Linux-Setup Digest #73, Volume #19 Tue, 4 Jul 00 15:13:11 EDT
Contents:
Fun with large disk drives ([EMAIL PROTECTED])
Re: Bad Xit (MDLamberts)
Re: Need a clean hard disk ([EMAIL PROTECTED])
Re: Bad Xit ("D F")
Re: dns setup ("Chuck Swiger")
Re: LILO (Sjoerd >)
Remote printing ("Peter C. Lee")
Re: dns setup (Paul Kimoto)
Settings problems (Re: My Linux Adventure) (Laura Goodwin)
Re: dns setup (Homer Jay)
Re: Win98SE running inside Linux ???
Re: dns setup ("Chuck Swiger")
Re: Changed from Storm to Mandrake-have some new questions ("Ken Acker")
----------------------------------------------------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: de.comp.os.unix.linux.hardware,rec.video.desktop,de.comp.tv+video
Subject: Fun with large disk drives
Date: Tue, 04 Jul 2000 17:38:56 GMT
(This lengthy article deals with installing big IDE disk drives under
Windows98 and Linux. Maybe also useful for people interested in video
editing.)
See also: http://www.linuxdoc.org/HOWTO/Large-Disk-HOWTO.html
I bought a 60 GByte IDE disk drive the other week, because digital
video is just too painful with anything less than 20 GByte.
And I underestimated the effort to get it going by at least one
magnitude.
First complication: After hooking up the new disk drive, my PC did not
even boot. It turned out that the BIOS can't deal with anything larger
than about 30 GByte. This is another hard limit coded in many older
BIOSses. I thought that the BIOS programmers might have learned
something from the 520 MByte incident or the 2 GByte barrier or the 8
GByte barrier, but no! It is probably a big surprise to them that the
disk capacity is increasing all the time.
After wasting one or two hours I searched the net and I learned that
large disk drives usually have a special jumper setting that limits the
size to about 30 GByte. I went to the Maxtor website, found the
settings, installed the jumper and I could at least boot, although my
BIOS would not see the full 60 GByte disk space. But at least I was not
forced to buy a new motherboard.
Second complication. After playing with this 30 GByte configuration for
a few hours I discovered that Maxtor provides a small program that
installs in the boot sector and fixes this limit. I downloaded this
program (MaxBlast) from www.maxtor.com and let it install itself in the
boot sector. I had to do this twice because the program requires a
bootable windows95 or 98 disk. So I aborted it, found a windows98 disk
drive, booted that, made an emergency disk and rerun the MaxBlast
installation again. Good.
Third complication. My Linux disk is a SCSI disk, the IDE disk is
actually only storage for those bulky digital video files. Since the
startup disk is on SCSI, the BIOS launches the code from the SCSI disk,
which is the Linux loader. So Linux would start, but the MaxBlast code
is not installed and Linux can only see 30 GByte of my new expensive
disk. And since it is already formatted for 60 GByte it can't even
mount it.
My next idea was changing the boot order to IDE first and I was
successful in running the MaxBlast code. However the IDE drive did not
contain an operating system (with the exception of command.com, but
this probably does not count as an operating system by anybody`s
imagination), so I was staring at the dos prompt without having any
means to start Linux.
I then spent a couple of hours trying to install Lilo somehow on the
IDE drive so that it would go out to the SCSI disk and fetch the kernel
from there, but I could not get it running. Maybe there is a way to do
it but I could not figure it out to install both MaxBlast and lilo on
the same disk.
Needless to say that MaxBlast won't install on the SCSI disk either.
Anyway. After searching the net again I discovered that I can use
certain settings in lilo.conf that will override the broken BIOS
values. For my Maxtor 96147U8 I found these values: append
= "hda=7473,255,63" . You can calculate these values by dividing the
total number of blocks by 255 (for 255 heads) and then by 63 (for 63
sectors per track). Sometimes these values are also printed on the disk
drive label. It is also possible to use other values for heads and
sectors but the logic behind that is beyond be. However this particular
set of values works for me and my particular disk drive.
So here is my complete working /etc/lilo.conf file:
boot = /dev/sda
vga = normal
read-only
compact
prompt
timeout = 30
disk = /dev/sda
bios = 0x80
image = /boot/vmlinuz
label = linux
root = /dev/sda3
initrd = /boot/initrd
append = "hda=7473,255,63"
After booting Linux I could finally mount the drive and see all 60
GByte of it! Great! I was so happy.
But the problem was fixed only halfway. I want to download digital
videos from my Camcorder using Linux and dvgrab (shameless plug:
http://www.schirmacher.de/arne/dvgrab/index_english.html ), but there
are currently no really good nonlinear video editing programs available
for Linux. I have to use one of the Windows programs like ULead
VideoStudio for this job. The easiest solution for this is to use a
separate partition just for data, format this using Windows98 and mount
it under Linux using the vfat file system type.
Here's the entry in /etc/fstab for my big 60 GByte disk partition:
/dev/hda1 /hd vfat noauto,user 0 0
You need to create the directory /hd before that. The noauto option
keeps it from being mounted during startup, because it would be mounted
as root at that time. The vfat filesystem knows only one user, which is
whoever mounted it, and therefore you could access it only using the
root account if it is mounted at startup. The user option allows the
mount operation for everybody instead of only root.
So when I need the disk space I just type : mount /hd and the disk is
there.
At this time I had still no way to access the drive from Windows98SE,
where my video editing software lives. The MaxBlast description
suggests a full new installation of Windows from CDROM, but this was
not an option because I had plenty of special drivers and settings and
I did not want to redo everything again. So I copied the whole W98SE
disk to a file server, started Linux again and copied the whole thing
back from the server to my /hd directory. I then shut down Linux and
tried to start Windows. Guess what happened? It came up. I was very
impressed.
Well, I was not quite there yet. It turned out that during the file
copy certain file names were mangled (files having the german umlaut
characters for example). I ran the DOS chkdsk (or was it dskchk?
scandisk? I can't remember) and let it delete all bogus files and
directories. After that I started the Windows98 master disk (for the n-
th time), and this time I used the backup utility to make a regular
backup of my original Windows98 installation. I started the new 60 GB
disk for my final move and reinstalled everything from the backup image
file on the network server. Luckily the backup program and network
driver worked (they don't contain any umlaut characters in their file
names).
Wow. After about one weekend of my precious limited time I had this
whole setup in an useable state (sort of). I am really impressed. The
only thing I wanted was an IDE disk shared between my Linux SCSI disk
and my Windows98SE SCSI disk.
Btw, from time to time I have write errors when writing to this disk :
Jun 25 11:19:10 hal kernel: hda: dma_intr: status=0x51 { DriveReady
SeekComplete Error }
Jun 25 11:19:10 hal kernel: hda: dma_intr: error=0x10 {
SectorIdNotFound }, LBAsect=66055133, sector=66055070
Jun 25 11:19:10 hal kernel: hda: DMA disabled
Jun 25 11:19:10 hal kernel: ide0: reset: success
Jun 25 11:19:10 hal kernel: hda: write_intr error1: nr_sectors=128,
stat=0x51
Jun 25 11:19:10 hal kernel: hda: write_intr: status=0x51 { DriveReady
SeekComplete Error }
Jun 25 11:19:10 hal kernel: hda: write_intr: error=0x10 {
SectorIdNotFound }, LBAsect=66055133, sector=66055070
Jun 25 11:19:10 hal kernel: hda: write_intr error1: nr_sectors=128,
stat=0x51
I could not really identify the problem. When this happens, Linux hangs
and Windows98 scandisk reports corrupt files or directories. It might
be either a bug in the vfat filesystem or my lilo.conf settings are not
quite ideal.
If you are working on similar problems and this article saves some of
your time, good. If you have any suggestions how to improve this setup,
please let me know (reply to this thread or mail to
mailto:[EMAIL PROTECTED] .)
Thanks, Arne
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: [EMAIL PROTECTED] (MDLamberts)
Subject: Re: Bad Xit
Date: 04 Jul 2000 17:51:36 GMT
I know that "shutdown -h now" shuts down the machine. My point is that once I
am in X, I cannot get out cleanly - not to exit back to the command prompt and
cannot properly shutdown the machine. Any attempt to exit, logout, or whatever
results in a blank screen and no response from the computer. After that, I have
no choice but to do a hard shutdown.
I have not tried using the <Ctrl><Alt><F1-6>, so I'll give that a shot.
Thanks.
------------------------------
Crossposted-To:
alt.comp.hardware,comp.os.linux.hardware,comp.os.ms-windows.nt.setup.hardware,comp.sys.ibm.pc.hardware.storage
Subject: Re: Need a clean hard disk
From: [EMAIL PROTECTED]
Date: Tue, 04 Jul 2000 17:54:14 GMT
lament <[EMAIL PROTECTED]> writes:
> To get rid of old information use eraser, at http://www.tolvanen.com/eraser/
Or `dd if=/dev/zero of=<drive> ibs=<sector size> count=<sectors on the
drive>'. For example, on a standard 1.4M floppy, `dd if=/dev/zero
of=/dev/fd0 ibs=512 count=2880'.
You asked in at least two Linux groups, so I assume you have Linux.
--
Eric P. McCoy ([EMAIL PROTECTED])
"Dude... my hands are huge. They can touch anything but themselves...
oh, wait."
------------------------------
From: "D F" <[EMAIL PROTECTED]>
Subject: Re: Bad Xit
Date: Tue, 4 Jul 2000 13:55:35 -0400
MDLamberts wrote in message
<[EMAIL PROTECTED]>...
>I know that "shutdown -h now" shuts down the machine. My
point is that once I
>am in X, I cannot get out cleanly - not to exit back to the
command prompt and
>cannot properly shutdown the machine. Any attempt to exit,
logout, or whatever
>results in a blank screen and no response from the
computer. After that, I have
>no choice but to do a hard shutdown.
>
> I have not tried using the <Ctrl><Alt><F1-6>, so I'll give
that a shot.
>Thanks.
If all else fails, Ctrl-Alt-Bkspc should kill the X server
if you originally logged in init 3.
Dave Fluri
North Bay, Ontario Canada
------------------------------
From: "Chuck Swiger" <[EMAIL PROTECTED]>
Subject: Re: dns setup
Crossposted-To: comp.os.linux.networking
Date: Tue, 04 Jul 2000 18:00:57 GMT
In comp.os.linux.networking Tomas Kral <[EMAIL PROTECTED]> wrote:
> Can anybody clarify what the following really means, shall I be really
> worried about it, or is it just an informative named message.
>
> ...
> Jul 4 15:31:13 scisys named[3012]: Lame server on 'bbnrel1.cns.hp.com'
> (in 'cns.hp.COM'?): [156.153.255.210].53 'atlrel1.hp.com'
> ...
It's an example of lame delegation. The nameserver which is supposed
to be authoritatively answering questions about the cns.hp.com domain
is not doing so.
You can ignore it, unless you want to send mail to an address in that
domain-- in which case, your email will probably not go through until
HP fixes that zone.
-Chuck
Chuck 'Sisyphus' Swiger | [EMAIL PROTECTED] | Bad cop! No Donut.
------------------------+-------------------+--------------------
I know that you are an optimist if you think I am a pessimist....
------------------------------
From: Sjoerd <"Sjoerd <skrol"@inter.nl.net>>
Subject: Re: LILO
Date: Tue, 04 Jul 2000 08:52:22 +0200
Joe Chiasson wrote:
>
> If LILO was installed in your master boot record for your Windows partition,
> boot to Windows from a floppy and run fdisk /mbr.
Or if that doesn't help, when it isn't installed in the mbr : run fdisk
(windows/dos version), and change the active partition to the windblows
one..
Greetings,
Sjoerd
------------------------------
From: "Peter C. Lee" <[EMAIL PROTECTED]>
Subject: Remote printing
Date: Tue, 4 Jul 2000 14:06:40 -0400
Hi there,
I have remote printing working under Linux.
The Linux printing server has a fixed IP address. The other linux machines
obtain IP addresses through DHCP. Once a machine obtains an IP address, it
usually re-obtains the same address after reboot. So I list those IP
addresses in /etc/hosts.lpd instead of all possible IP address under the
same domain. However the problem is sometimes that a machine may obtain a
new IP from the DHCP server after rebooting that is not on the list in
/etc/hosts.lpd.
Question:
Is these a better way to edit /etc/hosts.lpd that would include all possible
IP addresses in the same domain. e.g 192.168.1.[2 to 244]
Thanks, Peter
------------------------------
From: [EMAIL PROTECTED] (Paul Kimoto)
Crossposted-To: comp.os.linux.networking
Subject: Re: dns setup
Date: 4 Jul 2000 14:21:38 -0500
Reply-To: [EMAIL PROTECTED]
In article <tDp85.11217$[EMAIL PROTECTED]>,
Chuck Swiger wrote:
> In comp.os.linux.networking Tomas Kral <[EMAIL PROTECTED]> wrote:
>> Jul 4 15:31:13 scisys named[3012]: Lame server on 'bbnrel1.cns.hp.com'
>> (in 'cns.hp.COM'?): [156.153.255.210].53 'atlrel1.hp.com'
> It's an example of lame delegation. The nameserver which is supposed
> to be authoritatively answering questions about the cns.hp.com domain
> is not doing so.
>
> You can ignore it, unless you want to send mail to an address in that
> domain-- in which case, your email will probably not go through until
> HP fixes that zone.
(Won't your local named try to find a non-lame server first?)
If these messages are of no use to you, you can add a line to the logging{}
section of your named.conf, like
// reduce log verbosity on issues outside our control
logging {
category lame-servers { null; };
};
--
Paul Kimoto
------------------------------
From: Laura Goodwin <[EMAIL PROTECTED]>
Subject: Settings problems (Re: My Linux Adventure)
Date: Tue, 04 Jul 2000 13:43:58 -0400
Reply-To: [EMAIL PROTECTED]
Laura Goodwin wrote:
> Ironic: Normally I'd be screaming at Scotty for more bandwidth, but
> right now I'd be satisfied with a 16 bit, 800x600. Funny how an
> emergency situation makes you reorganize your priorities....
>
> (to be continued)
I figured out how to get more colors, but the display is still
distorted. And how to enable my modem (I repeat, NOT a Winmodem)?
I'm having problems figuring out how to set my monitor display to what I
want in Linux, and this is *after* consulting two different Linux
manuals. I can do it through the terminal emulation via KDE, but how?
I can give you my monitor specs:
Horizontal = (KHz)30-70
Vertical= (Hz)50-120
Vid bandwidth= 110 MHz
Maximum resolution= VESA 1280x1024 (fh KHz=63.981, fv Hz=60.02)
Max Colors= unlimited
Size= 17" (CRT)
dot pitch= .26
Viewable area=300x225
Need more details? I got 'em all in my monitor manual. All anyone can
tell me is "trial and error" but that is NOT GOOD ENOUGH. Trial and
error will not be necessary. Just tell me where I go to input the
freaking data! I can set and reset my monitor settings in sixteen
different ways with a few clicks and little risk in Windows, and if I
can't do that in LINUX then no wonder the OS isn't catching on! How the
display looks is pretty basic!
Being able to set up TCP/IP and simply sign on is also pretty basic and
should be a bit more obvious, IMHO. Any enlightenment available here, my
dears? With Windows anything and BeOS, all I had to do was poke around
a bit to find the place to input my dialup settings. Where is it in
LINUX? I got the info...where do I input it?
--
Laura Goodwin
"Pain is fleeting, glory is forever.
Remember: scars are sexy."
------------------------------
From: Homer Jay <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.networking
Subject: Re: dns setup
Date: Tue, 04 Jul 2000 18:34:11 GMT
> Can anybody clarify what the following really means, shall I be really
> worried about it, or is it just an informative named message.
>
> ...
> Jul 4 15:31:13 scisys named[3012]: Lame server on 'bbnrel1.cns.hp.com'
> (in 'cns.hp.COM'?): [156.153.255.210].53 'atlrel1.hp.com'
> ...
I don't think you should worry about it at all. It happens to me
all the time. I looked it up a while ago at:
http://www.acmebw.com/askmrdns
His answer is:
A "lame server" is one that should be authoritative for a zone, according
to the NS records that delegate authority for that zone, but really isn't,
perhaps because it hasn't been configured to load the zone or because it
is a secondary master that has expired the zone.
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: [EMAIL PROTECTED] ()
Subject: Re: Win98SE running inside Linux ???
Reply-To: [EMAIL PROTECTED]
Date: Tue, 04 Jul 2000 18:54:13 GMT
On Tue, 4 Jul 2000 15:18:10 +0800, Jeff Williams <[EMAIL PROTECTED]> wrote:
>Is it possible to run a Windows98SE session inside Linux? If it is, which
>version & build of Linux do I have to install to have this possible?
see www.vmware.com
------------------------------
From: "Chuck Swiger" <[EMAIL PROTECTED]>
Subject: Re: dns setup
Crossposted-To: comp.os.linux.networking
Date: Tue, 04 Jul 2000 19:01:14 GMT
In comp.os.linux.networking Paul Kimoto <[EMAIL PROTECTED]> wrote:
> In article <tDp85.11217$[EMAIL PROTECTED]>,
> Chuck Swiger wrote:
>> You can ignore it, unless you want to send mail to an address in that
>> domain-- in which case, your email will probably not go through until
>> HP fixes that zone.
>
> (Won't your local named try to find a non-lame server first?)
Sure. But I've noticed that when there's a lame delegation, there
generally won't be a non-lame server available, either. :-)
We had a client firewall off their DNS servers once, and we had to set
up false delegation for their zone just to be able to get mail
through. They didn't fix it until two or three weeks later, when they
stopped getting mail from most other sites (which did not demand
AA's), as they expired the cached MX records...
-Chuck
Chuck 'Sisyphus' Swiger | [EMAIL PROTECTED] | Bad cop! No Donut.
------------------------+-------------------+--------------------
I know that you are an optimist if you think I am a pessimist....
------------------------------
From: "Ken Acker" <[EMAIL PROTECTED]>
Subject: Re: Changed from Storm to Mandrake-have some new questions
Date: Tue, 4 Jul 2000 16:07:38 -0300
I'm not sure if this is an option, but you could look into getting the new
version of Mandrake (7.1). It has a new boot loader, it fixes that annoying
Num Lock problem, and sets up your wheel mouse better.
As for the ACDSee clone, you could try Compupic
( http://www.photodex.com/ ). It has many, if not all, of the functions that
ACDSee has and comes in a Linux version.
-Ken
------------------------------
** 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.setup) 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-Setup Digest
******************************
