Библиотека сайта или "Мой Linux Documentation Project"
There's More Than One Way To Do It - TMTOWTDI
|Larry Wall, Tom Christiansen & Randal L. Schwartz: Programming Perl, Sec. Ed. 1996 p. 10|
From the Battery-Powered-HOWTO : "Installing and using Linux on a laptop is usually no problem at all, so go ahead and give it a try. Unlike some other operating systems, Linux still supports and runs well on even very old hardware, so you might give your outdated portable a new purpose in life by installing Linux on it."
One of the great benefits of Linux are its numerous and flexible installation features, which I don't want to describe in detail. Instead I try to focus on laptop specific methods, which are necessary only in certain circumstances.
Most current distributions support installation methods which are useful for laptops, including installation from CD-ROM, via PCMCIA and NFS (or maybe SMB). Please see the documents which are provided with these distributions for further details or take a look at the above mentioned manuals and HOWTOs.
With modern laptops, the traditional Linux installation method (from one boot floppy, one support floppy and a package of CD-ROMs or one DVD) should be no problem, if there is a floppy drive and a CD-ROM drive available. Though with certain laptops you might get trouble, if you can not use the floppy drive and the CD/DVD-ROM drive simultaneously, or if the floppy drive is only available as a PCMCIA device, as with the Toshiba Libretto 100. Some laptops support also booting and therefore installation completely from a CD drive, as reported for the SONY VAIO in the VAIO+Linux-HOWTO . Note: Check the BIOS for the CD boot option and make sure your Linux distribution comes on a bootable CD.
Certain laptops will only boot zImage kernels. bzImage kernels won't work. This is a known problem with the IBM™ Thinkpad 600 and Toshiba Tecra series, for instance. Some distributions provide certain boot floppies for these machines or for machines with limited memory resources, Debian/GNU Linux for instance.
Newer laptops are able to boot a Linux distribution from a bootable CD/DVD-ROM. This allows installation without a floppy disk drive. If the CD/DVD drive is only available as a PCMCIA device, as with the SONY VAIO PCG-Z600TEK, see the chapter about installing from PCMCIA devices below.
This is a short description of how to install from a CD-ROM under DOS without using boot or supplemental floppy diskettes. This is especially useful for notebooks with swappable floppy and CD-ROM components (if both are mutually exclusive) or if they are only available as PCMCIA devices. I have taken this method from Installing Debian GNU/Linux 2.1 For Intel x86 - Chapter 5 Methods for Installing Debian :
Get the following files from your nearest Debian FTP mirror and put them into a directory on your DOS partition: resc1440.bin drv1440.bin base2_1.tgz root.bin linux install.bat and loadlin.exe.
Boot into DOS (not Windows) without any drivers being loaded. To do this, you have to press <F8> at exactly the right moment during boot.
Execute install.bat from the directory where you have put the downloaded files.
Reboot the system and install the rest of the distribution, you may now use all the advanced features such as PCMCIA, PPP and others.
This should work for other distributions as well. Maybe you have to do some appropriate changes.
Because of their small or nonexistent footprint, micro-Linuxes are especially suited to run on laptops, particularly if you use a company-provided laptop running Windows9x/NT. Or for installation purposes using another non Linux machine. There are several micro Linux distributions out there that boot from one or two floppies and run off a ramdisk. See Appendix A Appendix A for a listing of distributions.
I tried the following with muLinux ( available at muLinux ) to clone my HP OmniBook 800 to a COMPAQ Armada 1592DT. Thanks to Michele Andreoli, maintainer of muLinux for his support. Since muLinux doesn't support PCMCIA yet, you may use TomsRtBt instead. In turn TomsRtBt doesn't support PPP but provides slip. Note: Since version 7.0 muLinux provides an Add-On with PCMCIA support.
I have described how to copy an already existing partition, but it might also be possible to achieve a customized installation. Note: Usually you would try to achieve an installation via NFS, which is supported by many distributions. Or if your sources are not at a Linux machine you might try the SMB protocol with SAMBA, which is also supported by muLinux .
You need two machines equipped with Linux. With the laptop (client/destination) on which you want to install Linux use the muLinux floppy. The other machine (server/source) may be a usual Linux box or also using muLinux. Though its low transfer rate I use a serial null modem cable because its cheap. You may apply the appropriate method using a PCMCIA network card and a crossover network cable or a HUB, or a parallel "null modem" cable and PLIP. As the basic protocol I used PPP, but you may also use SLIP. For the data-transfer I used nc. Note: this is an abbrevation for netcat, some distributions use this as the program name. You may use ftp, tftp, rsh, ssh, dd, rcp, kermit, NFS, SMB and other programs instead.
Basic requirements are:
A good knowledge about using Linux. You have to know exactly what you are doing, if not you might end destroying former installations.
A null modem serial cable.
At your source machine issue the following commands (attention: IP address, port number, partition and tty are just examples!):
/etc/ppp/options, it should contain only:
/dev/ttyS0 115200 passive
With muLinux versions 3.x you may even use the convenient command setup -f ppp .
Start PPP: pppd .
Configure the PPP network device: ifconfig ppp0 192.168.0.1 .
Add the default route: route add default gw 192.168.0.1 .
Check the network connection: ping 192.168.0.2, though the destination machine isn't up yet.
Start the transfer from another console, remember <LEFT-ALT><Fx>: cat /dev/hda2 | gzip -c | nc -l -p 5555 .
After the transfer (there are no more harddisk writings) stop the ping: killall ping .
At the destination machine issue:
/etc/ppp/options, it should contain only:
/dev/ttyS0 115200 passive
With muLinux versions >= 3.x you may even use the convenient command setup -f ppp .
Start PPP: pppd .
Configure the PPP network device: ifconfig ppp0 192.168.0.2 .
Add the default route: route add default gw 192.168.0.2 .
Check the network connection, by pinging to the source machine: ping 192.168.0.1 .
Change to another console and get the data from the server: nc 192.168.0.1 5555 | gzip -dc >/dev/hda4 .
400 MB may take app. 6 hours, but your mileage may vary.
Stop the transfer, when it is finished with: <CTL><C> . This can probably be avoided (but I didn't test it) by adding a timeout of 3 seconds using the -w 3 parameter for nc at the destination machine nc -w 3 192.168.0.1 5555 | gzip -dc >/dev/hda4
After the transfer is completed, stop the ping: killall ping .
/etc/lilo.msgand start lilo .
Set the new root device to the kernel: rdev image root_device .
You may use bzip2 the same way as gzip (untested).
Since rshd, sshd, ftpd daemons are not available with muLinux, you have to build your own file transfer mechanism with nc also known as netcat, as described above.
I had to set up both PPP sides very quickly or the connection broke, I don't know why.
Speed optimization has to be done. Maybe these PPP options will help: asyncmap 0 or local.
I checked this only with a destination partition greater than the source partition. Please check dd instead of cat therefore.
Or do the following (untested): At the destination machine cd into the root directory
/and do nc -l -p 5555 | bzip2 -dc | tar xvf -. At the source machine cd into the root directory
/and do tar cvf - . | bzip2 | nc -w 3 192.168.0.2 5555. This should shorten the time needed for the operation, too. Because only the allocated blocks need to be transfered.
Don't mount the destination partition.
From Adam Sulmicki adam_AT_cfar.unc.edu I got this hint: Most but not all harddisks in laptops are removable, but this might be not an easy task. You could just buy one of those cheap 2.5" IDE converters/adapters which allow you to connect this harddisk temporarily to a PC with IDE subsystem, and install Linux as usual using that PC. You may do so using the harddisk as the first IDE drive or besides as the second IDE drive. But then you need to be sure that lilo writes to the right partition. Also you have to make sure that you use the same translation style as your laptop is going to use (i.e. LBA vs. LARGE vs. CHS ). You will find additional information in the Hard-Disk-Upgrade-HOWTO. You might copy an existing partition, but it is also possible to achieve a customized installation.
Since I don't have a laptop which comes with a PCMCIA floppy drive (for instance Toshiba Libretto 100), I couldn't check this method. Please see the chapter Booting from a PCMCIA Device in the PCMCIA-HOWTO. Also I couldn't check whether booting from a PCMCIA harddisk is possible.
Anyway, when you are able to boot from a floppy and the laptop provides a PCMCIA slot, it should be possible to use different PCMCIA cards to connect to another machine, to an external SCSI device, different external CD and ZIP drives and others. Usually these methods are described in the documentation which is provided with the distribution.
The Sony Vaio (PCG-Z600) comes with an external USB-Floppy and an external CD-ROM (PCMCIA). You can boot from the CD-ROM, but afterwards Linux doesn't recognize the same drive anymore so that you can't install from it. You'll have to add the bootparameter linux ide2=0x180,0x360 (or 0x180,0x386?) at the LILO boot prompt if you want Linux to recognize a PCMCIA CDROM after the kernel has booted.
I couldn't check this method by myself, because I don't have such a device. Please check the appropriate Install-From-Zip-HOWTO . Also I don't know how much these installation methods are supported by the Linux distributions or the micro Linuxes. I suppose you have to fiddle around a bit to get this working.
From Jeremy Impson <jdimpson_AT_acm.org>: I installed Red Hat 6.1 on a Libretto 50CT. It only has a PCMCIA floppy drive. (Which BTW isn't well supported by the default PCMCIA floppy driver. I needed to download a patch from some Linux on Libretto web site.)
Linux will boot off the PCMCIA floppy drive, however. It just can't go back to the floppy after loading the kernel. My Libretto (the 50CT) only has one PCMCIA slot (later models had two slots, or I could have gotten the enhanced port replicator, which gave it another slot). So I couldn't boot off a floppy and then mount a remote filesystem.
So I downloaded ZipSlack (Slackware designed for running from a ZIP disk) and used another PC to load it onto a ZIP disk. I attached the ZIP drive to the Libretto (via the parallel port on the regular port replicator that comes with it) and booted from the Slackware boot disk in the PCMCIA floppy drive. When booted, I removed the floppy drive and inserted and configured a network PCMCIA card. At this point the kernel is in memory and it is using the filesystem on the ZIP disk.
I partitioned and formatted the Libretto's harddrive and then ftp'd Red
Hat 6.1 installation source onto one of the new partitions (the
partition that would become
/home when everything
gets done). This is the key: if you don't have enough disk space to have
the installation files plus enough to actually install the OS on to,
this method won't work.
I shut down the ZipSlack kernel and rebooted it using a RedHat install disk in the floppy drive. I pointed it at the RH6.1 installation media already on the harddrive and started the install.
I had tried myself to install Linux using the MicroSolutions BackPack parallel CD-ROM drive. It is fully supported by Linux and I haven't had any major problem running it. Until version 2.0.36 it is supported by its own module (bpck) while in later versions it has been merged in the more general parallel port ide adaptors (the paride module that relays then of course on more specific low level drivers, which in the BackPack case is still called bpck).
In RedHat 5.x based installations the bpck module is available already at installation stage so you'll just have to select the BackPack cdrom from the Other CD-ROMs at the installation stage and then give it some more options (but autoprobe should work just fine).
In RedHat 6.x (which uses 2.2.x kernels and should then use paride), the BackPack support was dropped. So to install the distribution from such a device, you will have to customize the bootdisk (adding the necessary modules) and the installation will be done without any problem.
Federico Pellegrin has customized a RedHat bootdisk that includes all the parallel CDROM devices that are supported by the distribution Linux kernel version (2.2.12) that should then work on all the supported parallel CDROM devices (even if he only tested it on his MicroSolutions BackPack since he doesn't have other similar hardware). You can find some information on it and the bootdisk image.
As from RedHat 6.2 a supplementary driver disk was included in the
distribution to support the paride devices. You'll just have to create
the driver disk (the image file is paride.img and can
be found in the
images/drivers directory) in the
usual way and insert it when the installer will ask for it.
Of course I suppose there isn't any problem in installing any other Linux distribution using such a device as long as you can add and configure the appropriate modules at the very beginning of the installation stage, but I haven't tested any.
You should take care of the mode the parallel port uses (ECP, EPP, Output only, PS/2) since some of them may cause your laptop to suddenly freeze or cause serious data corruption. On the other side some modes make the communication dramatically slow (I found the best choice on my laptop the PS/2, but you should make some tests).
This chapter is a courtesy of Federico Pellegrin. Please check also the CDROM-HOWTO.
I got this courtesy by Nathan Myers <ncm_AT_cantrip.org>: "Many distributions support installing via a network, using FTP, HTTP, or NFS. It is increasingly common for laptops to have only a single PCMCIA slot, already occupied by the boot floppy drive. Usually the boot floppy image has drivers for neither the floppy drive itself, nor the PCMCIA subsystem. Thus, the only network interface available may be the parallel port.
Installation via the parallel port using the PLIP protocol has been demonstrated on, at least, Red Hat. All you need is a Laplink parallel cable, cheap at any computer store. See the PLIP-HOWTO for details on setting up the connection. Note that (uniquely) the RedHat installation requires that the other end of the PLIP connection be configured to use ARP (apparently because RedHat uses the DOS driver in their installer). On the host, either export your CD file system on NFS, or mount it where the ftp or web daemon can find it, as needed for the installation."
The PLIP Install HOWTO by Gilles Lamiral describes how to install a Linux distribution on a computer without ethernet card, nor CD drive, but just a local floppy drive and a remote NFS server attached by a nullmodem parallel cable.
If booting from an USB device is supported from the BIOS, it is possible to install Linux from this drive. Besides some old laptops, almost all laptops equipped with USB ports support this feature.
First you have to configure the BIOS to boot from an USB device. Sometimes it is possible to use a certain key combination (e.g. <ESC>) during the boot process to select the boot device.
Second you have to install Linux on the boot medium (let's say an USB-Stick) and make it bootable. There are some special Linux distributions available, which are dedicated for such purposes, e.g.:
Feather Linux is a Linux distribution which runs completely off a CD or a USB pendrive and takes up under 64Mb of space. It is a Knoppix remastered (based on Debian/GNU Linux), and tries to include software which most people would use every day on their desktop. See these instructions about installing Feather Linux on an USB drive.
Partboot is dedictated to USB floppy drives and tailored for Linux laptop and notebook installations (you may find tools to resize your partitions as well as PCMCIA support and more).
Damn Small Linux (DSL) is a business-card size (50MB) Live CD Linux distribution. Despite its minuscule size it strives to have a functional and easy to use desktop.
On most modern laptops and notebooks with integrated network card, a network installation via the PXE protocol is easy to achieve. This comes in handy especially if there is no CD or DVD drive available.
For my installation I have used a Knoppix CD in the source machine. Just enable the Terminal Server (KNOPPIX->Server-Dienste->Terminal-Server KNOPPIX-Services-Start-> KNOPPIX Terminal Server) For almost any laptop model the default network drivers should work. Disable secure options, otherwise you will not be able to become the root user on the target machine. Besides using Knoppix, there are numerous ways to prepare the source machine for PXE. I haven't checked the EtherBoot protocol yet, but this might work too.
Look up the BIOS for something like a NetBoot Option and set it on. Boot the machine and choose booting from the network device. This is usually achieved by pressing a certain key during boot up or by pre-selecting the network interface as the boot device in the BIOS. Now Knoppix should come up. Open a shell and do an su to become root. To achieve a hard disk installation do either knx-hdinstall for Knoppix <=3.3 or knoppix-installer for Knoppix >=3.3.
You might ask why do a laptop installation via the VNC protocol? Indeed I know only of one reason to do so. Imagine you want to use a laptop with a broken keyboard you may use the keyboard of the remote machine to achieve the installation. Though you have to do a few key stroke to initiate the VNC installation! You have to prepare the source machine accordingly (instructions how to do so will follow later). For recent SuSE versions the distribution is already prepared, see the handbook for details.
If you have less than 8MB memory and want to install via NFS you may get the message "fork: out of memory". To handle this problem, use fdisk to make a swap partition (fdisk should be on the install floppy or take one of the mini Linuxes described above). Then try to boot from the install floppy again. Before configuring the NFS connection change to another console (for instance by pressing <ALT><F2>) and issue swapon /dev/xxx (xxx = swap partition ). Thanks to Thomas Schmaltz.
Bruce Richardson has written the 4MB-Laptop-HOWTO on installing a modern Linux distribution (specifically Slackware 7.0) onto laptops with 4MB RAM and <= 200MB hard disks. Another HOWTO is Getting Linux into Small Machines - HOWTO by L.C. Benschop.
Macintosh PowerBooks these days come with a CD/DVD drive but not a floppy drive, but the Linux distributions for PPC support booting and installation off of a CD without any need for a floppy.
Sometimes, when you boot the installer on the PowerBooks, the screen is black; this is easily fixed by tapping the brightness key on the keyboard (somehow, the screen brightness gets reset to zero).
If you have a very recent PowerBook, it may not be supported by the kernel on the installation CD. You can get around this by booting off of a recent kernel downloaded onto your hard drive and using a ramdisk on the CD or hard drive, while still loading the installation packages from the CD (the default). (See the instructions available online for yaBoot or BootX, the Linux/PPC boot loaders; yaBoot is currently better-supported on the newest machines.)
They can also boot/install from the Macintosh (HFS) partition on the internal hard disk.
This part is a courtesy of Steven G. Johnson.
For Linux installation reports see Linux Laptop and Notebook Survey: Apple.
If you have a 2,5" to 3,5" IDE drive adapter you can install one of the laptops, and with a desktop computer clone this harddisk to the disks of the other 99 laptops. You can use the DOS utility GHOST (works pretty with ext2) or with tar if the desktop works in linux. You only need an additional boot disk for the reinstall of the lilo in each laptop and change the hostname and IP address. These adapter are usually quite cheap (app . ten dollar, but difficult to get) .
VA SystemImager is software that makes the installation of Linux to masses of similar machines relatively easy. It also makes software distribution, configuration, and operating system updates easy. You can even update from one Linux release version to another! VA SystemImager can also be used for content management on web servers. It is most useful in environments where you have large numbers of identical machines. Some typical environments include: Internet server farms, high performance clusters, computer labs, or corporate desktop environments where all workstations have the same basic hardware configuration.
You might want to take a look at FAI - Fully Automatic Installation.
The package ALICE - Automatic Linux Installation and Configuration Environment, offers CVS-based configuration files and configuration templates.
Replicator is a set of scripts to automate the duplication of a Debian GNU/Linux installation from one computer to another. Replicator makes an effort to take into account differences in hardware (like HD size, video card) and in software configuration (such as partitioning). After the initial configuration, the scripts will create a bootdisk that allows you to completely (re)install a Debian box by booting from the floppy and answering a yes/no question.
Also bpbatch seems to be a good alternative.
Partition Image is a Linux/UNIX utility which saves partitions in the ext2fs (the linux standard), ReiserFS (a new journalized and powerful file system) or FAT16/32 (MS-DOS and MS-Windows file systems) file system format to an image file. The image file can be compressed in the GZIP/BZIP2 formats to save disk space, and splitted into multiple files to be copied on floppies (ZIP for example).
Только зарегистрированные пользователи могут оценивать и комментировать статьи.