Linux-Setup Digest #841, Volume #20 Fri, 16 Mar 01 10:13:09 EST
Contents:
shell scripts and cron (Keith O'Connell)
Re: Install Linux on SCSI HD with Solaris OS (Rod Smith)
Problems with installing new harddisk on RedHat Linux 6.0 ("Roman Badertscher")
RAID problems (long), someone wrote a new superblock on an old raid
[SOLUTION] K2.4.2 usb-storage problems: mount: /dev/sda1 has wrong major or minor
number (Geir Harris Hedemark)
Re: USB ("Sivagurunathan U.")
Re: shell scripts and cron (Michael Heiming)
Re: shell scripts and cron (Michael Heiming)
Re: Problems with installing new harddisk on RedHat Linux 6.0 ("Eric")
quota on / file system (Dick Visser)
Re: uninstall (Adam Hulcoop)
Windows size (The_saint)
Dual booting (The_saint)
Re: Burning iso files for RH7 (The_saint)
Re: X and MS Intellimouse problem (mouse won't move from center) (The_saint)
Re: shell scripts and cron (Keith O'Connell)
Re: uninstall ([EMAIL PROTECTED])
----------------------------------------------------------------------------
From: Keith O'Connell <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: shell scripts and cron
Date: Fri, 16 Mar 2001 13:20:46 GMT
I am a novice, but was pleased with myself when I felt that I had got
the hang of elementary shell scripts and cron jobs. I wrote a two line
script that backed up my key directories to a tar file that used the
date in the file name. I ran it and it worked perfectly. (Well done
me!!)
I read about cron and set the script to run at 2300hrs each day. and it
ran just when I wanted it to (double well done me!!)
........but (you all knew there was going to be a "but"!)
When I looked at the resultant tar file it was way to small. It had not
tarred everything.
Why would a script work perfectly when run from the command line, but
not from a cron job?
ss_bkup
~~~~~~~
#!/bin/bash
tar cvpf /store/linux/aragorn_`date +%y%m%d`.tar /etc /home /root
I was so pleased with myself as well (that'll teach me!!)
Anyone help?
--
========================================================================
[EMAIL PROTECTED] (Keith O'Connell)
========================================================================
------------------------------
From: [EMAIL PROTECTED] (Rod Smith)
Subject: Re: Install Linux on SCSI HD with Solaris OS
Date: Fri, 16 Mar 2001 13:31:42 -0000
Reply-To: [EMAIL PROTECTED]
[Posted and mailed]
In article <[EMAIL PROTECTED]>,
songjie <[EMAIL PROTECTED]> writes:
> Hi,
>
> I have a PC with only one SCSI controller, one SCSI CDROM and two SCSI
> hard disks. The first SCSI HD (ID : 0) has been installed with NT 4.0
> and Solaris OS. The second SCSI HD(ID: 2) is installed with Redhat Linux
> 6.2. The CDROM has an SCSI ID of 4. The ID of the controller is 7.
>
> When I install the Linux, I removed the First SCSI HD of NT & SOLARIS.
> So the lilo is trying to boot fom /dev/sda and root is /dev/sda5. Then
> I connected the first HD together. With the SCSI controller, I config it
> is boot from the SCSI ID of 2. But the lilo cannot boot corectly. It
> prompted like "VFS cannot mount root from /dev/sda 08:02 ....".
[Caveat: I have no experience with Solaris, so if there are
Solaris-specific gotchas, I don't know about them.]
You're trying to do it the hard way. It's **MUCH** easier to just leave
BOTH hard disks in place and install Linux on the second one. Linux has
no problems booting from the second hard disk.
> So My question is :
>
> 1) Can I install Linux at the second SCSI HD without damage the solaris
> partition on the fist SCSI HD?
Yes. Depending upon how you're selecting between NT and Solaris, you
might want to keep using it and add Linux to the mix; or you might want
to install LILO in the MBR of the first disk and use it to select which
OS to boot, or use a Linux boot floppy or LILO on a floppy when you boot
Linux.
> 2) Now I can boot from a floppy , but can I boot from /dev/sdb directly?
It's possible in general, but you've got a pretty weird setup by now.
You could probably recover your existing system, but it's probably
easier to just re-install from scratch, since you haven't done any
configuration of it yet.
> 3) How to install the linux to ensure it can be boot from the second
> SCSI HD?
There's nothing terribly special about this. The critical question is
how you're currently selecting between your two OSs, and how you want to
add Linux to that mix. If you want to have LILO replace whatever boot
loader you're currently using, install LILO in the MBR of the first disk
(this is the default for most distributions). If you've got some other
boot loader right now (like System Commander, BootMagic, or even NT's OS
Loader), you should install LILO in the boot sector of the Linux boot
partition and add an entry in your current boot loader to boot from that
partition. (In the case of OS Loader, there are extra hoops you'll have
to jump through to get this to work.)
--
Rod Smith, [EMAIL PROTECTED]
http://www.rodsbooks.com
Author of books on Linux & multi-OS configuration
------------------------------
From: "Roman Badertscher" <[EMAIL PROTECTED]>
Subject: Problems with installing new harddisk on RedHat Linux 6.0
Date: Fri, 16 Mar 2001 14:37:22 +0100
Hello
I recently bought a new hard disk (Seagate ST336705LW, LVD SCSI, 36 GB). My
system consists already of 2 9GB IBM LVD Disks, the first with Windows NT,
the second with RedHat Linux 6.0 (I know I should update). The Disks are on
a Adaptec AIC 7890 controller (aic7xxx module in Linux kernel). I use LILO
to boot to Windows or Linux.
I installed the new disk and then booted into Windows, made a 24 GB
partition and fomatted it. No problems there.
Then I rebooted into Linux, but when it came to recognize the new disk all I
got were timeouts on the SCSI bus. I had to physically disconnect the new
disk, then i could boot into Linux again. After searching a bit I thought it
might be the aic7xxx driver in Linux, because in Windows it worked
perfectly, RedHat was updated some times already, and the disk i a fairly
new type. I then tried to install a new version of this driver, but I
failed. What is left that i can't boot into Linux anymore, even if the new
disk is disconnected. I can boot with the rescue disk, and mount all the
partitions on the Linux disk. What I get when I try to but the normal way is
an error message from kmod, that it could not execute /sbin/modprobe -s -k
<something>, with error number 2, and that it xould not mount root partition
on 08:17 (whatever that means).
So thats the current state of my box. I'm interested what the kmod error
message is, and how I can fix it, if possible.
I will try to update Linux to RedHat 7.0 . Maybe that helps.
Thanks for your help
Roman
------------------------------
From: <[EMAIL PROTECTED]>
Subject: RAID problems (long), someone wrote a new superblock on an old raid
Date: 16 Mar 2001 13:26:27 GMT
Hi,
I have a little problem here I was wondering if anyone could help me with.
I went out on a work order yesterday and I ran into something I'm not quite
sure how to deal with.
Basically, someone tripped on a network cable and jerked the network card
out of the PCI bus on a dell linux server, all 4 disks in a raid 5 were
marked bad I believe.
After the raid would not come back up, the administrator upgraded the
superblocks to raid .90 compatible superblocks. I'm not sure if this worked
and the disks were still marked bad, or if it didn't work, but next the
administrator did a mkraid --force on the partition using the new raidtools.
Currently the box comes up under .90 with all disks complaining of a bad
raid superblock checksum. I do believe there is a valid .90 superblock
somewhere because mkraid --upgrade says that an upgrade is not required.
My questions:
Has any of the data on the disk been wiped to make room for the new
superblock? ie. does doing a mkraid simply write over the space of
an old raid superblock, or did the size or the format of the actual
data on the disk change?
My thoughts are to write a new .90 valid superblock to the disks and
see if there is a somewhat valid filesystem left. Do you agree that
no additional data would be lost in this circumstance? If that doesn't
work then try writing an old superblock and see what the fs looks like.
Any suggestions on a course of action to take care of this? The owner
of this box is considering sending it off to a data reconstruction
company, anything I do needs to not further compromise any more data.
Is there a way just recalculate/write checksums on the superblocks so
the drives look good again?
Last thing, if a raid superblock checksum is bad on all 4 disks, what was
likely the cause? Software misconfiguration, or something physical?
Thank you very much for any help.
--Dustin
------------------------------
From: Geir Harris Hedemark <[EMAIL PROTECTED]>
Subject: [SOLUTION] K2.4.2 usb-storage problems: mount: /dev/sda1 has wrong major or
minor number
Date: 16 Mar 2001 14:49:40 +0100
[EMAIL PROTECTED] writes:
> >>mount -t vfat /dev/sda1 /mnt
> > mount: /dev/sda1 has wrong major or minor number
> Have you tried using /dev/sda4 instead of /dev/sda1 ?
Wrong tree. :)
1) I needed to load usb-uhci. This will take care of the major/minor problem.
2) I needed to define my USB device (SDDR09) in the kernel config files.
3) It takes about 1 min to mount this USB device.
4) The kernel driver has a memory leak, and will only allow a certain
amount of data transfer before making the kernel panic.
More info on http://www.theneals.net/linux/SDDR09-HOWTO
Geir
------------------------------
From: "Sivagurunathan U." <[EMAIL PROTECTED]>
Subject: Re: USB
Date: Fri, 16 Mar 2001 13:54:31 GMT
==============DFF839DC7A1B6017801E59B9
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Hi! Guys,
I have the same motherboard in a Compaq computer. This machine loads every
thing at the time of installation.(I tried Caldera edesktop 2.4, and SUSE 6.4)
But when it is starting Linux and loading modules it says VIA82x6.. module not
loaded.
But for me USB keyboard PS2 mouse Vanta video all works fine. But D-Link
DFE530TX ethernet is not detecting. Though it is a listed HW in all of the above
distributions.
When I tried with different LINUX distributions site, all do not say
VIA82X686a/b motherboard supported.
Hence I think it is rather another HW not supported!!
If anybody knows a workaround I will also appreciate.
Regards
Siva
Zayin Krige wrote:
> I disabled IRQ for usb in the BIOS, but the same thing still happens...
>
> "Steve Martin" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > Gregory Davis wrote:
> >
> > > > if i unplug the mouse and replug it in, the message repeats.
> > > >
> > > > as far as i can see, my network card, video card and usb hub ar
> sharing
> > > > the same irq. is this maybe th cause. any help will be appreciated
> >
> > > Maybe. In that case, try specifying IRQs for each in the BIOS. If they
> > > already are, try having the BIOS set them up automatically. Try setting
> > > your BIOS for PnP compatible OS and for PnP OS incompatible.
> >
> > Yes, I'd guess that the shared IRQ is probably the problem.
> > Get them each their own IRQ (whether by jumper-setting,
> > BIOS setting, whatever), then see if Linux detects them all
> > properly by looking at /proc/interrupts.
> >
> > Good luck.
==============DFF839DC7A1B6017801E59B9
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit
<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
Hi! Guys,
<p> I have the same motherboard in a Compaq computer.
This machine loads every thing at the time of installation.(I tried Caldera
edesktop 2.4, and SUSE 6.4) But when it is starting Linux and loading modules
it says VIA82x6.. module not loaded.
<p> But for me USB keyboard PS2 mouse Vanta video all
works fine. But D-Link DFE530TX ethernet is not detecting. Though it is
a listed HW in all of the above distributions.
<p> When I tried with different LINUX distributions site,
all <u>do not say</u> VIA82X686a/b motherboard <u>supported</u>.
<p> Hence I think it is rather another HW not supported!!
<p> If anybody knows a workaround I will also appreciate.
<p>Regards
<p>Siva
<br>
<p>Zayin Krige wrote:
<blockquote TYPE=CITE>I disabled IRQ for usb in the BIOS, but the same
thing still happens...
<p>"Steve Martin" <[EMAIL PROTECTED]> wrote in message
<br><a
href="news:[EMAIL PROTECTED]">news:[EMAIL PROTECTED]</a>...
<br>> Gregory Davis wrote:
<br>>
<br>> > > if i unplug the mouse and replug it in, the message repeats.
<br>> > >
<br>> > > as far as i can see, my network card, video card and usb hub
ar
<br>sharing
<br>> > > the same irq. is this maybe th cause. any help will be appreciated
<br>>
<br>> > Maybe. In that case, try specifying IRQs for each in the
BIOS. If they
<br>> > already are, try having the BIOS set them up automatically.
Try setting
<br>> > your BIOS for PnP compatible OS and for PnP OS incompatible.
<br>>
<br>> Yes, I'd guess that the shared IRQ is probably the problem.
<br>> Get them each their own IRQ (whether by jumper-setting,
<br>> BIOS setting, whatever), then see if Linux detects them all
<br>> properly by looking at /proc/interrupts.
<br>>
<br>> Good luck.</blockquote>
</html>
==============DFF839DC7A1B6017801E59B9==
------------------------------
Date: Fri, 16 Mar 2001 15:02:33 +0100
From: Michael Heiming <[EMAIL PROTECTED]>
Subject: Re: shell scripts and cron
Keith O'Connell wrote:
>
> I am a novice, but was pleased with myself when I felt that I had got
> the hang of elementary shell scripts and cron jobs. I wrote a two line
> script that backed up my key directories to a tar file that used the
> date in the file name. I ran it and it worked perfectly. (Well done
> me!!)
>
> I read about cron and set the script to run at 2300hrs each day. and it
> ran just when I wanted it to (double well done me!!)
>
> ........but (you all knew there was going to be a "but"!)
>
> When I looked at the resultant tar file it was way to small. It had not
> tarred everything.
>
> Why would a script work perfectly when run from the command line, but
> not from a cron job?
In which crontab did you set this and how? Perhaps it's not executed
with the proper rights?
>
> ss_bkup
> ~~~~~~~
> #!/bin/bash
> tar cvpf /store/linux/aragorn_`date +%y%m%d`.tar /etc /home /root
I would suggest using something like:
#!/bin/bash
DATE=`date +%y%m%d`
tar -cvIpf test_$DATE.tar.bz2 /etc /home /root > tar.log 2>&1
Too see what went wrong.
(The -I option will compress your data, to save some space)
Good luck
Michael Heiming
>
> I was so pleased with myself as well (that'll teach me!!)
>
> Anyone help?
>
> --
> ------------------------------------------------------------------------
> [EMAIL PROTECTED] (Keith O'Connell)
> ------------------------------------------------------------------------
------------------------------
Date: Fri, 16 Mar 2001 15:04:38 +0100
From: Michael Heiming <[EMAIL PROTECTED]>
Subject: Re: shell scripts and cron
Keith O'Connell wrote:
>
> I am a novice, but was pleased with myself when I felt that I had got
> the hang of elementary shell scripts and cron jobs. I wrote a two line
> script that backed up my key directories to a tar file that used the
> date in the file name. I ran it and it worked perfectly. (Well done
> me!!)
>
> I read about cron and set the script to run at 2300hrs each day. and it
> ran just when I wanted it to (double well done me!!)
>
> ........but (you all knew there was going to be a "but"!)
>
> When I looked at the resultant tar file it was way to small. It had not
> tarred everything.
>
> Why would a script work perfectly when run from the command line, but
> not from a cron job?
In which crontab did you set this and how? Perhaps it's not executed
with the proper rights?
>
> ss_bkup
> ~~~~~~~
> #!/bin/bash
> tar cvpf /store/linux/aragorn_`date +%y%m%d`.tar /etc /home /root
I would suggest using something like:
#!/bin/bash
DATE=`date +%y%m%d`
tar -cvIpf test_$DATE.tar.bz2 /etc /home /root > /tmp/tar.log 2>&1
Too see what went wrong.
(The -I option will compress your data, to save some space)
Good luck
Michael Heiming
>
> I was so pleased with myself as well (that'll teach me!!)
>
> Anyone help?
>
> --
> ------------------------------------------------------------------------
> [EMAIL PROTECTED] (Keith O'Connell)
> ------------------------------------------------------------------------
------------------------------
From: "Eric" <[EMAIL PROTECTED]>
Subject: Re: Problems with installing new harddisk on RedHat Linux 6.0
Date: Fri, 16 Mar 2001 15:32:44 +0100
> I recently bought a new hard disk (Seagate ST336705LW, LVD SCSI, 36 GB).
My
> system consists already of 2 9GB IBM LVD Disks, the first with Windows NT,
> the second with RedHat Linux 6.0 (I know I should update). The Disks are
on
> a Adaptec AIC 7890 controller (aic7xxx module in Linux kernel). I use LILO
> to boot to Windows or Linux.
> I installed the new disk and then booted into Windows, made a 24 GB
> partition and fomatted it. No problems there.
> Then I rebooted into Linux, but when it came to recognize the new disk all
I
> got were timeouts on the SCSI bus. I had to physically disconnect the new
> disk, then i could boot into Linux again.
terminated wrong?
perhaps a duplicate SCSI ID?
> After searching a bit I thought it
> might be the aic7xxx driver in Linux, because in Windows it worked
> perfectly,
That doesn't mean much.
> RedHat was updated some times already, and the disk i a fairly
> new type. I then tried to install a new version of this driver, but I
> failed.
How did you fail?
> What is left that i can't boot into Linux anymore, even if the new
> disk is disconnected. I can boot with the rescue disk, and mount all the
> partitions on the Linux disk. What I get when I try to but the normal way
is
> an error message from kmod, that it could not execute /sbin/modprobe -s -k
> <something>,
<something> ??
surely it mentioned something else there.
> with error number 2, and that it xould not mount root partition
> on 08:17 (whatever that means).
08:17 = sdb2
> So thats the current state of my box. I'm interested what the kmod error
> message is, and how I can fix it, if possible.
Show us the exact error message
> I will try to update Linux to RedHat 7.0 . Maybe that helps.
perhaps it will.
Eric
------------------------------
From: [EMAIL PROTECTED] (Dick Visser)
Subject: quota on / file system
Date: 16 Mar 2001 14:46:58 GMT
I have a linux box with just one / partition (and swap o.c.). I added
/quota.grp and /quota.usr, that works OK. But when i reboot the box,
it cannot unmount the fs, so when it comes up it starts fscking.
Any ideas how I make my box unmount its /-fs before shutting down?
--
dick visser
amsterdam
the netherlands
------------------------------
From: Adam Hulcoop <[EMAIL PROTECTED]>
Subject: Re: uninstall
Date: Fri, 16 Mar 2001 14:42:03 GMT
[EMAIL PROTECTED] wrote:
>
> Adam Hulcoop <[EMAIL PROTECTED]> wrote:
> > or how about rm -Rf / :)
>
> hummm... could work, but I think is faster (and maybe easyer)
> simply remove the partition...
> BTW, I saw once someone (the PHBIT) do something like that on
> an old DOS box... boy, it was shocked when the machine froze!
>
> Davide
Ya .. hehe I realize that its not really a *great* way of doing it.
Its just an old joke I guess. Like in the old DOS days when you told
someone to type:
echo y | format c:
thats probly what you are refering to right?
adam
--
Adam Hulcoop
------------------------------
From: The_saint <[EMAIL PROTECTED]>
Subject: Windows size
Date: Fri, 16 Mar 2001 22:57:49 +0800
Hi NG,
How do I set all windows size to fit into the desktop?
I've only a poor 800x600 TFT display and many times the button are beyond
the screen size, so when I move the pointer I get the KDE taskbar popping
up.
------------------------------
From: The_saint <[EMAIL PROTECTED]>
Subject: Dual booting
Date: Fri, 16 Mar 2001 22:58:36 +0800
Hi NG,
I'd like to set a dual booting from the M$dos for Linux and WinME.
How to write the autoexec.bat?
What should I copy onto c:\ (windows) drive?
Thanks in advance
------------------------------
From: The_saint <[EMAIL PROTECTED]>
Subject: Re: Burning iso files for RH7
Date: Fri, 16 Mar 2001 22:58:06 +0800
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] says...
> What do I need to do to burn the CDs under Windows using CdrWin or Nero
> ?
>
>
Nero will do it, just choose to burn a disk image and point to that file.
------------------------------
From: The_saint <[EMAIL PROTECTED]>
Subject: Re: X and MS Intellimouse problem (mouse won't move from center)
Date: Fri, 16 Mar 2001 22:57:56 +0800
In article <[EMAIL PROTECTED]>, [EMAIL PROTECTED] says...
> Thanks for the help. It does not work in console mode either. Strange that it
> does work fine during installation, but after installation, it is gone.
>
I've similar problem, since I've tried to set my mouse as an intellimouse.
BTW, is there any touchpad enhancement for my laptop?
------------------------------
From: Keith O'Connell <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: Re: shell scripts and cron
Date: Fri, 16 Mar 2001 14:50:34 GMT
Michael
> In which crontab did you set this and how? Perhaps it's not executed
> with the proper rights?
On the last occasion, it was as root
> I would suggest using something like:
>
> #!/bin/bash
> DATE=`date +%y%m%d`
> tar -cvIpf test_$DATE.tar.bz2 /etc /home /root > tar.log 2>&1
I'll give this a try and see what is spit out
Keith
--
========================================================================
[EMAIL PROTECTED] (Keith O'Connell)
========================================================================
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: uninstall
Date: 16 Mar 2001 15:08:47 GMT
Adam Hulcoop <[EMAIL PROTECTED]> wrote:
> echo y | format c:
> thats probly what you are refering to right?
Worse than that, he did a wonderful deltree C:\
Davide
------------------------------
** 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
******************************
