Linux-Networking Digest #51, Volume #11 Wed, 5 May 99 17:13:43 EDT
Contents:
Re: NIC: 3C509B-TX Drivers? (Paul Black)
Problem accessing a Linux machine from an address range (Barclay McInnes)
Crontab woes with RH5.2 ([EMAIL PROTECTED])
Re: dns reverse lookup (Clifford Kite)
Re: Using tar as full backup (Bob McLaren)
Re: ?Windows NT dialup to Linux PPP server? (Richard Birchall)
Re: PCI Network Card that definetly work in Linux and Window 98 (Chris Moseng)
Trouble setting up IntelEtherExpress PRO/10+ ISA ("Roy Magne Gjemble")
Using tcpdump to monitor OSPF routing messages (Roland Thienpont)
IPFWADM ("NEVSKI")
Re: Samba as a printer server->suggestions ("Lee Sharp")
Re: HELP!! DNS Issues (mist)
Re: Trouble with 3Com (Dr Vincent C Jones PE)
Re: Home Office Setup (Dr Vincent C Jones PE)
Re: Routing / NAT problems (Luca Filipozzi)
RASDL (Chris Moseng)
Re: Linuxconf http access error?? (Hamka Hj Suleiman)
Re: PPP problems using Redhat 5.2 Usernet (James Lee)
THTTPD ("Shawn Conrad")
----------------------------------------------------------------------------
From: Paul Black <[EMAIL PROTECTED]>
Subject: Re: NIC: 3C509B-TX Drivers?
Date: Wed, 05 May 1999 16:03:59 +0100
Reply-To: Paul Black <[EMAIL PROTECTED]>
"Clemens Buchacher" <[EMAIL PROTECTED]> wrote:
> The Ethernet-HOWTO of LDP says that "the support for the Cyclone 'B'
> revision was only recently added. To use this card with older v2.0 kernels,
> you must obtain the updated 3c59x.c driver from Donald's site at:
> Vortex-Page <http://cesdis.gsfc.nasa.gov/linux/drivers/vortex.html>
>
> Actually, I don't need that update because SuSE Linux 6.1 has Kernel 2.2.5,
> do I?
It worked fine on my machine with 2.0.36, 2.2.3 & 2.2.5 kernels.
> So do I need the 3c95x.c driver or which driver can I use? And how do I
> install it (before installing Linux, while installing Linux, after
> installing Linux,...)?
When I installed RH 5.2, everything was set up to load the 3c905 driver
as a module (not sure what decides to load this module and not another -
modules is am area I need to learn more about). I tend to compile my
kernels with everything in that I use normally which obviously includes
the network driver. In short, a kernel reconfigure & recompile should get
you to your destination.
> I was delighted if you could also reply this E-mail to me (cuz I usually
> don't have the time to read all the newsgroups to which I've sent this
> mail).
Time to learn about DejaNews then.
Paul
------------------------------
From: [EMAIL PROTECTED] (Barclay McInnes)
Subject: Problem accessing a Linux machine from an address range
Date: Wed, 05 May 1999 17:28:40 GMT
Reply-To: [EMAIL PROTECTED]
This is just weird. About 3 in the afternoon yesterday, our primary
mailserver for one of our buildings, which runs on RedHat 4.1,
sendmail 8.8.9, decided to stop accepting connections from anyone in a
certain class C address range. No Telnet, no FTP, and most important,
NO MAIL. any attempt to connect with my client (Lookout Express)
returned the following dialog:
connection to server failed,protocol.pop.3 port 110, secure(ssl) no,
socket
error:10060, error number:0x800CCC0E
Oddly enough, web access to it still works. We run an Intranet site
there, and everyone can still see that. If I switch to a different
class C address range, I access it fine. Example: I was using
204.174.40.99 and had no luck. I switch to 204.239.27.131, I get
mail, I can telnet in, I can FTP, life is good. The problem is
twofold, we have over a hundred workstations on the 204.174.40 class C
with not enough free addresses on the 204.239.27 class C to accomodate
them, *and* WHY DID THIS HAPPEN IN THE FIRST PLACE? My colleague and
I spent over 6 hours trying to repair the original server with no
success. We eventually forklifted everything to a different mail
server, recreated all the accounts, moved the old mail, and aliased
all the old accounts and changed the MX records to reflect the new
server. Everyone works now, but we *still* want to know what the hell
happened. Anyone out there run into something like this, or have an
idea as to where we can start poking around? We're fresh out of
ideas.
Nobody messed with it, it's not running ipfwadm, and there are no
routers in between my workstation and the server. Basically, nothing
that can affect access through an access list. We've also checked
IP_allow and deny rules for sendmail, they're all fine. hosts.allow
and hosts.deny as well.
Thoughts?
Barclay McInnes
Multiactive Software.
------------------------------
From: [EMAIL PROTECTED]
Subject: Crontab woes with RH5.2
Date: Wed, 05 May 1999 18:25:10 GMT
I'm hoping someone can help me here.
I am trying to set up a cron job using crontab -e.
The first time I tried this, I got an error message:
no crontab for root - using an empty one
/bin/sh: /usr/bin/vi: No such file or directory
crontab: "/usr/bin/vi" exited with status 126
I am using RH5.2, and found that vi has been "upgraded" to vim.
Ok.. I set up a symlink to point /usr/bin/vi to /bin/vi.
Now when I type crontab -e I get this:
no crontab for root - using an empty one
crontab: no changes made to crontab
Then it dumps me to the command line again.
Anyone have an idea why?
Any and all help would be greatly appreciated.
TIA
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: kite@NoSpam.%inetport.com (Clifford Kite)
Subject: Re: dns reverse lookup
Date: 5 May 1999 14:49:15 -0500
Luca Filipozzi ([EMAIL PROTECTED]) wrote:
: In article <7gq0f2$tvf$[EMAIL PROTECTED]>, francesc_guasch@my-
: > my /etc/resolv.conf is:
: > nameserver my.isp.dns.server # real numbers edited here
: >
: > I noticed I have not the problem if I remove the resolv.conf file.
: Put this at the top of resolv.conf
: order hosts bind
: This tells your machine to look in /etc/hosts first, then to use dns.
Not unless things have really changed recently. The /etc/host.conf file is
where "order hosts bind" belongs.
--
Clifford Kite <kite@inet%port.com> Not a guru. (tm)
/* 97.3% of all statistics are made up. */
------------------------------
From: Bob McLaren <[EMAIL PROTECTED]>
Subject: Re: Using tar as full backup
Date: Wed, 05 May 1999 16:31:58 GMT
Thanks! That worked great.
Bill Long wrote:
> tar -zcf mybackup.tgz . OR
> tar -zcf mybackup.tgz *.* .*
>
> tar -zxf mybackup.tgz
>
> thats all you need my friend.
>
> Bob McLaren wrote:
> >
> > I am testing tar as a full system backup solution.
> > I used this to make the backup of the current directory structure: #
> > tar -cz . > bobs.tar.gz
> > And used this to restore it: # tar -xz < bobs.tar.gz
> >
> > Why did I get these error messages?
> > gzip: stdin: decompression OK, trailing garbage ignored
> > tar: Child returned status 2
> > tar: Error exit delayed from previous errors
> >
> > Your help is appreciated.
> >
> > --
> > Bob McLaren
> > Network Administration
> > Financial Statement Services, Inc.
> > HTTP://WWW.FSSI-CA.COM
>
> --
> www.bellanet.com - very nice web hosting services
> - web application services
>
> www.giftsgalore.com : www.longboys.net
> telnet://undying.longboys.net:4000 http://undying.longboys.net
--
Bob McLaren
Network Administration
Financial Statement Services, Inc.
HTTP://WWW.FSSI-CA.COM
------------------------------
From: Richard Birchall <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup
Subject: Re: ?Windows NT dialup to Linux PPP server?
Date: Wed, 05 May 1999 15:37:37 GMT
In article <[EMAIL PROTECTED]>,
jianhong <[EMAIL PROTECTED]> wrote:
> I'm trying to network my laptop(Slackware Linux 3.5) with my
> desktop (Windows NT 4.0 SP3) through a Null Modem Cable.
> The problem is I can't get the Windows NT dialup to work with the
> Linux PPP server. NT dialup complained that "there is an error with
> the modem or other device", while the PPPD complained that
> "LCP: timeout sending Config-Requests,
> Receive serial link is not 8-bit clean:
> Problem: all had bit 7 set to 1"
Your NT box has to send 'CLIENT\n' and the Linux box has to respond 'SERVER'.
This handshaking is required by NT RAS for null modem connections.
Regards,
Richard
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: Chris Moseng <[EMAIL PROTECTED]>
Subject: Re: PCI Network Card that definetly work in Linux and Window 98
Date: Wed, 05 May 1999 11:41:00 -0500
I got myself a Netgear FA-310TX 10/100 PCI card and it installed like
warm butter, if warm butter were plug and play.
It runs on the tulip driver, and *even came with that driver and
instructions on how to compile and install it as a module or as a kernel
component in the install disk.*
It was very convenient and made my day.
------------------------------
From: "Roy Magne Gjemble" <[EMAIL PROTECTED]>
Subject: Trouble setting up IntelEtherExpress PRO/10+ ISA
Date: Wed, 5 May 1999 21:09:30 +0200
I am having trouble setting up my network card. my /etc/conf.modules says:
alias eth0 eepro
options eth0 io=ox300 irq=03
When I boot I have no problem (i have also tried eepro100 but with that
module the "bringing up interace eth0" fails.) with the eth0, it says:
bringing up interface eth0 [OK]
I can also ping localhost and the cards IP-adress but none of the other
machines in the network.
dmesg says:
eth0: Set Rx mode to 1 adress
eth0: Set Rx mode to 1 adress
eth0: Set Rx mode to 1 adress
And I don't know what to do from here so I hope some of You can help me
/Roy Magne Gjemble
------------------------------
From: Roland Thienpont <[EMAIL PROTECTED]>
Subject: Using tcpdump to monitor OSPF routing messages
Date: Wed, 05 May 1999 18:31:38 +0200
Hi,
I'm trying to monitor with tcpdump the OSPF routing protocol messages. I
managed to see the hello packets but the problem is that I cannot see
the content of the packet. Any suggestions?
Roland
------------------------------
From: "NEVSKI" <[EMAIL PROTECTED]>
Subject: IPFWADM
Date: Wed, 5 May 1999 22:08:23 +0200
I have a cable modem and a redhat5.2 Kernel 2.0.36 running as a server . My
network is working, all machines can Browse, use ICQ and play QUAKE, QUAKE2
etc. on the net. My problem;
How do I open TCP ports to use videoconference programs like ICUII or
NRETMEETING??
With ipfwadm I was able to get something open but not the right ports for my
programs.
Can anyone give me an example ?? If my external IP = xx.xx.xx.xx and my
internal IP = yy.yy.yy.yy and the port I wanna open is a TCP 2011
I hope someone can help me, it drives me crazy!
Thnx,
NEVSKI
------------------------------
From: "Lee Sharp" <[EMAIL PROTECTED]>
Subject: Re: Samba as a printer server->suggestions
Date: Wed, 5 May 1999 15:07:59 -0500
Jose L Gomez Dans wrote in message <7gpvs6$9mk$[EMAIL PROTECTED]>...
| After tinkering about with samba for a couple of days, I am unable
|to have a decent printer server up and running. All that is needed is
access
|to a single printer (two in the future, if that one works :D). I'd like to
|know how one can achieve that. I read the smb.conf documentation, and the
|other documentation, but got swamped. The FAQ was quite useless, and the
|only resource which was somewhat useful was the SMB-HOWTO. My question is:
|does anyone have one such thing set up, and would be willing to share
|his/her knowledge with me (regarding smb.conf's and stuff like that)? I'm
|getting desperate now!!!
Gee... I am working on similar. :-) I have a partially working setup.
First, authentication is passed off to an NT server. <No need to worry
about encrypted vs. not encrypted, and it fits into the domain> It also is
a WINS client, and announces itself to a few subnets. The only user it is
working for also has an account of the same name on the linux box. That
account has write access to the spool directory.
| Also, does anyone know how to go around on the clients side, when
|they're using WfWg? I downloaded a TCP/IP driver, and disk sharing is no
|problem. It's only printing that doesn't happen!
WfWg will work the same as 95. If it doesn't work in 95, WfWg has no
chance. :-) Look at the logs to see if you are getting authentication
failures.
Lee
--
SCSI is *NOT* magic. There are *fundamental technical reasons* why it is
necessary to sacrifice a young goat to your SCSI chain now and then. *
Black holes are where God divided by zero. - I am speaking as an individual,
not as a representative of any company, organization or other entity. I am
solely responsible for my words.
------------------------------
From: mist <[EMAIL PROTECTED]>
Subject: Re: HELP!! DNS Issues
Date: Wed, 5 May 1999 19:56:31 +0100
Reply-To: mist <new$[EMAIL PROTECTED]>
[EMAIL PROTECTED] scribed to us that -
>Hello All,
>
>Here is the problem, I have a Local DNS server setup on my network. The db
>file is private.net.hosts and works fine for name resolution. I can't seem to
>get reverse dns working, nor does nslookup work. But nslookup does reverse
>lookups so the problem probably lies with reverse dns.
You should still be able to use nslookup to look up the address normally
even if reverse-lookup is not working. What's your /etc/resolv.conf
like?
>
>Anyone who may be able to answer these questions I would appreciete the help,
>BTW: the dns server is running on a RedHat Linux box.
>
Check the section on my site where I talk about setting up DNS for an
account with Demon. [1] The principles should be the same. Remember
that when you are setting up the named.conf file, you need to put the IP
addresses in *backwards* for the reverse lookups. E.g. for an IP range
of 100.200.300.*** the associated zone entry would be something like -
zone "300.200.100.in-addr.arpa"
[1] With Bind 8.
[2] The URL is <URL:http://www.misthaven.demon.co.uk/mist/noframes/bind.
html> (That should be one line if your newsreader doesn't pick it up.)
--
Mist.
------------------------------
From: Dr Vincent C Jones PE <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.networking
Subject: Re: Trouble with 3Com
Date: Wed, 05 May 1999 16:52:59 GMT
In article <7gncdj$[EMAIL PROTECTED]>,
Vesa S�rkel� <[EMAIL PROTECTED]> wrote:
>
>I'm having real trouble with 3Com.
>
>I have found a problem with 3C905B-TX cards, but 3Com insists that they
>are OK. You may also have this problem.
>
>The problem is sensitivity to line noise when receiveing data from
>network. Symptoms are network speed slowing down, dropped connections and
>problems with connections that have worked OK in the past. 3COM NIC
>DOCTOR (or 3c90xcfg.exe) will show a high number of framing, overrun and
>CRC errors.
>
>-- ... lots more deleted ...
>
>Vesa Sa"rkela" Keskuslaboratorio Oy ( KCL) tel. +358-9-4371430
>P.O. Box 70, FIN-02151 Espoo FINLAND
>[EMAIL PROTECTED] The Finnish Pulp & Paper Research Institute
>GSM 0405116047 fax +358-9-4371302
Saw a similar problem at a client site a few years ago, except with a
different card vendor. Interesting aspect of it was that the problem
only showed up when installed in a particular brand of motherboard! It
appeared to be a noise induced failure problem which would cause cards
to fail after a few weeks to a few months. Rather than get into taking
the tops off of chips to verify the exact mechanism, the client settled
with the contractor to simply replace all the cards with a different
brand (3com, if memory serves me correctly!). End of problem, other then
the extremely bad aftertaste from all the vendors fingerpointing blame
in every direction except themselves.
--
Dr. Vincent C. Jones, PE
Networking Unlimited, Inc.
[EMAIL PROTECTED]
------------------------------
From: Dr Vincent C Jones PE <[EMAIL PROTECTED]>
Subject: Re: Home Office Setup
Date: Wed, 05 May 1999 16:52:58 GMT
In article <7ga2da$4t9$[EMAIL PROTECTED]>,
June Lim <[EMAIL PROTECTED]> wrote:
>Hi,
>
>I am trying to setup a home office like the following: Please advise.
>
> ISP
> |
> |
> 56K modem (dce)
> |
> |
> serial int (dte)
> 2501
> e0
> |
> |
> home office(private ip addr)
>
>1. Can I connect the 56k modem to the serial int ? What cable and connector
>I have to use ?
If by "56k modem" you mean a V.90 asynchronous analog modem, the answer
is no. The Serial Interface on the 2500 series modems only support
synchronous serial communications. For an asynchronous modem, you would
have to use the auxillary port which is limited to a maximum data rate
of 34kbps. http://www.cisco.com has all the cable specs and
configuration documentation you may need.
>2. Should I use static IP addr/DNS/default-route assignment from
>the ISP ?
The answer will depend upon what you are trying to accomplish with
your home network and what your ISP is willing to support. In general,
if people need to reach machines on your home network (eg. for an ftp or
web server) you will need a static IP address. For just general internet
access from your home office, a dynamic address will work fine.
>3. If so, what should be the router config ? DDR, NAT...
Again, the answer depends upon the size and use of your home office
network. DDR only makes sense if you are "reaching out" only. To run a
server, you need to be full time connected (or have your ISP DDR to
you).
>4. What should be the network config (e.g. dns) for the home office machines
>?
This will depend upon the answers to the above. Try reading some of the
FAQ's floating around, for example, consider the various Linux FAQs on
connecting to ISPs via PPP and substitute "2501" for "Linux Box" where
appropriate. While you won't get a cookie cutter solution, you will get
an idea of the options available and their tradeoffs.
--
Dr. Vincent C. Jones, PE
Networking Unlimited, Inc.
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Luca Filipozzi)
Subject: Re: Routing / NAT problems
Date: Wed, 5 May 1999 09:56:22 -0700
In article <7gpn04$oka$[EMAIL PROTECTED]>, [EMAIL PROTECTED]
says...
> Hello,
>
> I have a problem concerning ipchains (I believe).
> The setup is like this:
>
> Internet <-> Linux router <-> Class C network (192.168.x.x)
>
> I have used ipchains with MASQ so the clients on the C network
> can access the internet, that works fine... but I have problems
> when I want to access individual computers sitting on the C network
> from the outside (from the internet). Something like generic NAT.
> I understand that I can't use the 192. addresses, so I need to map
> real IP's to correspond with the computers on the 192. net, but how?
> The ipchains REDIRECT only seems to work with certain ports on
> the localhost, not a full redirect to another IP.
If you are masquerading your entire LAN behind the firewall (many-to-one
NAT), then you won't be able to reach a particular computer in the LAN
from the Internet. You can redirect certain ports to certain machines on
your LAN (like HTTP, SMTP, FTP, etc.), but that's it.
If you use one-to-one NAT, then you could get to a particular machine
from the Internet by using the real IP address. This assumes that you
have enough IP addresses for all of your internal machines.
> The actual real-world problem is logging on with Win98 clients that
> come in off the internet, to a protected NT domain that sits
> behind a linux (RH 6.0) router/firewall. The Win98 clients can have
> static IP's.
Either way, you still won't be able to log on with your Win98 machine.
Netbios over TCP/IP really does not tolerate NAT.
What you can do instead is this:
1) Install RAS on your NT server.
2) Install PPTP on your NT server.
3) Install Dial-Up-Networking (DUN) on your Win98 client.
4) Install PPTP on your Win98 client.
The Win98 client connects to the Internet in whatever way it wants. The
it "dials" your RAS/PPTP box by using an ip address or name, like
pptp.yourdomain.com.
This works (we use it at my workplace) and it establishes a VPN between
the remote machine and your LAN!! Cool, eh? (Yes, I'm Canadian.)
Note that PPTP is not very secure from a cryptographic point of view. At
work, we're switching to an IPSEC-compliant VPN solution. But we did
start with PPTP.
Be sure to check out John Hardin's web page:
ftp://ftp.rubyriver.com/pub/jhardin/masquerade/ip_masq_vpn.html
Hope this helps,
Luca
--
Luca Filipozzi <[EMAIL PROTECTED]>
------------------------------
From: Chris Moseng <[EMAIL PROTECTED]>
Subject: RASDL
Date: Wed, 05 May 1999 11:56:29 -0500
This has only peripherally to do with Linux, but I thought someone wise
in the ways of ASDL might have an idea:
I'm using a linux machine to masquerade my home network with one IP
authentic IP address assigned to me by my ASDL ISP
I got the ASDL router installed and set the appropriate routing on this
linux box, and tried to use the connection before embarking on
recompiling for masqerading. What resulted was a very slow, lossy
connection that made it seem like I was connected over a 2400 baud modem
through a tin can and string. 40% packet loss, intermittant tcp
connections, 36 bps FTP transfers, what have you.
Needless to say, I was puzzled. Here's the punchline, though: my service
is not due to be "installed" until the 6th. So this connection, spotty
as it was, is actually far better than the "none" I should expect to
have. Naturally, I'm assuming the quality of the connection will rise
dramatically when the magic date passes.
How is it that my connection worked, sort of? In my limited experience,
things like network communications either work or they don't. Is this a
unique feature to ASDL over POTS?
One last note: my ISP has been ready for my connection since March, we
were both waiting for the phone co. and this magic date.
Curious,
Chris Moseng
[EMAIL PROTECTED]
------------------------------
From: Hamka Hj Suleiman <[EMAIL PROTECTED]>
Subject: Re: Linuxconf http access error??
Date: Thu, 06 May 1999 00:35:05 +0800
Les Hazelton wrote:
> I am attempting to get the web access to linuxconf working. However,
> each time I try to access "http://localhost:98/" I get an error page.
> The error message says
>
> "500 access denied: Check networking/linuxconf network access"
did u put in the ip address of the machine ur using to web access the
linux server in the linux conf setup? u need to tell linux conf which ip
# is allowed to access tru the web!
------------------------------
From: James Lee <[EMAIL PROTECTED]>
Subject: Re: PPP problems using Redhat 5.2 Usernet
Date: 5 May 1999 15:55:14 -0500
Thomas Lambert <[EMAIL PROTECTED]> wrote:
: I'm using redhat 5.2
: Usernet version 1.0.8-1 to start/stop my ppp connection.
: I get connected fine and can surf the net.
: But, once connected I can no longer start applications.
: Everything is locked up!!!
: I keep getting the following error:
: Xlib: connection to ":0.0" refused by server
: Xlib: Client is not suthorized to connect to Server
: netscape: can't open display :0.0
Is everything run under user or root or a combination of both?
------------------------------
From: "Shawn Conrad" <[EMAIL PROTECTED]>
Subject: THTTPD
Date: Wed, 5 May 1999 13:53:23 -0400
I have to write a paper comparing THTTPD with Microsoft's Internet
Information Server 4.0. The problem is that info on THTTPD is very limited.
Most of the info I have found is is Swedish.
------------------------------
** 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
******************************
