diff --git a/LDP/howto/linuxdoc/Multi-Disk-HOWTO.sgml b/LDP/howto/linuxdoc/Multi-Disk-HOWTO.sgml index 5eade324..713e3ab1 100644 --- a/LDP/howto/linuxdoc/Multi-Disk-HOWTO.sgml +++ b/LDP/howto/linuxdoc/Multi-Disk-HOWTO.sgml @@ -122,6 +122,19 @@ Changelog: 2407 : Fixed a typo, sent in to ldp-submit ->0.31 + 2009 : User inputs to file systems and FEM correction ->0.32 + 1610 : Minor updates and release ->0.32a + 0511 : Fixed one typo and added link to scsidev development page ->0.32b + 1711 : Fixed typos and some links ->0.32c + 0312 : Another round of link checking, will this never end? ->0.32d + 1012 : Evidently not, more links updated -> 0.32e + 1012 : And again, more links updated -> 0.32f + 090101: Applied patch from Nakano-san -> 0.32g + 0901 : Added new link to INN optimising, fixed one link ->0.32h + 3006 : Added recovering disk failure, Win2000 RAID, iSCSI, corrections to mount point list ->0.32i + + + --> @@ -134,7 +147,7 @@ Changelog: HOWTO: Multi Disk System Tuning <author>Stein Gjoen, <tt/sgjoen@nyx.net/ -<date>v0.31 , 24 July 2000 +<date>v0.32i, 30 June 2001 <abstract> <nidx>disk</nidx> <nidx>partitions, disk (see disk)</nidx> @@ -178,7 +191,7 @@ the <bf/The Newer Generation/ release. --> this brand new release is codenamed the <bf/Daniella/ release. --> For unclear reasons this brand new release is codenamed <!-- the <bf/Sauchiehall/ release. --> -the <bf/Taylor2/ release. +the <bf/Taylor3/ release. New code names will appear as per industry standard guidelines to emphasize the state-of-the-art-ness of this document. @@ -207,12 +220,16 @@ private but a general release is hopefully in the near future. --> The followup to FSSTND is called the Filesystem Hierarchy Standard (FHS) -and covers more than Linux alone. FHS versions 2.0 and 2.1 have been -released but there are still a few issues to be dealt with and even +and covers more than Linux alone. FHS versions 2.0, 2.1 and 2.2 have been +released but there are still a few issues to be dealt with. Many recent +distributions are now aiming for FHS compliance. +<!-- removed 010630 + and even longer before this new standard will have an impact on actual distributions. FHS is not yet used in any distributions but Debian -has announced they will use it in Debian 2.1 which is their next -distribution. +has announced they will use it in Debian 2.1 which is the current +distribution. Also SuSE is aiming for FHS compliance and no doubt more +will come. --> It is also a good idea to read the Linux Installation guides thoroughly and if you are using a PC system, which I guess the majority still does, @@ -282,8 +299,10 @@ major installation and backups at regular intervals. <p> <nidx>disk!news on</nidx> +This is a maintenance release featuring minor but numerous updates +and additions to file systems and also tools for mount tables. -This release features a major restructuring and more additions +<!-- This release features a major restructuring and more additions than I can list here especially on backup systems, hints and tips and even more on file system support. Also there is now a new appendix with a shell script that helps @@ -297,9 +316,10 @@ and the old version will therefore not format this document properly. Also quite new is a number of new translations available. Now a Chinese and also an Italian translation are under way. + --> On the development front people are concentrating their energy towards -completing Linux 2.2 and until that is released there is not going to +completing Linux 2.4 and until that is released there is not going to be much news on disk technology for Linux. <!-- Debian 2.1 is readying for release and as I use Debian for my test @@ -340,7 +360,8 @@ in a number of formats: A European mirror of the -<url url="http://home.sol.no/˜gjoen/stein/disk.html" +<!-- <url url="http://home.sol.no/˜gjoen/stein/disk.html" --> +<url url="http://home.online.no/˜ggjoeen/stein/disk.html" name="Multi Disk HOWTO"> just went on line. @@ -397,6 +418,9 @@ mbaehr (at) email.archlab.tuwien.ac.at adc (at) postoffice.utas.edu.au pjm (at) bofh.asn.au jochen.berg (at) ac.com +jpotts (at) us.ibm.com +jarry (at) gmx.net +LeBlanc (at) mcc.ac.uk </verb></tscreen> @@ -433,7 +457,8 @@ to the translators for the job and the input they have given: </itemize> - +ICP Vortex is gratefully acknowledges for sending in-depth information +on their range of RAID controllers. Also DPT is acknowledged for sending me documentation on their controllers as well as permission to quote from the material. These quotes have been @@ -941,8 +966,8 @@ crowded cabinets. </itemize> More information on SCSI cabling and termination can be found at -<url url="http://resource.simplenet.com/files/68_50_n.htm" - name="other"> +<!-- <url url="http://resource.simplenet.com/files/68_50_n.htm" + name="other"> --> various web pages around the net. @@ -1117,7 +1142,8 @@ as well as <url url="http://spin.ch/˜tpo/bench/" name="this one"> and also -<url url="http://metalab.unc.edu/LDP/HOWTO/Benchmarking-HOWTO.html" +<!-- <url url="http://metalab.unc.edu/LDP/HOWTO/Benchmarking-HOWTO.html" --> +<url url="http://www.linuxdoc.org/HOWTO/Benchmarking-HOWTO.html" name="The Benchmarking-HOWTO">. There are also official home pages for @@ -1176,9 +1202,19 @@ PCI bus is getting closer. Currently the 64 bit version has a limit of 264 MB/s. The PCI transfer rate will in the future be increased from the current 33 MHz to 66 MHz, thereby increasing the limit to 528 MB/s. +The ATA development is continuing and is increasing the performance +with the new ATA/100 standard. Since most ATA drives are slower in +sustained transfer from platter than this the performance increase +will for most people be small. + +More interesting is the Serial ATA development, where the flat cable +will be replaced with a high speed serial link. This makes cabling +far simpler than today and also it solves the problem of cabling +obstructing airflow over the drives. + Another trend is for larger and larger drives. I hear it is possible -to get 55 GB on a single drive though this is rather expensive. -Currently the optimum storage for your money is about 6.4 GB but also +to get 75 GB on a single drive though this is rather expensive. +Currently the optimum storage for your money is about 30 GB but also this is continuously increasing. The introduction of DVD will in the near future have a big impact, with nearly 20 GB on a single disk you can have a complete copy of even major FTP sites from around the @@ -1191,6 +1227,15 @@ a problem at these speeds. About one month after that again the first commercial 24x CD-ROMs were available... Currently you can get 40x and no doubt higher speeds are in the pipeline. +A project to encapsulate SCSI over TCP/IP, called +<url url="http://www.ietf.org/internet-drafts/draft-ietf-ips-iscsi-06.txt" + name="iSCSI"> +has started, and one +<url url="http://www.cs.uml.edu/~mbrown/iSCSI" + name="Linux iSCSI implementation"> +has appeared. + + <sect1>Recommendations <label id="recommendations"> <p> <nidx>disk!recommendations</nidx> @@ -1275,7 +1320,7 @@ by all main Linux distributions. <nidx>disk!FHS</nidx> Later it was decided to make a successor that should also support operating systems other than just Linux, called -the <em/Filesystem Hierarchy Standard/ (FHS) at version 2.1 currently. +the <em/Filesystem Hierarchy Standard/ (FHS) at version 2.2 currently. This standard is under continuous development and will soon be adopted by Linux distributions. @@ -1339,8 +1384,10 @@ one you will find that extra swap space will buy you extra time between reboots. Also remember to take into account the type of programs you use. -Some programs that have large working sets, such as finite element -modeling (FEM) have huge data structures loaded in RAM rather than +Some programs that have large working sets, such as +<!-- finite element method (FEM) --> +image processing software +have huge data structures loaded in RAM rather than working explicitly on disk files. Data and computing intensive programs like this will cause excessive swapping if you have less RAM than the requirements. @@ -1511,7 +1558,7 @@ and <tt>/home</tt> partition. The Linux Quota subsystem is capable of limiting the number of blocks and the number of inode a single user ID can allocate on a per-filesystem basis. See the <url -url="http://metalab.unc.edu/LDP/mini" name="Linux Quota mini-HOWTO"> by +url="http://www.linuxdoc.org/HOWTO/mini/Quota.html" name="Linux Quota mini-HOWTO"> by Albert M.C. Tam <tt/bertie (at) scn.org/ for details on setup. @@ -1823,10 +1870,14 @@ home page. <p> <nidx>disk!file system!ext3fs</nidx> This is the name for the upcoming successor to <tt/ext2fs/ due to enter -development kernel in the near future. Many features will be added to +stable kernel in the near future. Many features are added to <tt/ext2fs/ but to avoid confusion over the name after such a radical upgrade the name will be changed too. You may have heard of it already -but source code is now in beta relsease . <!--not yet available. --> +but source code is now in beta release . <!--not yet available. --> + +Patches are available at +<url url="ftp://ftp.linux.org.uk/pub/linux/sct/fs/jfs" + name="Linux.org">. @@ -1860,7 +1911,7 @@ home page. name="Silicon Graphics Inc (sgi)"> has started porting its mainframe grade file system to Linux. Source is not yet available as they are busily cleaning out -legal encumberance but once that is done they will provide the +legal encumbrance but once that is done they will provide the source code under GPL. More information is already available on the @@ -1871,8 +1922,6 @@ at SGI. - - <sect2><tt/reiserfs/ <p> <nidx>disk!file system!reiserfs</nidx> @@ -1882,7 +1931,8 @@ Hans Reiser <tt/reiser (at) RICOCHET.NET/ has put up the source to his tree based <!-- <url url="http://idiom.com/˜beverly/reiserfs.html" 990919 --> <!-- <url url="http://devlinux.com/namesys/" 000501 --> -<url url="http://devlinux.com/projects/reiserfs/" +<!-- <url url="http://devlinux.com/projects/reiserfs/" 001203 --> +<url url="http://www.namesys.com" name="reiserfs"> on the web. While his filesystem has some very interesting features and is much faster than <tt/ext2fs/ and is in use by a number of people. @@ -1909,6 +1959,21 @@ file system and volume management into a single layer. The Enhanced File System project is now dead. +<sect2><tt/Tux2 fs/ +<p> +<nidx>disk!file system!Tux2 fs</nidx> +This is a variation on the <tt/ext2fs/ that adds robustness +in case of unexpected interruptions such as power failure. +After such an event <tt/Tux2 fs/ will restart with the file system +in a consistent, recently recorded state without fsck or +other recovery operations. To achieve this <tt/Tux2 fs/ uses +a newly designed algorithm called Phase Tree. + +More information can be found at the +<url url="http://tux2.sourceforge.net" + name="project home page">. + + <sect1>Microsoft File Systems <p> <nidx>disk!file system!Microsoft</nidx> @@ -2171,9 +2236,16 @@ for more information as well as documentation. <sect2>Coda <p> <nidx>disk!file system!Coda</nidx> -Work has started on a free replacement of <tt/AFS/ and is called +<!-- Major input from Dr. A V LeBlanc --> +<!-- Work has started on a free replacement of <tt/AFS/ and is called --> +A networking filesystem similar to <tt/AFS/ is underway and is called <url url="http://coda.cs.cmu.edu/" name="Coda">. +This is designed to be more robust and fault tolerant than <tt/AFS/, +and supports mobile, disconnected operations. +Currently it does not scale very well, and does not really have +proper administrative tools, as <tt/AFS/ does and <tt/ARLA/ is +beginning to. <sect2><tt/nbd/ @@ -2181,13 +2253,27 @@ Work has started on a free replacement of <tt/AFS/ and is called <nidx>disk!file system!nbd</nidx> <nidx>disk!device!network block device</nidx> The -<url url="http://atrey.karlin.mff.cuni.cz/˜pavel" +<url url="http://atrey.karlin.mff.cuni.cz/˜pavel/" name="Network Block Device"> (<tt/nbd/) is available in Linux kernel 2.2 and later and offers reportedly excellent performance. The interesting thing here is that it can be combined with RAID (see later). +<sect2><tt/enbd/ +<p> +<nidx>disk!file system!enbd</nidx> +<nidx>disk!device!enhanced network block device</nidx> +The +<url url="http://www.it.uc3m.es/˜ptb/nbd" <!-- 001213 --> + name="Enhanced Network Block Device"> +(<tt/enbd/) is a project to enhance the <tt/nbd/ with +features such as block journaled multi channel communications, +internal failover and automatic balancing between channels +and more. + +The intended use is for RAID over the net. + <sect2>GFS <p> <nidx>disk!file system!GFS</nidx> @@ -2271,7 +2357,10 @@ Currently there are 2 solutions to this problem in various stages of development: <descrip> <tag/scsidev/ works by creating a database of drives and where they -belong, check <em/ man scsifs/ for more information +belong, check <em/ man scsifs/ and the +<htmlurl url="http://www.garloff.de/kurt/linux/scsidev/" + name="scsidev home page"> +for more information <tag/devfs/ is a more long term project aimed at getting around the whole business of device numbering by making the <!-- <file>/dev</file> --> <htmlurl url="file:///dev/" @@ -2324,7 +2413,7 @@ check out the <url url="http://students.ceid.upatras.gr/˜gef/fs/oldindex.html" name="file system"> page -which has been superceded by +which has been superseded by <url url="http://www.penguin.cz/˜mhi/fs/" name="file system"> and the article @@ -2646,7 +2735,8 @@ that will reimplement many of the volume management functions found in IBM's AIX system. Unfortunately this project is currently on hold. Another project is the -<url url="http://linux.msede.com/lvm/" +<!-- <url url="http://linux.msede.com/lvm/" 001210 --> +<url url="http://www.sistina.com/lvm/" name="Logical Volume Manager"> project that is similar to a project by HP. @@ -2703,11 +2793,11 @@ which is a work in progress. A <url url="http://www-mddsp.enel.ucalgary.ca/People/adilger/online-ext2/" - name="patch for online growth of <tt/ext2fs/"> + name="patch for online growth of ext2fs"> is available in early stages and related work is taking place at <url url="http://ext2resize.sourceforge.net/" - name="the <tt/ext2fs/ resize project"> + name="the ext2fs resize project"> at Sourceforge. @@ -3135,7 +3225,8 @@ here is the power consumption, heat and noise. <p> <nidx>disk!technologies!yoke</nidx> There is also a -<url url="http://www.it.uc3m.es/˜ptb/cgi-bin/cvs-yoke.cgi" +<!-- <url url="http://www.it.uc3m.es/˜ptb/cgi-bin/cvs-yoke.cgi" --> +<url url="http://www.it.uc3m.es/cgi-bin/ptb/cvs-yoke.cgi" name="Linux Yoke Driver"> available in beta which is intended to do hot-swappable transparent binding of @@ -3249,10 +3340,10 @@ repair the system. Don't forget there are other alternatives to DOS, the most well known being -<url url="http://www.caldera/dos/" +<url url="http://www.caldera.com/dos/" name="DR-DOS"> from -<url url="http://www.caldera/" +<url url="http://www.caldera.com/" name="Caldera">. This is a direct descendant from DR-DOS from Digital Research. It offers many features not found in the more common DOS, such @@ -3323,8 +3414,20 @@ or by using a file explorer like tool called <sect1>OS/2 <p> <nidx>disk!operating systems, other!OS/2</nidx> -The only special note here is that you can get a file system driver for +The only special note here is that you can get file system driver for OS/2 that can read an <tt/ext2fs/ partition. +Matthieu Willm's ext2fs Installable File System for OS/2 can be found at +<url url="ftp://ftp-os2.nmsu.edu/pub/os2/system/drivers/filesys/ext2_240.zip" + name="ftp-os2.nmsu.edu">, +<url url="ftp://sunsite.unc.edu/pub/Linux/system/filesystems/ext2/ext2_240.zip" + name="Sunsite">, +<url url="ftp://ftp.leo.org/pub/comp/os/os2/drivers/ifs/ext2_240.zip" + name="ftp.leo.org"> and +<url url="ftp://ftp-os2.cdrom.com/pub/os2/diskutil/ext2_240.zip" + name="ftp-os2.cdrom.com">. + +The IFS has read and write capabilities. + <sect1>NT <p> @@ -3352,6 +3455,21 @@ You can now read <tt/ext2fs/ partitions from NT using name="Explore2fs">. +<sect1>Windows 2000 +<p> +<nidx>disk!operating systems, other!Windows 2000</nidx> +Most points regarding Windows NT also applies to its descendant Windows 2000 +though at the time of writing this I do not know if the aforementioned bugs +have been fixed or not. + +While Windows 2000, like its predecessor, features RAID, at least one +company, +<url url="http://www.raidtoolbox.com/" + name="RAID Toolbox">, +has found the bundled RAID somewhat lacking and made their own commercial +alternative. + + <sect1>Sun OS <p> <nidx>disk!operating systems, other!SunOS</nidx> @@ -3539,7 +3657,8 @@ NFS mounting onto nox. There are also some more advanced clustering projects going, notably <itemize> <item> -<url url="http://cesdis.gsfc.nasa.gov/linux/beowulf/beowulf.html" +<!-- <url url="http://cesdis.gsfc.nasa.gov/linux/beowulf/beowulf.html" --> +<url url="http://www.beowulf.org/" name="The Beowulf Project"> <item> @@ -3602,7 +3721,7 @@ Directory Suitability / | +-bin 0 -+-boot 0 ++-boot 3 +-dev 0 +-etc 0 +-home 5 @@ -3685,7 +3804,7 @@ With the cheap hardware available today it is possible to have quite a big system at home that is still cheap, systems that rival major servers of yesteryear. While many started out with old, discarded disks to build a Linux server (which is how this -HOWTO came into existence), many can now afford to buy 20 GB +HOWTO came into existence), many can now afford to buy 40 GB disks up front. Size remains important for some, and here are a few guidelines: @@ -3811,7 +3930,8 @@ the other hand be a disadvantage, if possible try to adjust based on the URL. For more information check up on the most used servers such as <tt/Harvest/, <!-- http://squid.nlanr.net/Squid --> -<url url="http://www.squid-cache.org/Squid" +<!-- <url url="http://www.squid-cache.org/Squid" 001203 --> +<url url="http://www.squid-cache.org/" name="Squid"> and the one from <url url="http://www.netscape.com" @@ -3886,7 +4006,8 @@ medium sized drives rather than one single huge disk. Also look into the High Availability (HA) project for more information. More information is available at -<url url="http://metalab.unc.edu/pub/Linux/ALPHA/linux-ha/High-Availability-HOWTO.html" +<!-- <url url="http://metalab.unc.edu/pub/Linux/ALPHA/linux-ha/High-Availability-HOWTO.html" 001203 --> +<url url="http://www.ibiblio.org/pub/Linux/ALPHA/linux-ha/High-Availability-HOWTO.html" name="High Availability HOWTO"> and also at related <url url="http://www.henge.com/˜alanr/ha/index.html" @@ -4081,11 +4202,15 @@ the many seeks among multiple spindles. It is recommended in the manuals and FAQs for the INN news server to put news spool and <tt>.overview</tt> files on separate drives for larger installations. -There is also a web page dedicated to +<!-- There is also a web page dedicated to 001210 gone. <url url="http://www.spinne.com/usenet/inn-perf.html" name="INN optimising"> -well worth reading. +well worth reading. --> +Some notes on +<url url="http://www.tru64unix.compaq.com/internet/inn-wp.html" + name="INN optimising under Tru64 UNIX"> +also applies to a wider audience, including Linux users. <tag/Database/ applications can be demanding both in terms of drive usage and speed requirements. The details are naturally application @@ -4368,7 +4493,7 @@ down into <tt>/etc</tt> directories. <sect>Implementation <p> <nidx>disk!implementation</nidx> -Having done the layout you should now have a detailled description on +Having done the layout you should now have a detailed description on what goes where. Most likely this will be on paper but hopefully someone will make a more automated system that can deal with everything from the design, through partitioning to formatting and @@ -4582,7 +4707,8 @@ This shows you what files are where so you can deal with them directly. A freeware alternative is -<url url="http://members.xoom.com/Zeleps" +<!-- <url url="http://members.xoom.com/Zeleps" 001203 --> +<url url="http://members.nbci.com/Zeleps/" name="Partition Resizer"> which can shrink, grow and move partitions. @@ -4606,7 +4732,8 @@ the Microsoft system will try to mount the last partition as if it were a FAT partition in place of the last primary FAT partition. There is more -<url url="http://www.v-com.com/support/osinstalls/notes/95notes.html" +<!-- <url url="http://www.v-com.com/support/osinstalls/notes/95notes.html" --> +<url url="http://www.v-com.com/" name="information"> available on the net on this. @@ -4615,7 +4742,8 @@ at the very end of your disk. More information on multi OS installations are available at <url url="http://www.v-com.com/" - name="V Communications">. + name="V Communications"> but they keep rearranging the +links continuously so no direct links can be offered here. Since some hardware comes with setup software that is available @@ -4708,7 +4836,13 @@ proc /proc proc defaults 0 0 This file is somewhat sensitive to the formatting used so it is best and also most convenient to edit it using one of the -editing tools made for this purpose. +editing tools made for this purpose, +such as +<url url="http://www.bit.net.au/˜bhepple/fstool/" + name="on the netfstool">, a Tcl/Tk-based file system mounter, +and +<url url="http://kfstab.purespace.de/kfstab/" + name="kfstab">, an editing tool for KDE. Briefly, the fields are partition name, where to mount the partition, type of file system, mount options, when to dump for backup @@ -5018,6 +5152,10 @@ Rescue disks can be gotten off the net, from your distribution or you can put one together yourself. Do make sure the boot and root parameters are set so the kernel will know where to find your system. +If you don't have a recovery floppy you can use the +<url url="http://www.gnu.org/software/grub/" + name="GRUB"> boot loader +to load from a Linux kernel somewhere on disk, with arguments. <sect>Advanced Issues @@ -5472,14 +5610,15 @@ could have a look at the <url url="http://bmrc.berkeley.edu/people/chaffee/index.html" name="development page">. <!-- Only minor details are missing before it comes into the kernel. --> -These drivers are now in the 2.1.x kernel development series as well as +These drivers are in the 2.1.x kernel development series as well as in 2.0.34 and later. +<!-- seems to be gone 001117 <item>For more information on booting and also some BSD information have a look at <url url="http://www.paranoia.com/˜vax/boot.html" name="booting information"> -page. +page. --> </itemize> @@ -6055,6 +6194,56 @@ Linux installations use modern file systems these schemes are not used, however, some other operating systems have retained such schemes. --> +<sect1>Crash Recovery +<p> +<nidx>disk!miscellaneous!recovery</nidx> +<nidx>disk!miscellaneous!crash recovery</nidx> +Occationally hard disks crash. A crash causing data scrambling can +often be at least partially recovered from and there are already +HOWTOs describing this. + +In case of hardware failure things are far more serious, and you +have two options: either send the drive to a professional data +recovery company, or try recovering yourself. The latter is of +course <em>high risk</em> and can cause more damage. + +If a disk stops rotating or fails to spin up, the number one +advice is first to turn off the system as fast as safely possible. + +Next you could try disconnecting the drives and power up the +machine, just to check power with a multimeter that power is +present. Quite often connectors can get unseated and cause all +sorts of problems. + +If you decide to risk trying it yourself you could check all +connectors and then reapply power and see if the drive spins up +and responds. If it still is dead turn off power quickly, +preferrably before the operating system boots. Make sure that +delayed spinup is not deceiving you here. + +If you decide to progress even further (and take higher risks) +you could remove the drive, give it a firm tap on the side so +that the disk moves a little with respect to the casing. This +can help in unsticking the head from the surface, allowing the +platter to move freely as the motor power is not sufficient to +unstick a stuck head on its own. + +Also if a drive has been turned off for a while after running +for long periods of time, or if it has overheated, the lubricant +can harden of drain out of the bearings. In this case warming the +drive slowly and gently up to normal operating temperature will +possibly recover the lubrication problems. + +If after this the drive still does not respond the last possible +and the highest risk suggestion is to replace the circuit board +of the drive with a board from am identical model drive. + +Often the contents of a drive is worth far more than the media +itself, so do consider professional help. These companies have +advanced equipment and know-how obtained from the manufacturers +on how to recover a damaged drive, far beyond that of a hobbyist. + + <!-- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% -->