Flags MSS Window irtt
Iface
49.1.1.215 0.0.0.0 255.255.255.255 UH 0 0 0
eth0
10.0.0.2 49.1.1.9 255.255.255.255 UGH 0 0 0
eth0
49.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0
eth0
127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0
lo
0.0.0.0 49.1.1.9 0.0.0.0 UG 0 0 0
eth0
brick:> netstat -r
Typ Destination Netmask Gateway Metric
Interface Proto
LOC 10.0.0.0 255.255.255.0 10.0.0.2 0
en2 local
DEF default 0.0.0.0 1
Crown local
LOC 212.6.157.1 255.255.255.255 212.6.157.200 0
Crown other
NET 49.0.0.0 255.0.0.0 10.0.0.1 2
en2 local
------------------------------
From: Basil Blume <[EMAIL PROTECTED]>
Subject: Re: SAMBA/PASSWORDS
Date: Thu, 01 Apr 1999 14:06:50 -0700
Reply-To: [EMAIL PROTECTED]
Phil-
I was haveing the same problem with Win98 and Linux. Here's what I did after
reading the entription.txt in samba docs:
- create the smbpasswd file as show in the encription.txt file then use
smbpasswd [username] to create a password on the linux box for use with samba
- if necessary edit the smbusers file to map smb users to linux users - I
did this becuase according to Win98 (issue the command 'net config' as a dos
prompt) my user name is BASIL and my linux user is Basil - two differnet user
names.
- edit the smb.conf file and uncomment (my Red Hat 5.2 Install already
had the entries) or add the entires in the Global section:
encrypt password = yes
smb passwd file = /etc/smbpasswd [or where ever]
user name map = /etc/smbusers [or where ever]
[I also uncommented the 'set passowrd level = 8' and 'set user level =
8' to allow uper and lower case password and username matches - this is
probably overkill given the above and is less secure.]
- stop and restart samba
- Next, at my Win98 box I went to network neighboorhood and double-clicked
on the linux box and entered the password created. Now I'm able to see the
linux box in explorer!
Basil Blume
phil morle wrote:
> Hi All,
>
> I have setup SAMBA with Linux so that I can successfully browse Network
> Neighborhood in Win98 and see the Linux box... when I go to map a network
> drive I am asked for a password. Each time I give it it tells me that the
> password is wrong... any ideas?
>
> Thanks in advance
>
> Phil Morle
------------------------------
From: "Scott R. Knight" <[EMAIL PROTECTED]>
Crossposted-To: comp.protocols.ppp
Subject: EQL only working one way
Date: Thu, 01 Apr 1999 16:37:48 -0500
This is a multi-part message in MIME format.
==============7488EB8D97C483AD21A05614
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Redhat 5.2
Kernel 2.2.4
ppp-2.3.7
eql-1.2
two v.34+ modems
I am looking for help to figure out how to get two modems working for a
poor man's double bandwidth solution. I have been fiddling with EQL and
having limited success with it. It seems to work for outbound traffic
just as it should, but does nothing for the inbound data. The
instructions for setting it up are:
http://home.indyramp.com/eql/eql-1.html
and were very helpful. They just stop short after describing how to set
it up, so there is no debugging help. I found out from my ISP that they
_are_ running Livingston Portmaster, but have yet to get to speak to a
sysadmin directly. Since I don't know anything about the setup on their
end, I wouldn't know what to ask for anyway. I am just assuming that
they know what they are doing (possible danger <g>).
What happens when I start the setup: I dial up with one modem on ppp0
and it connects like it should. Then I go about configuring the eql
device and enslaving ppp0 to it and making it the default route. The
functionality of the connection never changes. If I ftp files out or
in, the data goes as expected through the connection (I am monitoring
this with pload, BTW). Then I can connect with the second modem as ppp1
and enslave that to the eql device as well. Now, all the data coming in
from the ISP goes through ppp1 while ppp0 just sits idle. If I ftp
files OUT to a server, the data goes out twice as fast and equally
through both ppp0 and ppp1 just as it should. Now I am stuck.
BTW, if you have any ideas at all, please let me know right away since
they agreed to let me do this for free only for a couple days until I
decide whether I want to pay for 2 fast modems and another connection.
Anyone?
--
Scott R. Knight mailto:[EMAIL PROTECTED] http://www.redseaco.com/
4885-A McKnight Road, #298
Pittsburgh, PA 15237-3972
Phone 888-891-5862
==============7488EB8D97C483AD21A05614
Content-Type: text/x-vcard; charset=us-ascii;
name="sknight.vcf"
Content-Transfer-Encoding: 7bit
Content-Description: Card for Scott R. Knight
Content-Disposition: attachment;
filename="sknight.vcf"
begin:vcard
n:Knight;Scott R.
tel;fax:888-891-5862
tel;work:888-891-5862
x-mozilla-html:FALSE
url:http://www.redseaco.com/
org:Redsea Works Co.
version:2.1
email;internet:[EMAIL PROTECTED]
title:CTO
adr;quoted-printable:;;4885-A McKnight Rd., #298=0D=0A;Pittsburgh;PA;15237-3972;USA
x-mozilla-cpt:;-18672
fn:Scott R. Knight
end:vcard
==============7488EB8D97C483AD21A05614==
------------------------------
From: [EMAIL PROTECTED]
Subject: IP forwarding situation
Date: Thu, 01 Apr 1999 21:43:38 GMT
Hi! Here's my network config...
Company workstations
|
| +--------+ ( )
+~~~~~~~~~~~~+ eth1| LINUX | +~~~~~~~~~~~~+ ( )
|NT 4 Domain1|<--------->|REDHAT |<------->|NT 4 Domain2|---->(Internet)
+~~~~~~~~~~~~+ |FIREWALL|eth0 +~~~~~~~~~~~~+ ( )
+--------+ ( )
2 NT domains with a Linux RedHat firewall in between.
I'm an NT Admin who walked into this situation with no knowledge of Linux.
What I need: Employees from Domain1 need to be able to access files and shares
in Domain2. These two domains can be set up as trusts. Also, users dialing in
to a server equipped with RAS in Domain2 need to be validated to access files
and shares in Domain1.
I know it's probably a simple operation, but I'm a Linux newbie, and any help
would be greatly appreciated. I've been told that what I need to do here is IP
forwarding. Is that true, and if so, how?
Thanks in advance...
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: [EMAIL PROTECTED] (brian moore)
Crossposted-To: comp.os.linux.setup,comp.os.linux.hardware,comp.os.linux.misc
Subject: Re: Slow ethernet LAN driving me crazy!!
Date: 1 Apr 1999 22:19:14 GMT
On 1 Apr 1999 15:11:15 GMT,
Markus Wandel <[EMAIL PROTECTED]> wrote:
> In article <[EMAIL PROTECTED]>,
> brian moore <[EMAIL PROTECTED]> wrote:
>
> >(And, yes, you can see assymetric failures with bad twisted pair cables,
> >since one pair is Tx and one is Rx, the system with the icky Rx pair
> >will have problems receiving, though less problems sending.)
>
> Permit me to add that in my experience, many if not most 10baseT cables
> are bad. The pairing of pins on an RJ45 connector is
>
> 1,2
> 3,6
> 4,5
> 7,8
>
> of which 10/100baseT use 1,2 and 3,6. A lot of cables, though are wired with
> the twisted pairs hooked up to 1,2 / 3,4 / 5,6 / 7,8. This happens to work
> for some fairly nontrivial lengths and 10baseT but it is WRONG and can cause
> problems, and of course they will be asymmetrical ones because one direction
> is in fact on a twisted pair.
Indeed. We had such a pair here (NEVER let your phone guy wire your
LANS). Actually, his was a bit worse: he'd paired them as 1/6 and 2/3.
(He's a phone guy so he thinks cables have at most 4 wires.) This made
it check out just fine and dandy on a cheap ethernet tester, but when it
was in service it would get really slow sometimes and one additional
machine added to that segment pretty well killed it.
Actually, the phone guy did it more than once. (Every single cable he's
run he's had to come out and fix. If it didn't all have to go through
his unlabelled-punchdown-block-Hell we'd do it ourselves, but he's got
Job Security.)
> It's hard to check if your cable is one of the "bad ones" unless you can see
> enough of the wires though the connector. A twisted pair is usually a solid
> colour and white wire with stripes of the solid colour.
>
> I've seen 20-30ft runs of "bad" cable work fine for 10baseT but you never know.
Icky. I wouldn't even try 2ft.
--
Brian Moore | "The Zen nature of a spammer resembles
Sysadmin, C/Perl Hacker | a cockroach, except that the cockroach
Usenet Vandal | is higher up on the evolutionary chain."
Netscum, Bane of Elves. Peter Olson, Delphi Postmaster
------------------------------
From: Ron Watkins <[EMAIL PROTECTED]>
Subject: Re: NE 2000 ISA config problems (KNE2000LC works 98 not linux)
Date: Fri, 02 Apr 1999 14:31:38 -0800
IRQ 4 is a bad idea. That's where COM1 is. That shouldn't even work well
under 98, much less Linux, unless 98 moved your COM1 out of the way to some
other IRQ.
Generally, you want a card set to IRQ 9, 10, or 11. You need to make sure
that your PCI bios marks that IRQ as unavailable to PCI cards. If you try to
share IRQs, it won't work and may crash. Those settings are in CMOS setup
when you first start your computer. Assuming you want the card on IO 300, IRQ
9 (good default values), mark IRQ 9 as 'unavailable' in the PNP setup.
Exactly how you go about doing that varies from system to system.
Also, when you first get to CMOS Setup, be careful to write down all your old
settings -- if you change settings to a bad value you may end up with a
non-booting system. In that case, you'll need to open the case and clear the
CMOS by shorting a jumper, and then restore all your old settings manually.
So write them down.
Most network cards come with a DOS config utility. Boot up in DOS-only mode
in win98, and set the card to NON-plug and play, IRQ 9, IO 300. Then
configure Linux to look for the card at that IO port and IRQ. (you can often
pass this as an argument from LILO at bootup: personally I usually just
compile the network card driver, with appropriate settings, right into the
kernel.)
Each of these steps will probably take some time to figure out, but if you
successfully follow them, your card will almost certainly work. You'll then
have to configure your IP addressing and DNS and netmasks and things -- but
first get the card working.
<<RON>>
------------------------------
From: "mech" <[EMAIL PROTECTED]>
Subject: HUB TOPOLOGY QUESTION
Date: Fri, 02 Apr 1999 01:32:10 GMT
I have the 1mb modem, a linux box and 2 other computers.
Is it posible to just plug the one meg modem into the hub and plug the other
computers directly into the hub? When I do this the computers talk but
there is no internet connectivity.
If I multihome the linux box and plug it to the hub and plug the other
computers directly into the hub, it works.
Is this the only way it can work?
Thanks
------------------------------
From: [EMAIL PROTECTED]
Subject: NETZERO setup on Linux
Date: Fri, 02 Apr 1999 16:23:14 GMT
Has anyone been able to configure Netzero on Linux. Netzero is a "free" ISP
that uses java to connect to their site.
Any input appreciated!
PS: please email responses to [EMAIL PROTECTED]
(I can't get newsgrps from work ;-) )
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: redhat.networking.general
Subject: Re: RedHat Lousy Support
Date: Fri, 02 Apr 1999 23:24:18 GMT
Allen! Bite your tongue!
In article <isFM2.29493$[EMAIL PROTECTED]>,
"Allen" <[EMAIL PROTECTED]> wrote:
>
> I have business to run. I don't care how much I have to pay to get the
> service.
(snip)
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: "Brian E. Parker" <[EMAIL PROTECTED]>
Crossposted-To: linux.redhat.misc
Subject: Re: Odd line (to me) in messages file
Date: Fri, 2 Apr 1999 09:07:05 -0600
Brian E. Parker wrote in message ...
>If there is something else I need to do to kill any news access (including
>deleting the news user), please let me know.
Thanks to everyone for pointing out the cron.hourly job that was activating
an inn-* script.
-BEP
------------------------------
From: Graham Murray <[EMAIL PROTECTED]>
Subject: Re: winmodems
Date: 02 Apr 1999 15:12:54 +0000
Sami Yousif <[EMAIL PROTECTED]> writes:
> Not only that, but two models, one a "windmodem" the other a
> "standard" modem were only a couple of rows apart, with only a
> couple of lines on the side of the box hinting that one was a
> winmodem ("win95/98/NT4.0 required. May not work with other
> operating systems").
>
> I can understand the frustration of many users since I (and IMNSHO I
> consider myself to be a knowledgeable power user) had to "dig a
> little" to figure out which were winmodems and which were "real"
> modems; its no wonder the average user would not know the difference
> and go for the "better deal"....
It does not help when the published specification for an *external*
modem (which cannot be a Winmodem) states that the minimum
requirements are a Pentium processor and Win95/98. I assume that this
was referring to the bundled software not the hardware.
------------------------------
From: Graham Murray <[EMAIL PROTECTED]>
Subject: Re: Finding a process on a port
Date: 02 Apr 1999 15:42:26 +0000
"Jim Laird" <[EMAIL PROTECTED]> writes:
> How do you figure out which process has attached to a given port? I was port
> scanning my machine to see what was open and I came across port 635 (the NFS
> mount service). Luckily I found it in /etc/services, but what if I didn't?
> Is there an easy way (i.e. a program) to figure this out, i.e. find a
> process id attached to a given port?
This capability was introduced into netstat in net-tools 1.51. Use
netstat -p
------------------------------
From: Graham Murray <[EMAIL PROTECTED]>
Subject: Re: STAC compression???
Date: 02 Apr 1999 15:54:18 +0000
[EMAIL PROTECTED] (Erik Hensema) writes:
> You'll have to upgrade your kernel to version 2.2, and use the new isdn4linux
> kernel drivers and utilities from ftp.suse.com/isdn4linux/, compile the
> utilities, compile the kernel (in that order), and compile the compressor
Is it possible to use this compression with "standard" serial port
ppp?
My ISP supports Stac compression (of various types) but not any of the
ones which come standard with pppd or kernel 2.2.x.
------------------------------
From: [EMAIL PROTECTED] (Andreas Dilger)
Subject: Setting transceiver for eth card
Date: 3 Apr 1999 03:37:52 GMT
I've looked in the Ethernet HOWTO, but I couldn't find any info on this.
I bought a cheapo Western Digital 8013 card to use as the second card in
my Linux firewall/masq setup, and I can't seem to get it to work on the
10baseT port. I know the card worked under Win95, and the original card
in my system (ne2k_pci) works fine too. I think it may be that the WD
card is using the wrong transceiver (there is a 10baseT, AUI, and BNC
connector on the card). I can ping the interface on the card, so I
believe the driver is installed OK.
I tried using the "ifport" program that comes with the PCMCIA tools, but
it didn't help me at all. Anyone have any hints?
Cheers, Andreas
--
Andreas Dilger University of Calgary \"If a man ate a pound of pasta and
Micronet Research Group \ a pound of antipasto, would they
Dept of Electrical & Computer Engineering \ cancel out, leaving him still
http://www-mddsp.enel.ucalgary.ca/People/adilger/ hungry?" -- Dogbert
------------------------------
From: Lew Pitcher <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup,hk.comp.os.linux
Subject: Re: Different IP in one network
Date: Thu, 25 Mar 1999 21:39:36 -0500
Jack Cheng wrote:
>
> Hello,
>
> I have a network system using RH5.2 and running fine. The IP is 192.168.16.1
>
> Because I want ot learn more about linux, so I hooked up one more linux box
> in my existing lan (of course running as a SERVER and IP is 192.168.18.1),
> so i can change anything in the new linux box for my testing.
>
> I cannot ping the 192.168.16.1 from the new linux box (192.168.18.1), is
> possible set two different IP in one lan system?
>
> Best Regard
> Jack Cheng
If you have defined your 'subnet mask' to be more than 16bits (i.e.
255.255.0.0),
you'll have to stick a router between the two Linux boxes. You see, they
look to
be on different subnets, and would ignore each other if they were.
Rather than add more hardware to your lan, either
a) change the IP address on the new Linux box to a 192.168.16.x address
(like your first Linux box has), or
b) change the subnet masks on *both* Linux boxes to be 255.255.0.0
(in order to put both machines on the same subnet).
Of course, option (b) above may have detrimental effects on your
existing LAN.
--
Lew Pitcher | If everyone has an angle, why
JOAT-in-training | are most of them so obtuse?
------------------------------
** 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
******************************
