Linux-Setup Digest #438, Volume #21 Thu, 14 Jun 01 09:13:14 EDT
Contents:
Re: Newbie Monitor help ("Andy Walker")
Re: Buying a CDR/RW (Stanislaw Flatto)
Re: Buying a CDR/RW ([EMAIL PROTECTED])
RAID-1 boot question ("Sven Lassen")
Remote network printing overseas ("Scotty")
Re: newbie - Port 80 (Robert Davies)
Re: Can I LINUX on this PC? (Robert Davies)
Re: Creating Raw Devices (Robert Davies)
Re: "Stable" 2.4.X Linux Kernels (Robert Davies)
Re: install linux on one harddrive and windows on another? (Robert Davies)
Re: Cannot open master raw device '/dev/rawctl' (No such device) (Robert Davies)
Re: Can't Boot Machine! (Robert Davies)
Re: rpms (Robert Davies)
Re: New HD Install (Robert Davies)
Re: I can't mount ...? hardware problems (Robert Davies)
Re: linux ftp, ncftp etc (Robert Davies)
----------------------------------------------------------------------------
From: "Andy Walker" <[EMAIL PROTECTED]>
Subject: Re: Newbie Monitor help
Date: Thu, 14 Jun 2001 05:07:21 -0000
Steve wrote in message <3b281e02$0$18898$[EMAIL PROTECTED]>...
>I had used the graphic monitor and resolution change in KDE to
>change to my specific video card and to a 1024x768 refresh rate.
>The settings worked fine for the rest of the session. However, on
>reboot, my machine now boots to text mode rather than graphical.
>When I do a startx, it tries to go graphical but then bails back
>out to text mode and gives me the following error:
>
>Fatal Server Error:
>Could not open font 'fixed'
>
>I tried to change the settings back using xf86config but that did
>not help. Can someone help me out? I am using Mandrake 7,2
>
>
>
>
Mandrake7.2 comes with XFree86-3.3.6 and XFree86-4.0.1. Which one were/are
you using because, if you were using X-4.0.1, the Mandrake software will
screw it up and change it back to X-3.3.6 instead (or at least that's what
happened to mine).
If you want more help then I need more info. such as the log file
/var/log/XFree86.log to go further. Take a look at the file
/etc/X11/XF86Config as this can be set up manually.
------------------------------
From: Stanislaw Flatto <[EMAIL PROTECTED]>
Subject: Re: Buying a CDR/RW
Date: Thu, 14 Jun 2001 21:54:20 +1000
Just a comment:
You want them to perform fast, then get SCSI.
BTW who are the manufacturers of those units, not distributors?
Stanislaw.
Slack user from Ulladulla.
Julian Bordas wrote:
> Hello
> I'm looking at buying a CDR/RW and have narrowed it down to three
> options
>
> Creative 32 8 4
> LG 32 8 4
> Cyber 32 8 4
>
> Any comments or suggestions on these models?
>
> TIA
>
> Julian
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Buying a CDR/RW
Date: 14 Jun 2001 11:59:09 GMT
Stanislaw Flatto <[EMAIL PROTECTED]> wrote:
> BTW who are the manufacturers of those units, not distributors?
AFAIK Creative and LG are Mitsumi and Cyber should be Matsushita.
Davide
------------------------------
From: "Sven Lassen" <[EMAIL PROTECTED]>
Subject: RAID-1 boot question
Date: Thu, 14 Jun 2001 13:59:52 +0200
Hi!
I successfully set up a RAID-1 Konfiguration on two SCSI disks (sda, sdb)
using raidtool 0.9. It seems to work fine even if one disk fails.
My question is now: How can I boot the system (without a boot floppy) if the
first disks fails?
I�ve already tried to find a solution following the instructions in the
"Boot + Root + Raid + Lilo : Software Raid HOWTO" but it does not work. I
think there is something wrong with my Lilo Setup.
My current setup looks as follows:
- two lilo.conf files in /etc (lilo.conf.sda and lilo.conf.sdb)
lilo.conf.sda:
disk=/dev/md6 # raid device is available but not used
bios=0x80
sectors=63
heads=255
cylinders=527
partition=/dev/md1 # my /boot partition
start=526
boot=/dev/sda
lba32
map=/boot/map
install=/boot/boot.b
image=/boot/raid_ipsec
root=/dev/md0
read-only
label=raid_ipsec
lilo.conf.sdb:
disk=/dev/md6 # raid device is available but not used
bios=0x80
sectors=63
heads=255
cylinders=555
partition=/dev/md1 # my /boot partition
start=526
boot=/dev/sdb
lba32
map=/boot/map
install=/boot/boot.b
image=/boot/raid_ipsec
root=/dev/md0
read-only
label=raid_ipsec
Lilo has been installed in the MBR of both disks using the following
commands:
lilo -b /dev/sda -C /etc/lilo.conf.sda
lilo -b /dev/sdb -C /etc/lilo.conf.sdb
If I disconnect sda from the SCSI bus and try to boot I get the following:
L1 01 01 01 01 01 ...... (and so on)
Any help to find a solution would be greatly appreciated
Sven
------------------------------
From: "Scotty" <[EMAIL PROTECTED]>
Subject: Remote network printing overseas
Date: Thu, 14 Jun 2001 12:36:51 GMT
Question on Network Printing to a specific Port - - -
My client has a network in Michigan, running a DSL router with a fixed ip
address. (using Network Address Translation)
He has a linux box on the network. 192.168.50.10
DSL router ip (ethernet) address - 192.168.50.1
DSL router ip (WAN) address - 10.10.10.1 (just for this example)
Linux box is configured to be able to hit the internet - has the DSL router
as gateway, etc....
He also has an office in the Netherlands with a ISDN router, with a fixed ip
address. (using Network Address Translation)
This router supports port / ip address forwarding.
ISDN router ip (ethernet) - 192.168.60.1
ISDN router ip (WAN) - 20.20.20.1 (just for this example)
Local IP for printer in netherlands 192.168.60.10
What we need is to be able to have the linux box in Michigan have a print
queue - ie - NETHER1 - print to an
address on a specific port. The address would be the wan ip of the router
in Netherlands, (20.20.20.1) the port would be 9015.
If the linux print queue NETHER1 could print to 20.20.20.1 on port 9015,
the router in Michagan would just pass the request on
to the internet. Then the router in Netherlands could very easily be setup
to redirect the inbound
request coming in on port 9015 to 192.168.60.10, on port 9100 on the local
network to the print
server. (9100 is standard for HP jet direct print servers)
We just can't figure out if, and if so, how to get Linux to print to an ip
address on a particular port.
All the information I can find is that Network printing to an HP is done to
a port called RAW, or something to that effect.
Please email / respond to newsgroup if you think you could provide
assistance
Scotty
Technical Support Consultant
REALTIME SOFTWARE CORPORATION
http://www.realtimesw.com
[EMAIL PROTECTED]
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: newbie - Port 80
Date: Thu, 14 Jun 2001 11:41:20 +0100
Bob Bourne wrote:
> Robert Davies, in part wrote (and I thankyou very much):
> Installed wwwoffle (luv that name) and I've obviously got a another
> mountain of reading to go through but... /etc/locale.gen (see PS below)
> and run local-gen as root - well, locale.gen is a complete mystery for me,
> could you please give me some simple pointers...
Just use the defaults for now, and in wwwoffle.conf, look for the lines
with /Proxy in them. Worry about locale's later, you speak English, and Oz
not too different from US and UK, no?
To set that proxy up, it should look like this in wwwoffle.conf :
#
# Proxy
# -----
#
# This contains the names of the HTTP (or other) proxies to use external to
the
# WWWOFFLE server machine.
#
# [<URL-SPEC>] proxy = (host[:port])
# The hostname and port on it to use as the proxy.
#
<snip>
Proxy
{
<http://*> proxy = none
There you go Bruce! Just change that to read :
# proxy.sale.vic.australis.com.au - Port 80
<http://*> proxy = proxy.sale.vic.australis.com.au:80
Then in your browser set localhost:8080 as your proxy, assuming you took
the wwwoffle defaults.
Simple!
I do agree that this ISP seems a bit amateurish, for a start proxies are
usually run on 8080 which is service 'webcache'.
If you have KDE2 running on your system KInternet supports wwwwoffle, it
will tell wwwoffle for you when you are on or off line.
The alternative is to put 'wwwoffle -online' in /etc/ppp/ip-up script, and
-offline in ip-down script. You can also use an -autodial option if you
use demand dialing.
Rob
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: Can I LINUX on this PC?
Date: Thu, 14 Jun 2001 11:08:01 +0100
Guz wrote:
> 815E chipset on CUSL2 MoBoard
> BIOS 1003
> PIII 800EB -133
> 512SDRAM memory 133.(non ecc).
> Sound Card is SB Live! Platinum.
> OS W98SE
>
> And I have the 3 CDs packet from Debian-Linux.
>
> Can I install this OS on this PC? ...
Yes.
> In the file "boot.bat" It says that
> " CPU is in V86-Mode (May be WINDOWS, EMM386, QEMM, 386MAX,...) You need
> pure 386/486 real mode or a VCPI server to boot Linux..."
>
> One more question:
> I can't boot from my CD-ROM Drive, How can I make a bootable Floppy Disc
> in order to have acces to my CD and install all the rest?
Why not? The Asus CUSL2 surely supports it. Did you scan for all disks in
one of the BIOS menus? I did that and had same thing once.
In the first BIOS menu, change all IDE stuff to AUTO.
>From Win or DOS, there's a program included on the Linux CD-ROMS, called
rawrite, there are various boot image files depending on what hardware you
have included on the CD-ROMs as well.
Sometimes with less usual SCSI cards, a boot floppy is necessary to be able
to load a module, which drives the card the SCSI CD-ROM is on.
I think this is academic in your case, as long as the CD-ROMs, show lots of
files, and not one .iso then they should be bootable.
If it's your first time with Linux, then you might find obtaining a
Mandrake 8.0 CD (2 CD set, one inst for basic installation, and one ext for
extra packages) will provide a more graphically oriented system, with
easier configuration, it detects most things automatically. They can be
downloaded and burned with Win CD-burning software.
Rob
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: Creating Raw Devices
Date: Thu, 14 Jun 2001 10:54:35 +0100
Jaishni Govender wrote:
> Can anyone please help or point me in the right direction. I need info on
> how to create raw devices as I am trying to install oracle parallel server
> on RedHat 6.2. I can't find any info or documentation on this. any help
> would be greatly appreciated.
See if you have a command called raw(8), ie type 'man 8 raw' to view docs.
RAW(8) RAW(8)
NAME
raw - bind a Linux raw character device
The other answerer misunderstood what you wanted, you're trying to bind a
parition with a block device, to a unbuffered character device.
If it doesn't work, just use the block device eg) /dev/hda6, which will
give Oracle a device file.
But most Oracle installations, use Tablespaces in file systems, because
they're much easier to maintain.
Rob
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: "Stable" 2.4.X Linux Kernels
Date: Thu, 14 Jun 2001 10:50:03 +0100
[EMAIL PROTECTED] wrote:
> On Wed, 13 Jun 2001 08:29:25 -0700 (PDT), patrick
> <[EMAIL PROTECTED]> wrote:
>>I just updated my Linux kernel from 2.2.16-22 to 2.2.18 and
>>life is good. (I was having a problem with the 2.2.16-22 kernel.)
>>
>>However, there are now 2.4.0-2.4.5 kernels. Are there any
>>2.4.X kernels which are known to be better than others?
>>
>>Please advise.
>
> I'd say to go with the latest'n'greatest version. I currently run 2.4.4
> and
> it works like a champ with one exception: it's matrox mga driver is out
> of date and I had to go to xfree.org's cvs site to get a usable driver.
>
> The 2.4.x kernel is the most reliable software I've ever experienced in my
> 25 years of computing.
Well you can't be putting your system into swap then!
2.2.19 is far more stable on most ppl systems, and the 2.4 currently needs
much greater swap space size, 2xRAM at a minimum, so it could be an awkward
upgrade.
2.4.6pre series have some important fixes for VM system. Upgrading a 2.2
system to 2.4 is not so easy, you need to read Moshe Bar's article at
Byte.com, and Documentation/Changes in the new source.
My advice, is wait until 2.4 has settled down a little more. 2.4.6 might
be the best release, all the others seem to have had problems, but look
after it's been release for any patches issued in a 2.4.7pre series.
The other issue is whether Linus kernels or -ac series are best ones to
use. Most distros will proably go with Alan Cox, as he is supporting new
major device number assignments, whilst Linus has refused any new numbers.
Rob
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: install linux on one harddrive and windows on another?
Date: Thu, 14 Jun 2001 10:54:38 +0100
Erick Woertz Jr. wrote:
> I've got ME on one harddrive right now, and I just came across a free
> harddrive at work, and want to put linux on that one, keep windows on the
> one in my computer now, and set up a dual boot that way. is it possible
> at
> all? can anyone help?
There's no magic to this. Just install LILO on the MBR of your new disk,
and then after make sure you have an entry like the following.
>From lilo.conf too boot a windows in 2nd primary partition, on ide2.
other = /dev/hde2
label = windows
map-drive = 0x80
to = 0x81
map-drive = 0x81
to = 0x80
table = /dev/hde
Change hde2 to whatever partition, your Linux install thinks win is, and
hde to the disk that windows lives on.
Rob
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: Cannot open master raw device '/dev/rawctl' (No such device)
Date: Thu, 14 Jun 2001 11:13:53 +0100
Sean Marshall wrote:
> After upgrading from the standard redhat 7.0 kernel to kernel 2.2.19 I
> find I cannot access rawdevices such as hard drives. When I run the raw
>
> -qa command I get the following error message Cannot open master raw
> device '/dev/rawctl' (No such device)
>
> I need to be able to access raw SCSI devices because I'm running
> sanergy, can anyone help me fix this problem?
RedHat patch their sources heavily, did you download a source RPM from
them, or an rpm binary? If you built from source, perhaps you forgot a
configuration option you need, for raw I/O.
I think you would be wiser to stick with distribution source or binary
kernels, rather than download 'Linux' standard, as they have added support
for distribution features, things like Large Files (>2GB), LVM, and
improved NFS, long before they make it into Linus standard.
Rob
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: Can't Boot Machine!
Date: Thu, 14 Jun 2001 11:18:40 +0100
David wrote:
> Rand Simberg wrote:
>>
>> I just upgraded the kernel on my firewall from 2.2.19 to 2.4.5.
>> Almost...
Did you read Documentation/Changes and upgrade all the utilities, Moshe Bar
has an article on moving to 2.4 at Byte.Com.
>> I did a successful upgrade on my work station, so I went into the
>> project with full (and apparently unjustified) confidence. Everything
>> went well, make menuconfig, make dep, make clean, make modules, make
>> modules_install.
>>
>> I copied the bzImage to /boot, added it to lilo, keeping the original
>> one.
And /boot/System.map?
You should allways make new entries for experimental kernels, and only
replace your default kernel, once you are really sure it is solid.
>> When I rebooted, it said "loading linux, and uncompressing linux," and
>> then nothing. Just died. No keyboard even.
>>
>> So I reset, and tried booting into the old kernel. Same thing. Died
>> upon "uncompressing linux."
>>
>> So I got a rescue disk, and typed rescue on the prompt.
>>
>> "uncompressing linux"
>>
>> Nothing.
>>
>> What now?
>
>
> You should be able to use the installation boot disk and/or CD to boot
> the system. This will boot the kernel on the CD.
>
> At the installation boot prompt enter:
>
> linux rescue root=/dev/hdaX
>
> Once you have it back up and running you can fix the problem or
> recompile the kernel if needed. After you edited /etc/lilo.conf did you
> run "lilo" to update the MBR with the changes you made to the
> /etc/lilo.conf file?
Another thing is to make sure you do not have modules lying around, that
you have compiled into the kernel, often modules.conf needs updating if you
do compile in.
Rob
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: rpms
Date: Thu, 14 Jun 2001 11:01:24 +0100
J Hayward wrote:
> Yes, you can use --prefix also, It works slightly different from
> --relocate. --relocate only changes the install location of the files in
> the directory you specifiy, ie oldpath. --prefix changes the install
> locations of all files in the rpm.
>
> To display a list of files, this will list the full path where they are
> installed, in an rpm that hasn't been installed yet use:
>
> rpm -qpl foo.rpm
Great explanation J!
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: New HD Install
Date: Thu, 14 Jun 2001 11:21:39 +0100
Cameron Loranger wrote:
> I'm running a Win98 box right now, and to rid myself of this os, I've
> bought a new 40gb hard drive to put linux on. I have also purchased Red
> Hat Linux 7.1 Deluxe, so I have cd's. There is no way linux will need that
> much space, so, I'd like to partition this new drive for both fat32 (10gb)
> and leave the linux partition 30gb. I'm wondering how I should go about
> partitioning this new hard drive.
>
> Should I format the fat32 first and then the linux partition, or install
> linux first? I'd really like to do this, but if the difficulty really
> negates the reason, feel free to let me know.
Just let your distro parition the drive for you. You can usually run an
option, to create a spare partition, which you can later change to FAT32,
and Format with windows.
The RH Disk Druid software is graphical partitioner, I think it will let
you create a FAT32 partition, during the Linux install.
Rob
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: I can't mount ...? hardware problems
Date: Thu, 14 Jun 2001 11:22:36 +0100
VP wrote:
> I can't mount my sdb7 reiserfs partition (I have there some importante
> files, my sources codes)
>
> When I going to mount the partition the kernel(2.4.5) crash with this
> logs
Go to linuxtoday.com and read the release notes for 2.4.6pre series.
Rob
------------------------------
From: Robert Davies <[EMAIL PROTECTED]>
Subject: Re: linux ftp, ncftp etc
Date: Thu, 14 Jun 2001 11:30:00 +0100
Riyaz Mansoor wrote:
> David Efflandt wrote:
>
>>
>> Yes, but have you edited $HOME/.ncftp/firewall for your firewall-host and
>> firewall-port, maybe with firewall-type=1? Can Netscape access ftp://
>> URL's when properly configured for your proxy.
>>
>
> i tried 3 and 6 which seemed the closest to me. but i'll try 1.
>
>
>>
>> > any other tips?
>>
>> Get a different ISP. Otherwise you may never be able to telnet or ssh to
>> an internet host (free shell acct, etc.), or use any other POP3 servers,
>> or do irc, or use time servers...
>>
>> >
>
> hehe. can't switch. :(
>
> on a different note, aren't there linux ftp clients that handle http
> proxy/firewalls?
Yes there are, ncftp for one! The question is though, wether your
proxy/firewall using some nasty undocumented proprietary format with
special windows only clients, that noone can write programs for? I find
that unlikely, more likely is that you have made a configuration error.
As an alternative, does your browser use a proxy? These support ftp as
well, so you could use the browser for ftp downloads if you tell it to use
the proxy for all protocols.
The right solution is to find out what proxy you are offered, and configure
your ftp client to use that.
Rob
------------------------------
** 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 by posting to comp.os.linux.setup.
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-Setup Digest
******************************
