Linux-Networking Digest #419, Volume #12 Mon, 30 Aug 99 23:13:31 EDT
Contents:
Re: linux box vs switched hub ("Sagolsem C")
Re: Cable Modem Performance Probelms ("Robert Lowry")
Does anyone have a "3com/US Robotics Cable Modem CMX" working with (David Brandon)
State of NS16C650N support in the 2.0.3x kernel (Richard Ward)
PPP disconnect question (Matt Schmill)
Re: Vmware and Linux (Mircea)
ISP disconnects after about 30 secs.. (NuB2Linux)
Re: Online hardware reference? (Nick Rout)
eth0 and modules (Brian)
Re: Samba help required ([EMAIL PROTECTED])
Re: Can't mount Windows drives ("Hiawatha Bray")
pppd 2.3.9 won't compile ("dps")
Re: CHAP for Worldnet (Clifford Kite)
Re: VMware - wow! (Christopher Browne)
----------------------------------------------------------------------------
From: "Sagolsem C" <[EMAIL PROTECTED]>
Subject: Re: linux box vs switched hub
Date: Mon, 30 Aug 1999 09:40:09 -0400
What Switching hub are you guys talking about ? I thought a switching hub is
slightly more inteligent than the normal hub for the fact that it can
maintain a table of the IP addresses of the host connected on its port.
Therefore it can directly switch an incoming port trafic to the specific
destination port rather than broadcasting the packets to all ports like a
normal hub does. Moreovet when a communication takes place between two ports
other free ports also can also communicate with other free ports
simeltaneously thus increasing throughput.
I also think a 10/100 switching hub can switch packets between 10baseT and
100baseT hosts.
Sagolsem
Graham Fountain <[EMAIL PROTECTED]> wrote in message
news:4TDx3.73$[EMAIL PROTECTED]...
> As one other poster mentioned, a switch & a linux router/firewall are
> completely different, however there is one function of a switch that can
be
> emulated with a linux computer.
>
> One thing that is achieved with switches is the ability to allow 10MBit
and
> 100Mbit clients on the same network, that can communicate with each other.
> With a normal 10/100 Hub the 100 clients can see each other, and the 10
> clients can see each other, but the 10's can't see the 100's & vice versa.
>
> A Linux computer can be set up with both 10 and 100 MBit cards in it to
> provide the link between your 10MBit and 100MBit networks. This allows
you
> to save on the expense of buying a big switch, allowing you to use the
much
> cheaper hubs. This still won't give you as much bandwith as a hub, and it
> will impact on the performance of the linux computer, although I can't
give
> you any figures of how much it will impact it. I use a P133 and it
doesn't
> seem to hurt it too much.
>
> You can set up the linux box two ways - either as a router or a bridge.
> There are some subtle differences between each.
>
> A router will only route TCP/IP packets over the network. You set up your
> 10MBit clients on one subnet (say 192.168.1.0/24) and your 100MBit clients
> on another subnet (say 192.168.2.0/24). Given this, you would then set
the
> net cards up in the linux box as 192.168.1.1 for the 10MBit and
192.168.2.1
> for the 100Mbit. You would then use ipchains to get it to route data to
and
> from the 192.168.1.0 and 192.168.2.0 networks. Each workstation will need
> to be told that the relevant ip address in the linux box is its "gateway"
> address. If you want Windows Networking browsing available to you, you
will
> also need to set up a "WINS" server on your network. Routing also has the
> advantage that you can have some level of control over the packets. For
> example you can stop certain types of packets traversing the network.
>
> A bridge operates a little differently, in that it makes the two different
> network topologies to appear as one physical network. Instead of looking
at
> the destination IP addresses, the linux computer looks at the Ethernet
> address of each packet and routes it accordingly. All of your computers,
> both 10's and 100's would be on the same subnet, because as far as each is
> concerned, it thinks it is communicating direct to the other. One
advantage
> of a bridge is that it can also route Netbeui, IPX/SPX, and any other
> ethernet protocols. In this respect it functions almost the same as the
> ability of a switch to link 10 and 100 networks. I'm not 100% sure, but I
> think a bridge configuration would require more CPU power on the Linux
> machine than a router configuration.
>
> Another circumstance where you may want to use a switch or a linux router
is
> if you have a few departments (eg CAD, accounting, typing). Each of these
> departments shifts a lot of data within itself, and doesn't shift a lot of
> data to other departments (the assumption is that each department has its
> own server). All departments also need a low level of shared access to
some
> other common resources. In this circumstance if you had everyone
connected
> to a normal hub, the 10 or 100MBit bandwith would be shared between
> everyone, so someone saving a big CAD file to a server, would hog bandwith
> from every user. The better way would be to set this type of department
up
> as follows:
>
> TYPING
> | | |
> HUB
> |
> Common Resources-----SWITCH -------HUB====Accounting
> |
> HUB
> | | |
> CAD
> I hope the ASCII "art" lines up, but if not, all hubs are linked to the
> switch. If you used a Linux Router or Bridge instead of the switch in
this
> configuration, you would set it up with 4 network cards. In this type of
> enterprise, if you wanted to use only one server (either linux or NT) you
> would put it in the central position with four network cards, and set up
> with routing.
>
> I'm starting to get a bit long winded - there are a zillion different
> configurations and many aspects to be taken into account when determining
> the best solution. Hope this helps you.
>
>
> chas <[EMAIL PROTECTED]> wrote in message
news:7q6ebg$pcn$[EMAIL PROTECTED]...
> >
> >
> > Preparing for a discussion with our network administrator...
> >
> > What are the advantages (if any) of a switched hub over a linux
> > router/firewall?
> >
> > --
> > ===============================
> > Charles Lewis ([EMAIL PROTECTED])
> > Director of IS (SWAU)
> > ===============================
> >
> >
> > Sent via Deja.com http://www.deja.com/
> > Share what you know. Learn what you don't.
>
>
------------------------------
From: "Robert Lowry" <[EMAIL PROTECTED]>
Subject: Re: Cable Modem Performance Probelms
Date: Tue, 31 Aug 1999 01:06:39 GMT
I'll try adjusting the IP Chain timeout. Thanks for the Idea. I did force my
Ethernet card to half duplex and
that helped some. I get further on complex pages. The performance through
the lunix box is still lacking though.
RWL
Robert_Glover <Please_reply_to@newsgroup> wrote in message
news:7qeaj5$[EMAIL PROTECTED]...
> I had another thought. Just try setting your masquerading timeouts
> extremely low, and see if that doesn't reliably produce the symptoms
> you are experiencing. If it does, then you're onto something.
> Something like this is certain to cause problems on the _masqueraded_
> PC's:
>
> ipchains -M 10 1 1
>
> Note: I've seen -M 1800 10 60 suggested as good, normal values.
>
> If you don't use iphcains, I'm pretty sure that ipfwadm (or
> ipmasqadm?) will also allow you to set these values too.
>
> Robert_Glover wrote in message <7qdu54$[EMAIL PROTECTED]>...
> >That makes me suspect your masquerading timeouts. I recently set
> mine
> >way too low and experienced the same symptoms. I think mine are
> >now set to 900 10 60 (for ipchains).
> >
> >Robert Lowry wrote in message ...
> >>No DNS Lookup problems. from WIN95, I get a connection reset from
> >host
> >>message
> >>after a while. The first part of most pages come across. On complex
> >(large
> >>amounts of
> >> graphics etc), the transfer seems to die.
>
> [snip]
>
> >>> >Any help would be appreciated.
> >>> >
> >>> >[EMAIL PROTECTED]
> >>>
> >>
> >>
> >
>
------------------------------
From: David Brandon <[EMAIL PROTECTED]>
Subject: Does anyone have a "3com/US Robotics Cable Modem CMX" working with
Date: Mon, 30 Aug 1999 21:12:14 -0400
I've had two other cable modems, first with MediaOne, and then with
RCN. Now RCN has replaced their first cable modem (the brand was
"Hybrid") with a 3com/USR "Cable Modem CMX". The other two were no
trouble but for some reason I can NOT get this one to work with Linux
(Redhat 5.2). Since the setup worked fine with two previous modems I'm
pretty sure that routing, NICs, etc, are okay. Also, if I plug my iMac
into the modem, it works fine after I power cycle the modem.
I've tried everything I can think of:
- Power cycle the cable modem with the linux box up (so the modem will
see the MAC address of the linux box)
- try dhcpcd with "-h hostname"
- try dhcpcd with "-r" (desperation)
My ethernet card is an SMC EtherEZ. I also have another ethernet card,
and old DEC (I forget the model off of the top of my head). The other
modems were working with the dec card, but the link on that card goes
down when the driver isn't loaded, so I was worried that the cable modem
wasn't seeing the NIC when the cable modem booted. However, it doesn't
seem to matter which NIC I use.
Has anyone else managed to get one of these 3com modems working?
Thanks,
David
--
David [EMAIL PROTECTED]
------------------------------
From: Richard Ward <[EMAIL PROTECTED]>
Subject: State of NS16C650N support in the 2.0.3x kernel
Date: Mon, 30 Aug 1999 20:33:59 -0400
I am considering purchasing a Lava Multi-port Serial I/O card with 16650
chips on it and when checking the serial.c file to see if the kernel had
native support for this chipset, it appeared to in someplaces have this
chips FIFO's disabled, i.e. set to 1, and in other places it is set to
32, as it should be.
So, has anyone used the single port, ISA, Lava 16650 I/O card under
Linux? Is the 16650 chip supported at all.
Thanks
Richard Ward
------------------------------
Date: Mon, 30 Aug 1999 09:45:59 -0400
From: Matt Schmill <[EMAIL PROTECTED]>
Crossposted-To: 5col.comp.linux
Subject: PPP disconnect question
RedHat 6.0's handling of PPP has gotten better, but I have a question
regarding bringing the interface down.
When I click on the "toggle interface" button on Usernet, it breaks its
connection with the PPP server, but does not hang up the phone. Since I
dial-up to a terminal server, breaking with the PPP server does not
result in freeing up the modem. I recall hearing about a "lo" interface,
but I don't get that in Usernet. Is there a surefire way of hanging up
the PPP connection?
That failing, I use Usernet to "toggle" the connection, then login as
root and get in with minicom and hang up that way. Unfortunately,
sometimes the serial port gets locked, and there are
"ifup-ppp" processes floating around doing nothing. How do I unlock the
serial port and prevent ifup-ppp processes from wanding aimlessly?
thanks in adnance!
------------------------------
From: Mircea <[EMAIL PROTECTED]>
Subject: Re: Vmware and Linux
Date: Mon, 30 Aug 1999 21:12:31 -0400
Peter wrote:
>
> Hi
>
> If there is anyone using vmware could they answer a few of my
> questions, I have a dual boot win98/Linux machine and would like to
> know if it is possible to use the existing win98 partition in a vm or
> if I have to install another copy of Windows as the instructions seem
> to imply.
You can do both, although the first one is a little tricky.
>
> I have Linux running on a spare 1gig disc and there is no room to
> install another copy of Window, if it works out OK I will install
> Linux on a bigger drive and run Win98 in a vm but want to try it out
> first before reorganizing my discs.
>
> I have downloaded a 30 day trial of the program but don't see any
> obvious way of doing what I want from the help files.
You may want to lurk around the newsgroups in news.vmware.com - it's
vmware's public news server, and some of the guys that wrote it are
always around, you can ask them.
MST
------------------------------
From: NuB2Linux <[EMAIL PROTECTED]>
Crossposted-To: redhat.config,redhat.networking.general
Subject: ISP disconnects after about 30 secs..
Date: Tue, 31 Aug 1999 01:31:03 GMT
Hello,
I was wondering if there is a terminal window that can pop up after the
modem connects to see what exactly the computer is send to my ISP server?
It acts as though the name and password that it's suppling isn't correct
and the modem disconnects after about 30sec (it's and external and I can
see it sending info to the server). I'm currently using RH Linux 5.2 (RH6
should be here within the week). Also, there is a way to log your connect
responses and things for troubleshooting purposes. What is that command?
If you could reply via e-mail, it'd be much appreciated...
================== Posted via CNET Linux Help ==================
http://www.searchlinux.com
------------------------------
From: [EMAIL PROTECTED] (Nick Rout)
Subject: Re: Online hardware reference?
Date: Tue, 31 Aug 1999 13:25:23 +1200
Try www.pcguide.com
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] says...
> Hi group,
>
> Does anyone know of a computer hardware reference site? Specifically looking
> for information online dealing with scsi and ide drive compatibility issues.
>
> Thanks!
>
> Badmango
>
------------------------------
From: Brian <[EMAIL PROTECTED]>
Subject: eth0 and modules
Date: Mon, 30 Aug 1999 21:50:02 -0400
does any know how to get an ethernet card to work in RH 5.2?
--
Brian Savacool
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Samba help required
Date: Tue, 31 Aug 1999 01:25:19 GMT
In article <7qdvcg$7ih$[EMAIL PROTECTED]>,
"Paco" <[EMAIL PROTECTED]> wrote:
> Hi there
>
> I have a NT Domain controller, some 95 workstations and one Redhat
Linux 6.0
> box.
>
> I have Samba installed and all the win95 machines can access any share
of
> the linux box. I can however not connect from or to the NT PDC from
the
> Linux machine. network functionality is there and i can ping both ways
with
> no hassles.
>
> When from within NT i try to connect to the Linux box i get the
following
> error:
>
> This workstation does not have permission to access
>
> Any ideas
>
> Thanks
> JP
>
>
You may have EnablePlainTextPassword registry "off" on the NT.
Try setting
\HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Rdr\Parameters\
EnablePlainTextPassword to "1" (in hex).
It worked for me. :-)
Goodluck,
Jaran
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: "Hiawatha Bray" <[EMAIL PROTECTED]>
Crossposted-To: linux.samba
Subject: Re: Can't mount Windows drives
Date: 30 Aug 1999 19:35:10 PDT
So...I'm supposed to put 2 more slashes before mypc?
Cliff <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Hello,
>
> Try this..
>
> smbmount \\\\mypc\\c /mnt/win -N
>
> (make sure /mnt/win exists)
>
> Cliff
>
> Obs wrote:
>
> > you should do smbmount ////computer-name/sharename /mount-point
> >
> > Hiawatha Bray <[EMAIL PROTECTED]> wrote in message
> > news:7qcp4c$[EMAIL PROTECTED]...
> > > I got my Samba working and I can mount my Linux drives on my Windows
> > > machine. Now...how do I do it the other way around, mounting Windows
> > drives
> > > on my Linux box? I'm using RH 6 and Samba 2.0.3, withWin 98 on the
> > Windows
> > > machine.
> > >
> > > I created a /mnt/win mountpoint and then tried the smbmount command.
> > Here's
> > > what I got...
> > >
> > > [root@linux watha]# smbmount file://mypc/c /mnt/win
> > > Added interface ip=10.0.2.16 bcast=10.255.255.255 nmask=255.0.0.0
> > > Server time is Sun Aug 29 21:42:48 1999
> > > Timezone is UTC-4.0
> > > security=share
> > > smb: \>
> > >
> > > I don't even know what this means...is this the correct response?
Thanks.
> > >
> > >
>
>
------------------------------
From: "dps" <[EMAIL PROTECTED]>
Subject: pppd 2.3.9 won't compile
Date: Mon, 30 Aug 1999 22:49:57 -0400
Hello all ...
Can't get pppd 2.3.9 to compile on my slackware (kernel 2.2.12) box. I've
followed the readme directions but it still bombs out with the following:
============================================================
sys-linux.c: In function `ppp_available':
sys-linux.c:1640: warning: implicit declaration of function `makedev'
cc -O2 -pipe -Wall -g -D_linux_=1 -DHAVE_PATHS_H -DIPX_CHANGE -I../include -
DCHAPMS=1 -DUSE_CRYPT=1 -DHAS_SHADOW -c ipxcp.c -o ipxcp.o
cc -O2 -pipe -Wall -g -D_linux_=1 -DHAVE_PATHS_H -DIPX_CHANGE -I../include -
DCHAPMS=1 -DUSE_CRYPT=1 -DHAS_SHADOW -c md4.c -o md4.o
cc -O2 -pipe -Wall -g -D_linux_=1 -DHAVE_PATHS_H -DIPX_CHANGE -I../include -
DCHAPMS=1 -DUSE_CRYPT=1 -DHAS_SHADOW -c chap_ms.c -o chap_ms.o
cc -O2 -pipe -Wall -g -D_linux_=1 -DHAVE_PATHS_H -DIPX_CHANGE -I../include -
DCHAPMS=1 -DUSE_CRYPT=1 -DHAS_SHADOW -o pppd main.o magic.o fsm.o lcp.o
ipcp.o upap.o chap.o md5.o ccp.o auth.o options.o demand.o utils.o
sys-linux.o ipxcp.o md4.o chap_ms.o
sys-linux.o: In function `ppp_available':
/usr/include/sys/stat.h:169: undefined reference to `makedev'
make[1]: *** [pppd] Error 1
make[1]: Leaving directory `/usr/src/ppp-2.3.9/pppd'
make: *** [all] Error 2
============================================================
I've just upgraded to gcc ver 2.7.2.3, enabled PPP in my new kernel rebuild,
and I am at a standstill here.
Can anyone help?
Thanks for any and all assistance,
peter steele
------------------------------
From: kite@NoSpam.%inetport.com (Clifford Kite)
Subject: Re: CHAP for Worldnet
Date: 30 Aug 1999 21:20:23 -0500
Jeffrey Kiefer ([EMAIL PROTECTED]) wrote:
: OpenLinux 2.2 KDE. I have finally configured my modem. How do I set CHAP? I
: have the chap-secrets file - now what? I'm such a newbie, and I need help.
http://axion.physics.ubc.ca/ppp-linux.html
--
Clifford Kite <kite@inet%port.com> Not a guru. (tm)
------------------------------
From: [EMAIL PROTECTED] (Christopher Browne)
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup
Subject: Re: VMware - wow!
Reply-To: [EMAIL PROTECTED]
Date: Tue, 31 Aug 1999 02:18:11 GMT
On Sat, 28 Aug 1999 23:28:21 +0100, Spike! <[EMAIL PROTECTED]> wrote:
>Christopher Browne <[EMAIL PROTECTED]> wrote:
>> The VM approach goes in the opposite direction; instead of having
>> fine-grained threading where you share as much as possible, you move
>> to having completely separate virtual machines, where in order to get
>> data between VMs, some Special Code needs to be invoked.
>
>Have you ever heard of uQLx? It's the linux QL emulator, and one of the
>recent additions to that is the ability to make it fork into separate copies
>of itself so you can have multiple QLs working on the same machine.
>
>I think the inter-process communication between these forks is done rather
>cleverly with the QL network emulation... This means all the virtual QLs can
>talk to each other using the standard SuperBASIC interpreter...
>
>(could be wrong though)
>
>Just thought I'd mention it because it seems to be a similar thing to what
>you were talking about.
While I was generally aware of the existance of a QL emulator, I was
not familiar with that; it sure sounds a lot like what I was
describing.
The killer question, of course, is whether or not people are writing
code for Virtual QL instances running atop Linux... :-)
--
OS/2: Why marketing matters more than technology...
[EMAIL PROTECTED] <http://www.hex.net/~cbbrowne/lsf.html>
------------------------------
** 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
******************************
