Linux-Setup Digest #54, Volume #19 Sun, 2 Jul 00 13:13:17 EDT
Contents:
automount does nothing ("Andrew E. Schulman")
Re: Moving /home ("Jeffrey Lacy")
Re: Why is my harddisk so slow? (Hal Burgiss)
How to configure Motorola ModemSurf 56K ("qxd")
Re: Destroyed VFAT Filesystem (Svend Olaf Mikkelsen)
Network printing ([EMAIL PROTECTED])
Re: network question (Michael Nadler)
Large HD's ("Scott Wood")
Re: 1024 cylinder limit info please? (Ken Knecht)
Re: Moving /home (ray)
Help! ([EMAIL PROTECTED])
HLP Newbie: kppp and password (Romano)
----------------------------------------------------------------------------
From: "Andrew E. Schulman" <[EMAIL PROTECTED]>
Subject: automount does nothing
Date: Sun, 02 Jul 2000 11:13:09 -0400
I've installed the latest version of autofs (can't remember the version
number) in Debian. When I boot, autofs appears to start the automount
daemons, but then no automounting takes place. When I try the automount
command directly, also nothing happens, and "ps aux" shows no automount
daemons running. In fact no matter how I call automount, even with bad
parameters, it just returns without error, e.g.
$ automount bogus bogus bogus || echo error
$
automount just appears to be doing nothing. Does anyone know why? Any
help appreciated.
For reference here are my auto.master and auto.misc files, although I
don't think these are the problem:
$ cat /etc/auto.master
/cdrom /etc/auto.misc
/zip /etc/auto.misc
/floppy /etc/auto.misc
$ cat /etc/auto.misc
cdrom -fstype=iso9660,defaults,users,ro :/dev/hdc
zip -fstype=vfat,defaults,users :/dev/hdd4
floppy -fstype=vfat,defaults,users :/dev/fd0
------------------------------
From: "Jeffrey Lacy" <[EMAIL PROTECTED]>
Subject: Re: Moving /home
Date: Sun, 02 Jul 2000 15:20:16 GMT
OK. Thanks a lot! :-D
"Craig Kelley" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> "Jeffrey Lacy" <[EMAIL PROTECTED]> writes:
>
> > Hi. I am running Redhat 6.2. I will be upgrading my harddrive soon,
and I
> > need to know how to move /home to my new harddrive. How can I do this?
> > Thanks in advance!
>
> mnt /dev/hdblah /mnt/newdisk
> cd /home
> cp -a * /mnt/newdisk
> cd /
> umount /home
> umount /mnt/newdisk
> mount /dev/hdblah /home
>
> --
> The wheel is turning but the hamster is dead.
> Craig Kelley -- [EMAIL PROTECTED]
> http://www.isu.edu/~kellcrai finger [EMAIL PROTECTED] for PGP block
------------------------------
From: [EMAIL PROTECTED] (Hal Burgiss)
Subject: Re: Why is my harddisk so slow?
Reply-To: [EMAIL PROTECTED]
Date: Sun, 02 Jul 2000 15:36:41 GMT
On Sun, 02 Jul 2000 12:00:04 +0200, Cliff Pennock <[EMAIL PROTECTED]> wrote:
>I've posted this on other newsgroups before, but nobody seems to be able
>to answer my question so I'll try in this newsgroup too...
>
>Hardware: Celeron500, 128 Mb Samsung 15.3Gb UDMA mode4 HDD, SiS5513 IDE
> controller.
>Kernel : 2.2.16, patched with ide-2.2.16.20000630
>BIOS : Detects the harddisk as PIO4 and UDMA4
>hdparm : 3.9-1, params: -d1 -m16 -c1 -A1 (even tried -X66)
> (for some reason, after installing 3.9-1 my hdparm
> manpages were gone)
>
>Enabled SiS5513 support in the kernel, including CONFIG_IDEDMA_AUTO
>(yes, and even recompiled and installed the new kernel :) ).
Have you played around with some of the other kernel config options
included with the ide patch? Unfortunately, some of these are not well
documented AFAICT. I had a problem with a drive that would not keep DMA
set, it would turn it off automagically. I know this isn't the same
problem, but I eventually solved it with that patch and trial and error
on various IDE config settings.
--
Hal B
[EMAIL PROTECTED]
--
------------------------------
From: "qxd" <[EMAIL PROTECTED]>
Subject: How to configure Motorola ModemSurf 56K
Date: Sun, 2 Jul 2000 08:33:37 -0700
Hi all,
I have a Pentium 3 computer with dual boot, Window 98 & Linux-Mandrake 7.0.
The computer reconized the serial port, com3 at 0x3E8 (no irq info), but not
the modem itself. From a book, it's suggested to add a couple of line to
RC.SERIAL to account for IRQ and ports if they are non-standard. I believe
my modem is not a WinModem and it should work with Linux. This modem is set
at IRQ 5 for window application.
Can you give me explicit instructions on how to set up this modem? Thank
you.
Henry
------------------------------
From: [EMAIL PROTECTED] (Svend Olaf Mikkelsen)
Subject: Re: Destroyed VFAT Filesystem
Date: Sun, 02 Jul 2000 15:59:57 GMT
[EMAIL PROTECTED] wrote:
>The machine has two hard drives, the first used only for operating
>systems (currently Windows 98 and Linux) and the second used only for
>data. The second drive is partitioned in MSDOS using a single extended
>partition with various logical partitions.
>
>The Linux is the VALinux distribution 6.2.1 (RedHat 6.2) and the kernel
>is 2.2.14-5.0.14 booted by LILO or Loadlin. I have reinstalled this
>distribution several times with no trouble as I have adjusted my
>partitioning scheme on the first hard drive.
>
>Now that the partitioning scheme on the first hard drive appears
>reasonable, I am moving on to other issues. I added the following
>lines to /etc/fstab.
>
>/dev/hdb5 /home/windows vfat defaults 0 0
>/dev/hdb6 /home/music vfat defaults 0 0
>
>The second of the two lines worked fine and I could play the MP3s on
>the partition fine in Linux.
>
>The first of the two lines appeared to destroy the formatting of the
>first logical partition of the second hard drive (/dev/hdb5). A couple
>of days ago, Linux went so far as to destroy the partition table by
>adding an additional logical drive to the second hard drive. (It
>looked to Windows as though the 27Gb HD had logical drives totaling
>31Gb.) I spent all day yesterday repartitioning, reformatting, and
>surface scanning the entire second hard drive (not to mention restoring
>the data files). Thereafter, booting Linux destroyed the formatting
>of /dev/hdb5 again (but not the partition table), and I found I could
>reformat the partition in Windows and restore the data.
>
>I removed all lines from /etc/fstab that mount FAT32 partitions.
>Nevertheless, now I find that simply booting Linux destroys the
>formatting of the first partition of the second hard drive, even
>without the lines in /etc/fstab that I thought were causing the
>problem. (Needless to say, the second hard drive and partitioning
>scheme work fine in Windows, and I have no interest in partitioning or
>formatting them in Linux.)
>
>Does anyone have any idea what is going on? I will very much
>appreciate any help.
>
>(I am new to using Linux and have spent the last several months
>installing various distributions on two different machines and have
>done a lot of reading. Now I'm wondering if this has been a waste of
>time (apart from the knowledge gained about Unix), if Linux can't do
>something as basic as booting without destroying filesystems. Is Free
>BSD more reliable?)
>
>Bill
This could be a case of the
"if an extended partition type 0F exists, and the last logical
partition in an extended partition known by DOS is non DOS, no
extended partition type 05 can exist"
Windows bug. Windows typically will write the first logical partition
on disk number two at a wrong location.
If the type of the extended partition on disk number 1 is 05, you can
change it to 0F using the Linux fdisk t command. When you are sure you
did the change correct, write the partition table using the w command.
--
Svend Olaf
------------------------------
From: [EMAIL PROTECTED]
Subject: Network printing
Date: Sun, 02 Jul 2000 15:54:12 GMT
I have a Epson Laser printer on my network with IP 192.168.1.26 and the
IP of my Red Hat Linux 6.0 machine is 192.168.1.2. So how do I print
from this machine to the printer.
I know that you may answer to use printtool but what should I select,
ofcourse not a local printer, then a remote Unix printer or SMB/Win...
printer. Please do try to explain assuming other parameters. I would be
very greatful since I require it very desparately.
Customised configuration links can be given but nothing like your
explanation.
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: Michael Nadler <[EMAIL PROTECTED]>
Subject: Re: network question
Date: Sun, 02 Jul 2000 09:08:52 -0700
Arne Bohle wrote:
> Sorry for jumping into this discussion, but I also have some problems at the
> moment configuring my Linux box (Red Hat 6.0) in a local environment I am
> setting up.
>
> The local network will be a Win98 and a Linux node, connected to a hub, both
> nodes running a Linksys Combo EtherPCI LAN Card.
>
> Unfortunately, I tried out different network drivers to try to get the PCI
> card on my Linux node working (since Linksys did not mention Linux specific
> drivers in their set-up guide), which now has resulted in my Linux node not
> booting. It hangs when "Bringing up interface eth0".
>
> Later last night, after looking through Linksys' WEB site, I realised what
> driver I should have used (Tulip).
>
> My question now is, how do I restore my network settings so I can load the
> correct driver ? I have tried my rescue disk, but it does not seem to work
> (or I am doing it wrongly).
>
Does it hang forever? It ought to eventually time out and continue with boot
up. Then, you could "modprobe tulip" to get the driver loaded & then ifconfig
the interface, to verify that all is well?
I run Slack7 so am not really certain what you do with RedHat. Seems like
you'll have to boot up in single user mode, so that you can edit your
configuration prior to the network hanging.
The files that changed in Slack 7 when I changed NICs was
/etc/rc.d/rc.netdevice -- this script is called from rc.modules inside rc.S
(single-user init) to cause the right drivers to get loaded.
I'd search around in the /etc/rc.d (or /etc/init.d in RedHat?) directories
looking for "modprobe" commands. Perhaps you'll run into the spot where
network driver modules are loaded & can fix that up accordingly.
>
> Any ideas ?
>
> Thanks in advance,
> Arne
>
> Michael Nadler <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > sylvain hutchison wrote:
> >
> > > Hello, I am loking for a driver for my network card which is a 3Com
> > > Etherlink 10/100 PCI NIC (3C905C-TX), I've been trying to phone Dell for
> > > the last fucking 30 min, they are getting on my nerves, and I can't find
> > > anything on their web page, so I was wondering if any of you knew where
> > > I could download this driver!!! BTW I've got Red Hat 6.1, I don't know
> > > if that makes a difference!!!!!
> > >
> > > Thanks a lot,
> > >
> > > Sly.
> >
> > I wonder if this card is supported by the same driver as the 3c905b --
> > 3c59x.o -- have you tried that module? It was in my Slack7 distro, might
> > be in yours. The Ethernet-HOWTO tells how to find it otherwise.
> >
------------------------------
From: "Scott Wood" <[EMAIL PROTECTED]>
Subject: Large HD's
Date: Sun, 02 Jul 2000 16:14:46 GMT
I just installed a 20gig Maxtor HD, but ended up with 8 and a half gig of
space. I see that the cylinders, sectors, heads settings add up to 8.5 gig.
How do I get the remaining 12gig? Thanks, --sw
------------------------------
From: [EMAIL PROTECTED] (Ken Knecht)
Subject: Re: 1024 cylinder limit info please?
Date: 2 Jul 2000 16:21:58 GMT
[EMAIL PROTECTED] (Homer Jay) wrote in
<8jlda7$dht$[EMAIL PROTECTED]>:
>You can usually boot from a floppy without worrying about the
>1024 restriction. The BIOS is sometimes incapable of booting
>an OS beyond the 1024th cylinder, but a floppy circumvents
>that pretty well.
>
>The 1024 limit works out to 504MB. Put your OS(s), or at
>least the boot info, within the first 504MB and you'll have
>no problem. Some BIOSs can handle booting beyond that,
>however. I guess the limit originated because the original
>scheme for ATA disks imposed a 504 MB limit. BIOSs got around
>that with several different "translation" schemes. (SCSI is a
>little more intelligible, handling the translation in the
>drive electronics.) From the lilo docs:
>
>The most common BIOS restrictions that affect LILO are the
>limitation to two hard disks and the inability to access more
>than 1024 cylinders per disk. LILO can detect both
>conditions, but in order to work around the underlying
>problems, manual intervention is necessary.
>
>The drive limit does not exist in every BIOS. Some modern
>motherboards and disk controllers are equipped with a BIOS
>that supports more (typically four) disk drives. When
>attempting to access the third, fourth, etc. drive, LILO
>prints a warning message but continues. Unless the BIOS
>really supports more than two drives, the system will _not_
>be able to boot in that case.*
>
> * However, if only "unimportant" parts of the system are
> located on the
> "high" drives, some functionality may be available.
>
>The cylinder limit is a very common problem with IDE disks.
>There, the number of cylinders may already exceed 1024 if the
>drive has a capacity of more than 504 MB. Many SCSI driver
>BIOSes present the disk geometry in a way that makes the
>limit occur near 1 GB. Modern disk controllers may even push
>the limit up to about 8 GB. All cylinders beyond the 1024th
>are inaccessible for the BIOS. LILO detects this problem and
>aborts the installation (unless the LINEAR option is used,
>see section "Global options").
>... [From Global options section:]
> LINEAR Generate linear sector addresses instead of
> sector/head/cylinder
> addresses. Linear addresses are translated at run time
> and do not depend on disk geometry. Note that boot disks
> may not be portable if LINEAR is used, because the BIOS
> service to determine the disk geometry does not work
> reliably for floppy disks. When using LINEAR with large
> disks, /sbin/lilo may generate references to inaccessible
> disk areas (see section "BIOS restrictions"), because 3D
> sector addresses are not known before boot time. LINEAR
> may conflict with COMPACT, see section "Other problems".
>
>Using Linear sounds like the choice of last resort to me. I
>recommend putting a 100-200 MB root partition first, then
>your DOS partition, making it as big as you need. Then you
>can use the rest of the disk for /usr /var, whatever scheme
>you want. If "/" (root) is within the first 500MB then
>they'll be no problem. (E.g.: I have a 13GB disk and my BIOS
>only supports up to 2GB, but it works b/c root is below
>504MB.) Just for the pedantic:
>The restriction to 500 MB resulted from the original BIOS
>disk interface, INT13 (originally designed for floppies of
>all things), using 10 bits to represent the cylinder number
>(2^10=1024) and 6 bits for the number of sectors/track
>(2^6=64). Heads, real heads, are limited to 16 to avoid
>incompatiblities with a disk controller from the original AT
>from 1984! Bytes/sector is 0.5K by convention. 0.5KB * 16 *
>63 * 1024 = 504MB (63 because sectors always number from 1)
>This would be a hard limit but new BIOSs use phantom heads,
>etc. to make the max disk size larger. Many allow larger
>cylinder number, etc. This makes me wonder, if the _BIOS_
>supports access to cylinders beyone 1024, why can't lilo,
>which depends upon the BIOS? Or, is that what the LINEAR
>option is for? Anyone know?
>
Thanks much. Very educational!
According to the manual, my Linux (TurboLinux 6.0) gives the
Linear option when it is installed. However, the manual says
this is only for SCSI HDs, as does the LILO section of a Linux
encyclopedia I have. I printed out the Large Disk Howto
this morning and will read it later - should be some more
info in there - if I can understand it!<g> Likely I'll end
up just redoing the HD with a 10M DOS partition and eliminate
the problem that way.
I still haven't figured out how to add another Linux to the HD
later as I intended (I have a recent Red Hat book coming with a
recent (I hope) Linux install CD in it) so I can play with it.
Guess I'll have to use a floppy to boot it.
Ken
------------------------------
From: ray <[EMAIL PROTECTED]>
Subject: Re: Moving /home
Date: Sun, 02 Jul 2000 16:45:45 GMT
Jeffrey Lacy wrote:
> OK. Thanks a lot! :-D
>
> "Craig Kelley" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > "Jeffrey Lacy" <[EMAIL PROTECTED]> writes:
> >
> > > Hi. I am running Redhat 6.2. I will be upgrading my harddrive soon,
> and I
> > > need to know how to move /home to my new harddrive. How can I do this?
> > > Thanks in advance!
> >
> > mnt /dev/hdblah /mnt/newdisk
> > cd /home
> > cp -a * /mnt/newdisk
> > cd /
> > umount /home
> > umount /mnt/newdisk
> > mount /dev/hdblah /home
> >
> > --
> > The wheel is turning but the hamster is dead.
> > Craig Kelley -- [EMAIL PROTECTED]
> > http://www.isu.edu/~kellcrai finger [EMAIL PROTECTED] for PGP block
Yup, that'll do it. Now, also, might be a good time to think about LVM,
now supported in the 2.4.0
series kernels. I have it running, and, man, being able to resize a "drive" is
a wonderful thing.
No wonder this has been in AIX and HP/UX (and others) for years. The issue is
not "How big does /home need to be?". it is "How big will it EVER need to
be?". LVM makes that issue moot.
--
Ray R. Jones
Linux 2.4.0-test2
[EMAIL PROTECTED]
HTTP://gordo.penguinpowered.com
------------------------------
From: [EMAIL PROTECTED]
Subject: Help!
Date: Sun, 02 Jul 2000 16:41:25 GMT
Hi there,
I'm new to linux. I am installing OPEN_Linux2.2 on Compaq presario 7222
(CPU P133, Mem 48m, HD 1.6g newly formatted). I follow the instruction
and everything
seems OK. (boot with lizard install, then put modules disk, then
loading kernel OK
Booting kernel OK
Automatic hardware detection
Probing for PCMICIA cards none
Probing for SCSI hosts none
Probing for ATAPI CDROMS OK
Probing for other CDRDMS none
Probing for Ethernet cards OK
Automatic install source detection
Installation from CDRM OK
PACKAGES 100%
POSTINSTALL 100%
Then choose
root password
login name and password
enable DHCP
the rest select default )
Then the grey "Finish" turns dark at the end of postinstall, I click
it.
The screen becomes command interface and bla, bla. Finally
...
...
...
INIT: Switching to runlevel: 3
INIT: Sending processes the TERM signal
Then nothing appears alive for a few hours (at least 3 hours).
The computer seems down, but it will respond to Ctrl Alt Del. And I
will have to start installation all over again
What is
wrong? Any help is appreciated.
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: Romano <[EMAIL PROTECTED]>
Subject: HLP Newbie: kppp and password
Date: Sun, 02 Jul 2000 19:01:09 +0200
Is there any way to run kppp and in general pppd connections in
normal user mode?
I mean I do not want to insert the root password every time I connect to
my ISP ( I have a PC and I'm the unique user.. It is not conformable to
type password every time.......
Thank you
------------------------------
** 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
******************************
