Linux-Networking Digest #58, Volume #11 Thu, 6 May 99 07:13:28 EDT
Contents:
Re: NT faster than Linux? (Edward_hill)
Re: no pingies in very simple network ([EMAIL PROTECTED])
Re: restarting network services ("Jochen Chauchet")
2 modems 2 ISP accts. is it possible? (Jack)
Compaq Microcom 550 Pcmcia (kc)
LinuxPing but no gateway (Matt)
Re: connecting two networks w/o a router?? ([EMAIL PROTECTED])
routing through ppp ??? ("hEllRaiSer")
Strange problem while connected to foreing host (Joerg Sauer)
ethernet card not detected. ioport occupied? (Winrider)
Re: "no daemon present"? ... lpc> status (Martin Schulz)
Re: Crontab woes with RH5.2 ("Leopold Toetsch")
Re: ethernet card not detected. ioport occupied? (Winrider)
Re: LinuxPing but no gateway (Paul Black)
Re: connecting two networks w/o a router?? (Luca Filipozzi)
Help on Samba network ... ("Flex Line")
----------------------------------------------------------------------------
From: Edward_hill <[EMAIL PROTECTED]>
Crossposted-To: uk.comp.os.linux
Subject: Re: NT faster than Linux?
Date: Thu, 06 May 1999 09:18:17 +0100
[EMAIL PROTECTED] wrote:
:
: In the sacred domain of uk.comp.os.linux didst Richard Corfield
:[EMAIL PROTECTED]> eloquently scribe:
: : In article <[EMAIL PROTECTED]>,
: : Anthony W. Youngman <[EMAIL PROTECTED]> wrote:
: :>
: :>Rumour? says that M$-basic was basically DEC-basic or somesuch that
he
: :>came across while on work experience at college ...
:
: : Way back then BBC Basic was the thing - especialy Version 5 with
what were
: : (for BASIC) advanced features such as switch and while and
procedures
: : with local variables and hence recursion. It was hardly OO but not
bad.
:
: Nahhh... M$ BASIC was crap long before the BBC or spectrum showed
people
: what could be...
:
: (And yes, I know, at the time the BBC model A and B came out, it was
the
: best BASIC out there... The speccy was pretty good too though)
Sold my old BBC B to a school quite a few years ago I still miss it
sometimes
(sniff), The BBC Micro was in a class of it's own if you'll excuse the
pun.
Actually I had a double disk drive that switched between 40 and 80 track
the equivalent of having a .. erm er no can't think of one
well the equivalent of having something really really swish nowadays
:->.
Elite on the BBC B is still the best version ever made...
Ed
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: no pingies in very simple network
Date: Thu, 06 May 1999 09:17:57 GMT
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (Luca Filipozzi) wrote:
> In article <7gremp$6r4$[EMAIL PROTECTED]>, [EMAIL PROTECTED] says...
> >
> >
> > I'm having trouble with what must be a bloody trivial problem. I've got two
> > machines, one desktop, one laptop, a hub, two ehternet cards, and can not
get
> > the two machine talking to each other in Linux.
> >
SNIPPED
> FIRST (ensure kernel likes Ethernet card)
>
> Your routing table and ifconfig statements look good. Try rebooting and
> using dmesg to see if the kernel complained about your Ethernet card at
> all.
>
> NEXT (attempt to isolate problem by packet sniffing)
>
> 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.
>
> Do the same for the other machine. At least this way we can hopefully
> isolate the problem.
>
> NEXT (ensure that the firewall rules don't block icmp packets)
>
> Check to ensure that you don't have firewall filters blocking traffic.
> You can use the following to list the rules
> "ipfwadm -I -l -n" to check the Incoming rules
> "ipfwadm -O -l -n" to check the Outgoing rules
> "ipfwadm -F -l -n" to check the Forwarding rules
> I don't know the syntax for ipchains, sorry.
>
> You can use the following to flush the rules and accept by default
> "ipfwadm -I -f -p accept"
> "ipfwadm -O -f -p accept"
> "ipfwadm -F -f -p accept"
>
> NEXT (post again if the above doesn't help)
>
> Luca
>
> --
> Luca Filipozzi <[EMAIL PROTECTED]>
>
Thanks, I'll try your first suggestion first. I now realize I should have
mentioned that I have a 3com509b card in one of the machines (the one which
has Windows and Linux.) In my naivety I thought that the ifconfig info was
evidence that the card was working properly but I take it that ain't
necessarily so (even if it works in Windows.)
I shall return.
Dennis
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: "Jochen Chauchet" <[EMAIL PROTECTED]>
Subject: Re: restarting network services
Date: Thu, 6 May 1999 10:48:02 +0200
killall -hup smbd
killall -hup nmbd
Have a lot of fun...
Jochen
Chris Snyder schrieb in Nachricht <[EMAIL PROTECTED]>...
>Perhaps this question is a bit of a newbie question but I will ask any
>way.
>If one changes configurations for Samba (or any other networking
>service) how can one restart the services with the new configuration
>without rebooting?
>Thanks in advance,
>-chris snyder
>
------------------------------
From: Jack <[EMAIL PROTECTED]>
Subject: 2 modems 2 ISP accts. is it possible?
Date: 6 May 1999 09:32:33 GMT
anybody tried modem bonding in Linux? where you use 2 modems and 2 ISP
accts in one computer.
is it possible now in Linux?
================== Posted via SearchLinux ==================
http://www.searchlinux.com
------------------------------
From: kc <[EMAIL PROTECTED]>
Subject: Compaq Microcom 550 Pcmcia
Date: 6 May 1999 09:32:36 GMT
Hi there,
Has anybody out there ever tried to install a microcom 550 pcmcia card from
compaq ? This card looks like a Xircom card but if you try to load this
module manualy (because the defined cards from the pcmcia package doesn't
recognise it) I don't receive an "eth0" from it.
Anybody has experience with this partical card ??
Help me out please !!
Greetinx KC
================== Posted via SearchLinux ==================
http://www.searchlinux.com
------------------------------
From: Matt <[EMAIL PROTECTED]>
Subject: LinuxPing but no gateway
Date: Thu, 06 May 1999 10:43:29 +0100
Hi,
I managed to get Linux to ping itself ok but I am unable to ping
the other box ie.. NT (NT box can ping itself too).
I have given the same gatway addresses to both machines
but still no connection.
Has anyone got any advice
Many thanks
Matt
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.protocols.tcp-ip
Subject: Re: connecting two networks w/o a router??
Date: Thu, 06 May 1999 09:24:00 GMT
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (Luca Filipozzi) wrote:
> In article <7grfo7$7mr$[EMAIL PROTECTED]>, [EMAIL PROTECTED]
> says...
> > Yes. I don't think you can do that on the client under NT, though. Does
> > anyone know if NT supports static routes? Something in 'lmhosts', maybe?
> Yes, Bill, you can add static routes from a command prompt.
Really? Shit. I wish I'd known that two years ago [which was, fortunately
for me, the last time I've had to admin an NT box].
> The two DNS server is my preference. Allows for very clean separation
> between "internal address space" and "external address space".
Same here. It doesn't sound like a second box for internal DNS is a viable
option in this case, though. Is there a way to do this on one box, by
running a second copy of `bind` on a different port, and configuring the
clients to use that DNS? I don't have the equiptment here to test it out.
> > The simplest solution is to use /etc/hosts, since this is the main reason
> > it's still around and hasn't been completely replaced by DNS.
> True enough. Maintaining an /etc/hosts file on every machine becomes a
> pain in anything but small networks.
Not necessarily. It's rather straight-forward to write a series of scripts
to update the local hosts file via `scp` or `rsync` or whatnot from a copy on
a master server [the DNS server would be a good choice]. On Windows it might
be a bit trickier, depending on the software available, but it's still very
feasible.
> > Unfortunately, that
> > won't work on any windows box, for obvious reasons.
> Sure it will.
I meant having an /etc/hosts. Windows doesn't have an /etc directory. Yeah,
I know, that's a rather silly point, but when I say "obvious", I really mean
it. :)
> \winnt\system32\drivers\etc\hosts
Didn't know about that one. In light of the pathname, I see that my prior
"obvious" point wasn't so obvious after all.
So, to sum up, there are now three proposed solutions to the original poster's
problem:
1) Use \winnt\system32\drivers\etc\hosts on client Windows machines,
/etc/hosts on Unix boxes, and Macs I think are screwed [except OS X]. Store
the internal name translations in that file, using the 192.168/16 IP address
of the Linux box. This solution is only really viable if the network is
relatively small or if an automated updating system can be put in place. [But
it's the easiest]
2) Set up a second server to be used for internal DNS.
2a) If it is possible, run two copies of `bind` on the same machine.
3) Add static routes mapping the external IP of the Linux box to the internal
IP, and rely solely on external DNS.
That sound about right?
-Bill Clark
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: "hEllRaiSer" <[EMAIL PROTECTED]>
Subject: routing through ppp ???
Date: Thu, 6 May 1999 09:49:25 +0200
hi,
only one question, how can I route through ppp.
I have 3 hosts, 1 host (ip:xxx.xxx.xxx.xxx) is connected over ppp to the
internet, if i add this at the 2 other hosts all pings from this 2 hosts to
the i-net are lost:
route add default gw xxx.xxx.xxx.xxx
also at the 1 host the ppp connection is the default gw, but that doesnt
work, why?
thanx for reply
sEEk
------------------------------
From: [EMAIL PROTECTED] (Joerg Sauer)
Crossposted-To: de.alt.comm.isdn4linux
Subject: Strange problem while connected to foreing host
Date: Thu, 06 May 1999 08:50:24 GMT
Hi,
I recently posted a question about my problem, but didn't get any
working answers and also couldn't find out the solution by myself.
Problem:
After logging into an remote host with telnet the connection is
immidiatly closed after invoking applications like vi, more, less, vnc
I just get the message "connection closed by foreign host"
This happens allways when I connect throug my remote access dial in
port isdn0 (as described below) when I connect through the Internet
this don't happen and everything works fine.
Here is my network map.
Each Linux Box has 3 Network connections.
1. internal net trough eth0
2. internet connection through ippp0 (PPP), with DHCP and Masquerading
3. direct dial in throug isdn0 (RawIP)
=============My Site=======================
Network 192.168.1.0 Mask:255.255.255.0
Win95 PC
| 192.168.1.2
|
eth0
|
| 192.168.1.1
Linux
| |
| |
| |_______ippp0 (PPP) -->Internet Provider (DHCP, Masquerading)
|
|___isdn0(Raw IP)--PtP-|
192.168.3.1 |
|
|
=======Remote Site======+======= Network 172.16.1.1 Mask 255.255.255.0
_________________|
|
isdn0 (192.168.3.2) Raw IP
|
Linux
| |
| |
| |___ippp0 -->Internet Provider (DHCP,Masquerading)
|
eth0 172.16.1.1
|
|
Win 95 PC 172.16.1.2
I am using Suse 6.0 Linux on both sides, The installations of the
linux boxes are allmost the same: (Isdn4Linux, Bind, Sendmail,
Hardware...) Just the network addresses and nameserver tables are
different for each side.
In the Suse rc.config file I made the following entries:
MSQ_START=Yes
MSQ_NETWORKS=192.168.1.0/24
MSQ_DEV=ippp0
FW_START=no
IP_DYNIP=yes
Do you have any ideas?
Thanks in advance
Joerg
PS: response also per email appretiated.
------------------------------
From: Winrider <[EMAIL PROTECTED]>
Subject: ethernet card not detected. ioport occupied?
Date: Thu, 06 May 1999 02:28:37 +0100
Reply-To: "[EMAIL PROTECTED];Winrider"@ix.netcom.com
Hi all
During startup, dmesg did not report my ethernet card (LinkSys
EtherPCI LAN II) and when I check /proc/ioport
I see
...
3000-3007 IDE DMA
3008-300f IDE DMA
<end>
I read the doc saying that we should keep those port clear.
How do I get rid of those IDE DMA?
Arnold
I
------------------------------
From: Martin Schulz <[EMAIL PROTECTED]>
Crossposted-To:
comp.protocols.smb,linux.samba,linux.redhat.misc,comp.os.linux.misc,comp.os.linux.hardware,alt.linux,alt.os.linux
Subject: Re: "no daemon present"? ... lpc> status
Date: 06 May 1999 12:06:09 +0200
"Bleh" <[EMAIL PROTECTED]> writes:
>
> Why does it says "no daemon present" for every single printer device? How
> do I start the "raw" device as a daemon?
As I understand it, this is okay as long as there are no entries.
HTH,
Martin.
--
Martin Schulz [EMAIL PROTECTED]
Uni Karlsruhe, Institut f. wissenschaftliches Rechnen u. math. Modellbildung
Engesser Str. 6, Karlsruhe
------------------------------
From: "Leopold Toetsch" <[EMAIL PROTECTED]>
Subject: Re: Crontab woes with RH5.2
Date: Thu, 6 May 1999 09:28:40 +0200
Hi,
[EMAIL PROTECTED] wrote in message <7gq2e1$vqj$[EMAIL PROTECTED]>...
>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
Export your favorite editor:
export EDITOR=mcedit
or
export EDITOR=joe
or whatever you like.
(The above is for bash)
and
man crontab tells you about this.
HTH
leo
------------------------------
From: Winrider <[EMAIL PROTECTED]>
Subject: Re: ethernet card not detected. ioport occupied?
Date: Thu, 06 May 1999 02:51:26 +0100
Reply-To: "[EMAIL PROTECTED];Winrider"@ix.netcom.com
Winrider wrote:
> Hi all
>
> During startup, dmesg did not report my ethernet card (LinkSys
> EtherPCI LAN II) and when I check /proc/ioport
>
> I see
>
> ...
> 3000-3007 IDE DMA
> 3008-300f IDE DMA
> <end>
>
> I read the doc saying that we should keep those port clear.
> How do I get rid of those IDE DMA?
>
> Arnold
> I
Some more info: in the dmesg, i see IDE device starting up
ide :i82371 PIIX (triton) on PCI bus 0 function 57
ide0:BM-DMA at 0300-0307
ide1:DM-DMA at 0308-030f
------------------------------
From: Paul Black <[EMAIL PROTECTED]>
Subject: Re: LinuxPing but no gateway
Date: Thu, 06 May 1999 11:20:02 +0100
Matt <[EMAIL PROTECTED]> wrote:
> I managed to get Linux to ping itself ok but I am unable to ping
> the other box ie.. NT (NT box can ping itself too).
>
> I have given the same gatway addresses to both machines
> but still no connection.
>
> Has anyone got any advice
Could do with more information: e.g. routing, ethernet cards etc.
What is the output of ifconfig, netstat -r?
Paul
------------------------------
From: [EMAIL PROTECTED] (Luca Filipozzi)
Crossposted-To: comp.protocols.tcp-ip
Subject: Re: connecting two networks w/o a router??
Date: Thu, 6 May 1999 03:23:51 -0700
In article <7grn3f$eej$[EMAIL PROTECTED]>, [EMAIL PROTECTED]
says...
> > The two DNS server is my preference. Allows for very clean separation
> > between "internal address space" and "external address space".
>
> Same here. It doesn't sound like a second box for internal DNS is a viable
> option in this case, though. Is there a way to do this on one box, by
> running a second copy of `bind` on a different port, and configuring the
> clients to use that DNS? I don't have the equiptment here to test it out.
I tried setting up two bind instances but gave up. Let me tell you why.
Bind 8.1.2 has an option called listen-on that would allow me to force a
particular instance to listen on one interface only. Therefore, having
two bind instances each "listening on" a different interface is possible.
The linux box itself probably should use the localhost interface, so one
of the two bind instances should listen on that interface, too.
The problem that made me give up ('cause I'm lazy) had to do with start-
stop-daemon. I couldn't get it to start two instances of the same
program. It's just a shell script problem that I didn't bother spending
time to rectify. An easy fix could be to copy bind to bind2 and be done
with it. But I'm very happy with the Debian packages, in general, and I
didn't want to screw around with little hacks like that.
> > > The simplest solution is to use /etc/hosts, since this is the main reason
> > > it's still around and hasn't been completely replaced by DNS.
>
> > True enough. Maintaining an /etc/hosts file on every machine becomes a
> > pain in anything but small networks.
>
> Not necessarily. It's rather straight-forward to write a series of scripts
> to update the local hosts file via `scp` or `rsync` or whatnot from a copy on
> a master server [the DNS server would be a good choice]. On Windows it might
> be a bit trickier, depending on the software available, but it's still very
> feasible.
could do, could do
> So, to sum up, there are now three proposed solutions to the original poster's
> problem:
>
> 1) Use \winnt\system32\drivers\etc\hosts on client Windows machines,
> /etc/hosts on Unix boxes, and Macs I think are screwed [except OS X]. Store
> the internal name translations in that file, using the 192.168/16 IP address
> of the Linux box. This solution is only really viable if the network is
> relatively small or if an automated updating system can be put in place. [But
> it's the easiest]
>
> 2) Set up a second server to be used for internal DNS.
> 2a) If it is possible, run two copies of `bind` on the same machine.
With shell script massaging, yes.
>
> 3) Add static routes mapping the external IP of the Linux box to the internal
> IP, and rely solely on external DNS.
>
> That sound about right?
Sounds good to me.
--
Luca Filipozzi <[EMAIL PROTECTED]>
------------------------------
From: "Flex Line" <[EMAIL PROTECTED]>
Subject: Help on Samba network ...
Date: Wed, 5 May 1999 10:50:43 +0200
We've a Linux as Domain Server and 4 Windows 98 client in 10Mbit Ethernet
connection with Samba 1.9 client and we've this problem: close connection
sometimes!
In LogFile:
...
1999/05/04 13:49:21 PIIPC (192.168.0.3) closed connection to service IPC$
Yielding connection to 118 IPC$
...
then Linux closes all open files and then reopen connection !?!
So, we have big trouble on execute dataflex reindex (in DOS session) in a
stable way!!
Copying file over Windows 98 work slow, ... but work.
Can You HELP us ?
Thank You in Advance.
------------------------------
** 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
******************************
