Linux-Networking Digest #198, Volume #10 Sun, 14 Feb 99 00:13:38 EST
Contents:
Compaq Deskpro 6000 and PCI BIOS ([EMAIL PROTECTED])
Re: Intranet Server ("William R. Mattil")
Re: SuSE Linux 6.0 & Routing (grumpy)
Re: Printing from Mac not possible (Rod Smith)
Re: IP masquerading ("tonni")
Re: Compaq Deskpro 6000 and PCI BIOS (Kevin Dick)
Re: NIC Works Part Time ?? ("jim")
Re: ip addresses ("tonni")
Re: IP Masquerading - Dropping Packets ("tonni")
Re: a *problem* with chatscript (Clifford Kite)
Re: Why is winsock ppp more robust that pppd? (Clifford Kite)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED]
Subject: Compaq Deskpro 6000 and PCI BIOS
Date: Fri, 12 Feb 1999 21:00:51 GMT
Hello all, OK, I think I've read all the readme's and Howto's before
coming here so I am desperate for any help. I need to get my onboard
NetFlex-3 PCI Ethernet card working in Linux. Everything else works great
with this RedHat 5.2 installation. I have a Compaq DeskPro 6000 with a
Pentium Pro 200, Adaptec 2940U PCI SCSI Host adapter with a 2 2GB seagate
hard drives, an IDE CD-ROM, 96MB RAM and a RedHat 5.2 Install CD. Currently I
have my 1st HD setup with 2 Partitions, NTFS for the c: drive and FAT for the
d: drive, both approximately 1GB each. On the second drive, I have a 24MB DOS
partition, a 47MB Swap partiton (sdb2) and a 1947MB Linux root partiton
(sdb1). Neither Linux, nor the Redhat Installer will recognize the
onboard PCI ethernet card. It is a NetFlex-3 PCI card. From what I've read,
this is because the PCI bios on this model is stored in Hi-MEM and is
inaccessible by Linux. Fair enough, I can find the Compaq movepci.exe file,
load it on a DOS bootable floopy and go. Problem is, hoe to get Linux
to boot from here so that I can take advantage of this change? Using
loadlin, I have gotten the error "Idle task could not sleep" which seems to
have a resolution and is posted all over the web in German. All I THINK I
need to do now is boot into Linux from this poitn and do an insmod and
ifconfig to get the card going. Is there an easier way to do all this? Am I
on the right track? Th kernel I'm using is copied right off the Linux root
partition and boots the machine fine from the NT boot.ini file. Any help
anyone could give me would be very appreciated. Thanks in advance. --Dave
Reuss [EMAIL PROTECTED] [EMAIL PROTECTED]
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: "William R. Mattil" <[EMAIL PROTECTED]>
Subject: Re: Intranet Server
Date: Fri, 12 Feb 1999 15:32:50 -0600
Duncan Simpson wrote:
>
> In <794dpr$n0b$[EMAIL PROTECTED]> "kevin" <[EMAIL PROTECTED]> writes:
>
> >Currently I am planning to setup an Intranet Server for my office... however
> >due to budget concern I have to use one of my office PC . The setup of this
> >is as such... 133 Pentium, 1.2G hardisk, 16M ram. Could anybody tell if this
> >is sufficient for an Intranet Server running Linux & Apache? Later I might
> >need to add an hard disk to make it run as SQL server as well. Any
> >recommendation?
>
> It should be enough if you have a budget probelm requiing you to use
> this hardware. If oyou are thinking of a server handling 100 hits per
> second then you probably want something fancier. SQL databases fall
> into two camps: almost SQL and cheap (mSQL, MySQL, et al) and the full
> whack and expensive (DB2, Informix, Sybase, Yard, etc).
>
This would not be very appropriate IMHO. 16 Mmeg RAM is nowhere near
enough for a production server unless you get less than 10 hits per
hour. With Memory being so cheap these days there really isn't much
reson for less than 64 Meg RAM. And even then I would use 128 Meg swap
... just in case.
Regards
Bill
--
William R. Mattil | You're one of those condescending Unix guru's,
SSCFI System Admin | Here's a nickel kid. Get yourself a better
(972) 399-4106 | computer. - Dilbert
------------------------------
From: grumpy <[EMAIL PROTECTED]>
Subject: Re: SuSE Linux 6.0 & Routing
Date: Fri, 12 Feb 1999 20:24:25 GMT
Assuming both cards are recongnized and working ,
enable ip forwarding, run a routing daemon (routed [RIP only] or gated
[RIP/OSPF/etc.])
check the man pages for the configuration file format.
Volker Kalthaus wrote:
>
> Hello all,
>
> i installed Linux (SuSE Linux 6.0) for the first time. the box consists of 2
> NICs. Both NICs have IP Adrs in different subnets of our class-c network.
>
> i want to setup routing now, so i can ping client in the one subnet from the
> other subnet, using the linux box as default router.
>
> i guess im really to stupid to set up this working. I tried for hrs ,
> without success yet.
>
> can any1 help ?????
> please reply to my mail adr.
>
> thx in advance...
>
> Volker Kalthaus
> [EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Rod Smith)
Subject: Re: Printing from Mac not possible
Date: 14 Feb 1999 04:38:49 GMT
Reply-To: [EMAIL PROTECTED]
[Posted and mailed]
In article <[EMAIL PROTECTED]>,
Martin Stenzel <[EMAIL PROTECTED]> writes:
>
> Hello,
>
> I am trying for several days now, still I am not succeeding:
>
> An iMac is linked to my Linux box with attached printer (Canon BJ
> 200e)at lp1 and I tried to make it execute print orders made from the
> iMac.
I recently set up something similar, but with Red Hat Linux 5.2, kernel
2.2.0-pre1, netatalk 1.4b2 with asun 2.1.2, Ghostscript 4.10, and MacOS
8.5.
> I am using the following:
>
> Kernel 2.0.36
> netatalk 1.4.2b
> ghostscript 3.33
> MacOS 8.1
Are you sure that's MacOS 8.1? My iMac came with 8.5. I just got it a
couple weeks ago, though....
Also, your version of Ghostscript is very old. I don't think your
problems are due to this, but you might want to look into an upgrade. If
you're using an RPM-based system, I recommend the Ghostscript RPMs from
http://www.users.dircon.co.uk/~typhoon/.
> Since my printer is not a PostScript printer I tried to set it up with
> help of the information given by Mike Pearson:
> <http://artoo.hitchcock.org/~flowerpt/projects/linux-netatalk/ghostscript.html>
>
> This is what I receive:
>
> Feb 9 06:15:48 plum xntpd[398]: kernel pll status change 89
> Feb 9 06:15:48 plum syslog: ROOT LOGIN ON tty1
> Feb 9 06:16:19 plum papd[478]: child 619 done
> Feb 9 06:16:40 plum papd[478]: child 633 for "gs10e" from 41201.173
> Feb 9 06:16:41 plum papd[633]: lp_init: lock: No such file or directory
>
> Feb 9 06:16:41 plum papd[633]: lp_open failed
> Feb 9 06:16:42 plum papd[478]: child 633 done
> Feb 9 06:17:30 plum afpd[617]: done
> Feb 9 06:17:30 plum afpd[480]: asp_chld 617 done
I don't know what these various error-like messages mean, I'm afraid.
> My Mac tells me that the document could not be printed due to a
> PostScript error (I like those "precise" error msgs!)
> This is all of the information I receive.
>
> I have to say that printing directly from Linux is possible.
Are you able to print PostScript files? For instance, if you set your
iMac to produce a PostScript file rather than to print to a printer, then
if you transfer that file over to the Linux box and try printing it with
lpr, does it print? Being able to do a "cat foo > /dev/lp1" or something
similar isn't enough.
> My papd.conf should be set up properly.
>
> #
> # Attributes are:
> #
> # Name Type Default Description
> # pd str ".ppd" Pathname to ppd file.
> # pr str "lp" LPD printer name.
> # op str "operator" Operator name, for LPD spooling.
> #
> gs10e:\
>:pr=lpr -Plp:op=apple:pd=/usr/local/atalk/etc/PPD/APLWNTR1.PPD:
>
> (I set it to pr=lp and pr=ps, too)
This is rather different from mine:
epson:\
:pr=epson:op=rodsmith:pd=/etc/atalk/ppds/epson800.ppd
My printer queue is called epson. If I understand correctly, the "op="
field has to point to a valid username on that system, so in your case,
you need to have a user called "apple" for your example to work. Also,
the PPD file you use is CRITICAL. You should use the same file on both
the Mac and the Linux sides, and you should make sure that the filename is
correct.
> My paths.h starts with:
This will only be relevant if you compile netatalk from scratch yourself.
I installed mine from an RPM.
One other tip: You can use the lpc command to disable printing but not
spooling on your queue to check to see if the file is making it to the
Linux queue or not. If it is but it's not printing correctly, then it's a
problem of matching the Mac's driver to Ghostscript, using the correct
PPD, or a more general Linux printing issue. If the file isn't getting to
the queue, then it's a netatalk configuration problem.
--
Rod Smith
[EMAIL PROTECTED]
http://www.channel1.com/users/rodsmith
NOTE: Remove the "uce" word from my address to mail me
------------------------------
From: "tonni" <[EMAIL PROTECTED]>
Subject: Re: IP masquerading
Date: Sun, 14 Feb 1999 04:39:13 GMT
i use like this with one nic and works fine no problem
echo "1" > /proc/sys/net/ipv4/ip_forward
/sbin/depmod -a
/sbin/modprobe ip_masq_ftp.o
/sbin/modprobe ip_raudio.o
/sbin/modprobe ip_irc.o
/sbin/ipfwadm -F -p deny
/sbin/ipfwadm -F -a m -S192.168.101.0/24 -D0.0.0.0/0
/sbin/ifconfig eth0:0 192.168.101.1 netmask 255.255.255.0
/sbin/route add -net 192.168.101.0
gateway to the windows box 192.168.101.1
jamie <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Kevin Martin <[EMAIL PROTECTED]> wrote:
> >In article <[EMAIL PROTECTED]>, it says
> >[EMAIL PROTECTED] wrote:
> >>Nightmare <[EMAIL PROTECTED]> wrote:
> >>>
> >>>ipfwadm -F -p deny
> >>>ipfwadm -F -a m -S 192.168.2.2/22 -D 0.0.0.0/0
> >>
> >>Perhaps it's me that's confused, as I'm just getting up to
> >>speed on subnet nomenclature and ipfwadm, but it appears
> >>as though you put in the gateway IP instead of the network
> >>address.
> >
> >It's you.
> >He's binding the gateway (192.168.2.2) to the Internet (0.0.0.0)
>
> Isn't that done with routing, not an ipfwadm rule?
>
> If 192.168.2.2 is the gateway, then that ipfwadm rule would cause the
> gateway to forward and masquerade it's own packets (?!) to anywhere,
> but won't forward and masquerade the network he's trying to gateway, no?
> Seems to me it should have been 192.168.0.0 or 192.168.2.0 (and I'm a
> bit weak on the 2-digit netmask suffix).
>
> --
> jamie ([EMAIL PROTECTED])
>
> "There's a seeker born every minute."
------------------------------
From: Kevin Dick <[EMAIL PROTECTED]>
Subject: Re: Compaq Deskpro 6000 and PCI BIOS
Date: Fri, 12 Feb 1999 15:32:26 -0600
If it's any consolation, we've had hours of trouble integrating these cards under
!!!WINDOWS 95!!!! Arg. Good luck. I suspect you'll be putting in a 3Com and
disabling the onboard NetFlex.
[EMAIL PROTECTED] wrote:
> Hello all, OK, I think I've read all the readme's and Howto's before
> coming here so I am desperate for any help. I need to get my onboard
> NetFlex-3 PCI Ethernet card working in Linux. Everything else works great
> with this RedHat 5.2 installation. I have a Compaq DeskPro 6000 with a
> Pentium Pro 200, Adaptec 2940U PCI SCSI Host adapter with a 2 2GB seagate
> hard drives, an IDE CD-ROM, 96MB RAM and a RedHat 5.2 Install CD.
------------------------------
From: "jim" <[EMAIL PROTECTED]>
Subject: Re: NIC Works Part Time ??
Date: Fri, 12 Feb 1999 11:40:02 -0800
RDF wrote in message ...
>Particulars:
>D-Link DE-220PCT cards
>PnP has been disabled
>Works fine in Win95
>Card in 486 computer works just fine Linux OS Debian 2.0.29
>NE2000 driver ne.o loaded as module
>Card in Pentium works part time Linux OS RedHat 2.0.31
>NE2000 driver ne.o loaded as module
>486 is Linux only. Pentium is dual boot Win95 and Linux
>
>Problem:
>When I boot Linux on the Pentium machine the NIC works only part of the
>time. I initialize on boot up but have to take the NIC down and then back
up
>several times before it will work at all.
>
>I have checked the interrupts and I/O address and they are OK.
>IRQ 10, I/O 0x300 and when the card works it does so very well. I have run
>out of ideas as to what the problem is and how to fix it.
>
>I have read and reread the HOWTO's and have searched Deja News for a
>description of the problem and a solution all to no avail.
>
>Any advice or suggestions would be appreciated.....
>
>Thanx
>Raymond
I just resolved a problem with a cheap ne2000 clone that works fine in a
Win95 machine... the card would not always respond to the same MAC address.
The problem was driving me nuts, I solved it when I accidentally enabled
arpwatch, which started mailing messages about 'changed Ethernet address'.
The fix was to buy another card.
-Jim
------------------------------
From: "tonni" <[EMAIL PROTECTED]>
Subject: Re: ip addresses
Date: Sun, 14 Feb 1999 04:42:26 GMT
copy and paste this to your /etc/rc.d/rc.local reboot using init 6 change
the ips in this case the 195.168.101.1 is the gateway to the winows box
good one nic will do the job
echo "1" > /proc/sys/net/ipv4/ip_forward
/sbin/depmod -a
/sbin/modprobe ip_masq_ftp.o
/sbin/modprobe ip_raudio.o
/sbin/modprobe ip_irc.o
/sbin/ipfwadm -F -p deny
/sbin/ipfwadm -F -a m -S192.168.101.0/24 -D0.0.0.0/0
/sbin/ifconfig eth0:0 192.168.101.1 netmask 255.255.255.0
/sbin/route add -net 192.168.101.0
Maat <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> hello,
>
> I have my own (little) network at home, a ftp server and a webserver but
>
> I have ordinary internet account and only one (static) ip address. But
> if I want to connect both machines connected to the internet they must
> have different ip addresses because they have different hostnames. Can I
>
> have two ip addresses on one line and how? And if not, is there another
> solution?
>
> Thanks!
>
> cheers,
>
> Johannes, http://www.chew.demon.nl
>
>
>
------------------------------
From: "tonni" <[EMAIL PROTECTED]>
Subject: Re: IP Masquerading - Dropping Packets
Date: Sun, 14 Feb 1999 04:35:13 GMT
copy in paste this in your /etc/rc.d/rc.local change the ips and reboot
using init 6
i use this in a cable modem setup in my house 1 nt server 3 nt wkst 1 linux
box 2 winnt wkst in the bedroom hook up tru the phone hires cat2 cables i
have link sys hub 8 port 10/100 it all work fine no packets drop or lost
and i have ssh server for some friends runing BitchX here is the file look
up for the gateway thats it
echo "1" > /proc/sys/net/ipv4/ip_forward
/sbin/depmod -a
/sbin/modprobe ip_masq_ftp.o
/sbin/modprobe ip_raudio.o
/sbin/modprobe ip_irc.o
/sbin/ipfwadm -F -p deny
/sbin/ipfwadm -F -a m -S192.168.101.0/24 -D0.0.0.0/0
/sbin/ifconfig eth0:0 192.168.101.1 netmask 255.255.255.0
/sbin/route add -net 192.168.101.0
PS: only one nic in the linux box save some money there hehe q:o)
phantom <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> 2 NIC's in the linux box would help so that the extra machines don't have
> any physical connection to the main cable:
>
> 128.210.195.1
> |
> |
> |
> +---128.210.195.38
> both in the linux box with ip forwarding/masq.
> +---192.168.38.1
> |
> |
> | rest of internal net.
>
> Its possible that there are just too many collisions going on.
> NB: the gateway for the internal machines would be
> 192.168.38.1 or whatever you give as the ip for the 2nd NIC
> in the linux box.
>
>
------------------------------
From: [EMAIL PROTECTED] (Clifford Kite)
Subject: Re: a *problem* with chatscript
Date: 13 Feb 1999 07:11:18 -0600
Julian Bordas ([EMAIL PROTECTED]) wrote:
: Hello
: I am having a bit of bother connecting to my ISP. The messages
: file says
: send (ATDT93972432^M)
: expect (CONNECT)
: ^M
: alarm
This is a timeout on finding CONNECT in the <CONNECT " "> following the
telephone number in your chat script. You apparently need more time to
dial, say the 45 second chat default. But you are on the ragged edge
of the minimum necessary timeout since you connect 1/3 of the time.
: Failed
: Connect script failed
: Exit
: Another error goes thus after ppp is running and the dynamic ip address
: has been assigned
: Hangup (SIGHUP)
: Modem hangup
: Connection terminated
: About a third of the time I dial in I do get a good connection. below
: is the chat script
: REPORT CONNECT
: TIMEOUT 3
: ABORT 'BUSY'
: "" /rAT
As the other post says the `/' should be a `/' but all you really need is
the `AT', and you might need to quote ('\r') to get the right thing sent.
I'd suggest instead resetting the modem before each call with ATZ for
the default user definable profile or AT&F for the factory default.
: TIMEOUT 30
: OK ATDT$93972432
: CONNECT " "
You will expect a space using this, which you may, or, more likely,
may not get. Better is "\c" which also suppresses an extra carriage
return that is otherwise sent and can cause problems. I need to qualify
this by saying in a small number of cases you actually need the extra
carriage return.
--
Clifford Kite <[EMAIL PROTECTED]> Not a guru. (tm)
/* Better is the enemy of good enough. */
------------------------------
From: [EMAIL PROTECTED] (Clifford Kite)
Subject: Re: Why is winsock ppp more robust that pppd?
Date: 13 Feb 1999 08:19:01 -0600
Chris Plachta ([EMAIL PROTECTED]) wrote:
: Hello,
: I recently installed Redhat 5.1 and Win98 on my P166 in a dual-boot
: configuration. I have a noisy phone line, and my pppd connection under
: linux frequently disconnects, especially when I'm trying to download
: large files. When I connect using PPP under Win98, I get a much more
: robust connection. I am getting very frustrated with the disconnects and
: I'm looking for any pointers on how to make my pppd connection more
: robust.
: The command I've been using is:
: /usr/sbin/pppd connect \'/usr/sbin/chat -v -f connect_script defaultroute
115200 crtscts user username debug /dev/modem
: When I disabled hardware flow control (i.e., changed "crtscts" to
: "nocrtscts") it seemed to get better, but I still can't ftp large files
: consistently with this connection.
: I'm looking for ways to make this connection more robust. Is there a way
: to configure pppd in exactly the same way as in Win98? I've been able to
: download large files much more reliably with my winsock PPP connection.
Unfortunately with a noisy phone line it's likely that you will have to
take a hit on modem speed by using less than 115200 for the pppd speed
option, perhaps less than the top speed for the modem itself. The
slower the speed the better the signal-to-noise ratio.
You can *try* using the pppd option "asyncmap 0" and removing any
"escape FF" or other escape option. If "asyncmap 0" just doesn't work
period, then try "asyncmap a0000". Don't disable hardware flow control
when using asyncmap.
--
Clifford Kite <[EMAIL PROTECTED]> Not a guru. (tm)
/* Governments should be changed like diapers - often and for the
* same reason. */
------------------------------
** 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
******************************
