NetBSD/sbin/newfs/newfs.8

375 lines
12 KiB
Groff
Raw Normal View History

Incorporate the enhanced ffs_dirpref() by Grigoriy Orlov, as found in FreeBSD (three commits; the initial work, man page updates, and a fix to ffs_reload()), with the following differences: - Be consistent between newfs(8) and tunefs(8) as to the options which set and control the tuning parameters for this work (avgfilesize & avgfpdir) - Use u_int16_t instead of u_int8_t to keep track of the number of contiguous directories (suggested by Chuck Silvers) - Work within our FFS_EI framework - Ensure that fs->fs_maxclusters and fs->fs_contigdirs don't point to the same area of memory The new algorithm has a marked performance increase, especially when performing tasks such as untarring pkgsrc.tar.gz, etc. The original FreeBSD commit messages are attached: ===== mckusick 2001/04/10 01:39:00 PDT Directory layout preference improvements from Grigoriy Orlov <gluk@ptci.ru>. His description of the problem and solution follow. My own tests show speedups on typical filesystem intensive workloads of 5% to 12% which is very impressive considering the small amount of code change involved. ------ One day I noticed that some file operations run much faster on small file systems then on big ones. I've looked at the ffs algorithms, thought about them, and redesigned the dirpref algorithm. First I want to describe the results of my tests. These results are old and I have improved the algorithm after these tests were done. Nevertheless they show how big the perfomance speedup may be. I have done two file/directory intensive tests on a two OpenBSD systems with old and new dirpref algorithm. The first test is "tar -xzf ports.tar.gz", the second is "rm -rf ports". The ports.tar.gz file is the ports collection from the OpenBSD 2.8 release. It contains 6596 directories and 13868 files. The test systems are: 1. Celeron-450, 128Mb, two IDE drives, the system at wd0, file system for test is at wd1. Size of test file system is 8 Gb, number of cg=991, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=35 2. PIII-600, 128Mb, two IBM DTLA-307045 IDE drives at i815e, the system at wd0, file system for test is at wd1. Size of test file system is 40 Gb, number of cg=5324, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=50 You can get more info about the test systems and methods at: http://www.ptci.ru/gluk/dirpref/old/dirpref.html Test Results tar -xzf ports.tar.gz rm -rf ports mode old dirpref new dirpref speedup old dirprefnew dirpref speedup First system normal 667 472 1.41 477 331 1.44 async 285 144 1.98 130 14 9.29 sync 768 616 1.25 477 334 1.43 softdep 413 252 1.64 241 38 6.34 Second system normal 329 81 4.06 263.5 93.5 2.81 async 302 25.7 11.75 112 2.26 49.56 sync 281 57.0 4.93 263 90.5 2.9 softdep 341 40.6 8.4 284 4.76 59.66 "old dirpref" and "new dirpref" columns give a test time in seconds. speedup - speed increasement in times, ie. old dirpref / new dirpref. ------ Algorithm description The old dirpref algorithm is described in comments: /* * Find a cylinder to place a directory. * * The policy implemented by this algorithm is to select from * among those cylinder groups with above the average number of * free inodes, the one with the smallest number of directories. */ A new directory is allocated in a different cylinder groups than its parent directory resulting in a directory tree that is spreaded across all the cylinder groups. This spreading out results in a non-optimal access to the directories and files. When we have a small filesystem it is not a problem but when the filesystem is big then perfomance degradation becomes very apparent. What I mean by a big file system ? 1. A big filesystem is a filesystem which occupy 20-30 or more percent of total drive space, i.e. first and last cylinder are physically located relatively far from each other. 2. It has a relatively large number of cylinder groups, for example more cylinder groups than 50% of the buffers in the buffer cache. The first results in long access times, while the second results in many buffers being used by metadata operations. Such operations use cylinder group blocks and on-disk inode blocks. The cylinder group block (fs->fs_cblkno) contains struct cg, inode and block bit maps. It is 2k in size for the default filesystem parameters. If new and parent directories are located in different cylinder groups then the system performs more input/output operations and uses more buffers. On filesystems with many cylinder groups, lots of cache buffers are used for metadata operations. My solution for this problem is very simple. I allocate many directories in one cylinder group. I also do some things, so that the new allocation method does not cause excessive fragmentation and all directory inodes will not be located at a location far from its file's inodes and data. The algorithm is: /* * Find a cylinder group to place a directory. * * The policy implemented by this algorithm is to allocate a * directory inode in the same cylinder group as its parent * directory, but also to reserve space for its files inodes * and data. Restrict the number of directories which may be * allocated one after another in the same cylinder group * without intervening allocation of files. * * If we allocate a first level directory then force allocation * in another cylinder group. */ My early versions of dirpref give me a good results for a wide range of file operations and different filesystem capacities except one case: those applications that create their entire directory structure first and only later fill this structure with files. My solution for such and similar cases is to limit a number of directories which may be created one after another in the same cylinder group without intervening file creations. For this purpose, I allocate an array of counters at mount time. This array is linked to the superblock fs->fs_contigdirs[cg]. Each time a directory is created the counter increases and each time a file is created the counter decreases. A 60Gb filesystem with 8mb/cg requires 10kb of memory for the counters array. The maxcontigdirs is a maximum number of directories which may be created without an intervening file creation. I found in my tests that the best performance occurs when I restrict the number of directories in one cylinder group such that all its files may be located in the same cylinder group. There may be some deterioration in performance if all the file inodes are in the same cylinder group as its containing directory, but their data partially resides in a different cylinder group. The maxcontigdirs value is calculated to try to prevent this condition. Since there is no way to know how many files and directories will be allocated later I added two optimization parameters in superblock/tunefs. They are: int32_t fs_avgfilesize; /* expected average file size */ int32_t fs_avgfpdir; /* expected # of files per directory */ These parameters have reasonable defaults but may be tweeked for special uses of a filesystem. They are only necessary in rare cases like better tuning a filesystem being used to store a squid cache. I have been using this algorithm for about 3 months. I have done a lot of testing on filesystems with different capacities, average filesize, average number of files per directory, and so on. I think this algorithm has no negative impact on filesystem perfomance. It works better than the default one in all cases. The new dirpref will greatly improve untarring/removing/coping of big directories, decrease load on cvs servers and much more. The new dirpref doesn't speedup a compilation process, but also doesn't slow it down. Obtained from: Grigoriy Orlov <gluk@ptci.ru> ===== ===== iedowse 2001/04/23 17:37:17 PDT Pre-dirpref versions of fsck may zero out the new superblock fields fs_contigdirs, fs_avgfilesize and fs_avgfpdir. This could cause panics if these fields were zeroed while a filesystem was mounted read-only, and then remounted read-write. Add code to ffs_reload() which copies the fs_contigdirs pointer from the previous superblock, and reinitialises fs_avgf* if necessary. Reviewed by: mckusick ===== ===== nik 2001/04/10 03:36:44 PDT Add information about the new options to newfs and tunefs which set the expected average file size and number of files per directory. Could do with some fleshing out. =====
2001-09-06 06:16:00 +04:00
.\" $NetBSD: newfs.8,v 1.33 2001/09/06 02:16:01 lukem Exp $
.\"
.\" Copyright (c) 1983, 1987, 1991, 1993, 1994
.\" The Regents of the University of California. All rights reserved.
1993-03-21 12:45:37 +03:00
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\" notice, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\" notice, this list of conditions and the following disclaimer in the
.\" documentation and/or other materials provided with the distribution.
.\" 3. All advertising materials mentioning features or use of this software
.\" must display the following acknowledgement:
.\" This product includes software developed by the University of
.\" California, Berkeley and its contributors.
.\" 4. Neither the name of the University nor the names of its contributors
.\" may be used to endorse or promote products derived from this software
.\" without specific prior written permission.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
.\" @(#)newfs.8 8.6 (Berkeley) 5/3/95
1993-03-22 11:04:00 +03:00
.\"
Incorporate the enhanced ffs_dirpref() by Grigoriy Orlov, as found in FreeBSD (three commits; the initial work, man page updates, and a fix to ffs_reload()), with the following differences: - Be consistent between newfs(8) and tunefs(8) as to the options which set and control the tuning parameters for this work (avgfilesize & avgfpdir) - Use u_int16_t instead of u_int8_t to keep track of the number of contiguous directories (suggested by Chuck Silvers) - Work within our FFS_EI framework - Ensure that fs->fs_maxclusters and fs->fs_contigdirs don't point to the same area of memory The new algorithm has a marked performance increase, especially when performing tasks such as untarring pkgsrc.tar.gz, etc. The original FreeBSD commit messages are attached: ===== mckusick 2001/04/10 01:39:00 PDT Directory layout preference improvements from Grigoriy Orlov <gluk@ptci.ru>. His description of the problem and solution follow. My own tests show speedups on typical filesystem intensive workloads of 5% to 12% which is very impressive considering the small amount of code change involved. ------ One day I noticed that some file operations run much faster on small file systems then on big ones. I've looked at the ffs algorithms, thought about them, and redesigned the dirpref algorithm. First I want to describe the results of my tests. These results are old and I have improved the algorithm after these tests were done. Nevertheless they show how big the perfomance speedup may be. I have done two file/directory intensive tests on a two OpenBSD systems with old and new dirpref algorithm. The first test is "tar -xzf ports.tar.gz", the second is "rm -rf ports". The ports.tar.gz file is the ports collection from the OpenBSD 2.8 release. It contains 6596 directories and 13868 files. The test systems are: 1. Celeron-450, 128Mb, two IDE drives, the system at wd0, file system for test is at wd1. Size of test file system is 8 Gb, number of cg=991, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=35 2. PIII-600, 128Mb, two IBM DTLA-307045 IDE drives at i815e, the system at wd0, file system for test is at wd1. Size of test file system is 40 Gb, number of cg=5324, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=50 You can get more info about the test systems and methods at: http://www.ptci.ru/gluk/dirpref/old/dirpref.html Test Results tar -xzf ports.tar.gz rm -rf ports mode old dirpref new dirpref speedup old dirprefnew dirpref speedup First system normal 667 472 1.41 477 331 1.44 async 285 144 1.98 130 14 9.29 sync 768 616 1.25 477 334 1.43 softdep 413 252 1.64 241 38 6.34 Second system normal 329 81 4.06 263.5 93.5 2.81 async 302 25.7 11.75 112 2.26 49.56 sync 281 57.0 4.93 263 90.5 2.9 softdep 341 40.6 8.4 284 4.76 59.66 "old dirpref" and "new dirpref" columns give a test time in seconds. speedup - speed increasement in times, ie. old dirpref / new dirpref. ------ Algorithm description The old dirpref algorithm is described in comments: /* * Find a cylinder to place a directory. * * The policy implemented by this algorithm is to select from * among those cylinder groups with above the average number of * free inodes, the one with the smallest number of directories. */ A new directory is allocated in a different cylinder groups than its parent directory resulting in a directory tree that is spreaded across all the cylinder groups. This spreading out results in a non-optimal access to the directories and files. When we have a small filesystem it is not a problem but when the filesystem is big then perfomance degradation becomes very apparent. What I mean by a big file system ? 1. A big filesystem is a filesystem which occupy 20-30 or more percent of total drive space, i.e. first and last cylinder are physically located relatively far from each other. 2. It has a relatively large number of cylinder groups, for example more cylinder groups than 50% of the buffers in the buffer cache. The first results in long access times, while the second results in many buffers being used by metadata operations. Such operations use cylinder group blocks and on-disk inode blocks. The cylinder group block (fs->fs_cblkno) contains struct cg, inode and block bit maps. It is 2k in size for the default filesystem parameters. If new and parent directories are located in different cylinder groups then the system performs more input/output operations and uses more buffers. On filesystems with many cylinder groups, lots of cache buffers are used for metadata operations. My solution for this problem is very simple. I allocate many directories in one cylinder group. I also do some things, so that the new allocation method does not cause excessive fragmentation and all directory inodes will not be located at a location far from its file's inodes and data. The algorithm is: /* * Find a cylinder group to place a directory. * * The policy implemented by this algorithm is to allocate a * directory inode in the same cylinder group as its parent * directory, but also to reserve space for its files inodes * and data. Restrict the number of directories which may be * allocated one after another in the same cylinder group * without intervening allocation of files. * * If we allocate a first level directory then force allocation * in another cylinder group. */ My early versions of dirpref give me a good results for a wide range of file operations and different filesystem capacities except one case: those applications that create their entire directory structure first and only later fill this structure with files. My solution for such and similar cases is to limit a number of directories which may be created one after another in the same cylinder group without intervening file creations. For this purpose, I allocate an array of counters at mount time. This array is linked to the superblock fs->fs_contigdirs[cg]. Each time a directory is created the counter increases and each time a file is created the counter decreases. A 60Gb filesystem with 8mb/cg requires 10kb of memory for the counters array. The maxcontigdirs is a maximum number of directories which may be created without an intervening file creation. I found in my tests that the best performance occurs when I restrict the number of directories in one cylinder group such that all its files may be located in the same cylinder group. There may be some deterioration in performance if all the file inodes are in the same cylinder group as its containing directory, but their data partially resides in a different cylinder group. The maxcontigdirs value is calculated to try to prevent this condition. Since there is no way to know how many files and directories will be allocated later I added two optimization parameters in superblock/tunefs. They are: int32_t fs_avgfilesize; /* expected average file size */ int32_t fs_avgfpdir; /* expected # of files per directory */ These parameters have reasonable defaults but may be tweeked for special uses of a filesystem. They are only necessary in rare cases like better tuning a filesystem being used to store a squid cache. I have been using this algorithm for about 3 months. I have done a lot of testing on filesystems with different capacities, average filesize, average number of files per directory, and so on. I think this algorithm has no negative impact on filesystem perfomance. It works better than the default one in all cases. The new dirpref will greatly improve untarring/removing/coping of big directories, decrease load on cvs servers and much more. The new dirpref doesn't speedup a compilation process, but also doesn't slow it down. Obtained from: Grigoriy Orlov <gluk@ptci.ru> ===== ===== iedowse 2001/04/23 17:37:17 PDT Pre-dirpref versions of fsck may zero out the new superblock fields fs_contigdirs, fs_avgfilesize and fs_avgfpdir. This could cause panics if these fields were zeroed while a filesystem was mounted read-only, and then remounted read-write. Add code to ffs_reload() which copies the fs_contigdirs pointer from the previous superblock, and reinitialises fs_avgf* if necessary. Reviewed by: mckusick ===== ===== nik 2001/04/10 03:36:44 PDT Add information about the new options to newfs and tunefs which set the expected average file size and number of files per directory. Could do with some fleshing out. =====
2001-09-06 06:16:00 +04:00
.Dd September 6, 2001
1993-03-21 12:45:37 +03:00
.Dt NEWFS 8
2001-06-05 15:22:41 +04:00
.Os
1993-03-21 12:45:37 +03:00
.Sh NAME
.Nm newfs ,
1994-07-16 09:18:11 +04:00
.Nm mount_mfs
1993-03-21 12:45:37 +03:00
.Nd construct a new file system
.Sh SYNOPSIS
1999-03-07 14:02:05 +03:00
.Nm ""
.Op Fl B Ar byte-order
.Op Fl FNOZ
.Op Fl S Ar sector-size
.Op Fl T Ar disk-type
.Op Fl a Ar maxcontig
.Op Fl b Ar block-size
.Op Fl c Ar cpg
.Op Fl d Ar rotdelay
.Op Fl e Ar maxbpg
.Op Fl f Ar frag-size
Incorporate the enhanced ffs_dirpref() by Grigoriy Orlov, as found in FreeBSD (three commits; the initial work, man page updates, and a fix to ffs_reload()), with the following differences: - Be consistent between newfs(8) and tunefs(8) as to the options which set and control the tuning parameters for this work (avgfilesize & avgfpdir) - Use u_int16_t instead of u_int8_t to keep track of the number of contiguous directories (suggested by Chuck Silvers) - Work within our FFS_EI framework - Ensure that fs->fs_maxclusters and fs->fs_contigdirs don't point to the same area of memory The new algorithm has a marked performance increase, especially when performing tasks such as untarring pkgsrc.tar.gz, etc. The original FreeBSD commit messages are attached: ===== mckusick 2001/04/10 01:39:00 PDT Directory layout preference improvements from Grigoriy Orlov <gluk@ptci.ru>. His description of the problem and solution follow. My own tests show speedups on typical filesystem intensive workloads of 5% to 12% which is very impressive considering the small amount of code change involved. ------ One day I noticed that some file operations run much faster on small file systems then on big ones. I've looked at the ffs algorithms, thought about them, and redesigned the dirpref algorithm. First I want to describe the results of my tests. These results are old and I have improved the algorithm after these tests were done. Nevertheless they show how big the perfomance speedup may be. I have done two file/directory intensive tests on a two OpenBSD systems with old and new dirpref algorithm. The first test is "tar -xzf ports.tar.gz", the second is "rm -rf ports". The ports.tar.gz file is the ports collection from the OpenBSD 2.8 release. It contains 6596 directories and 13868 files. The test systems are: 1. Celeron-450, 128Mb, two IDE drives, the system at wd0, file system for test is at wd1. Size of test file system is 8 Gb, number of cg=991, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=35 2. PIII-600, 128Mb, two IBM DTLA-307045 IDE drives at i815e, the system at wd0, file system for test is at wd1. Size of test file system is 40 Gb, number of cg=5324, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=50 You can get more info about the test systems and methods at: http://www.ptci.ru/gluk/dirpref/old/dirpref.html Test Results tar -xzf ports.tar.gz rm -rf ports mode old dirpref new dirpref speedup old dirprefnew dirpref speedup First system normal 667 472 1.41 477 331 1.44 async 285 144 1.98 130 14 9.29 sync 768 616 1.25 477 334 1.43 softdep 413 252 1.64 241 38 6.34 Second system normal 329 81 4.06 263.5 93.5 2.81 async 302 25.7 11.75 112 2.26 49.56 sync 281 57.0 4.93 263 90.5 2.9 softdep 341 40.6 8.4 284 4.76 59.66 "old dirpref" and "new dirpref" columns give a test time in seconds. speedup - speed increasement in times, ie. old dirpref / new dirpref. ------ Algorithm description The old dirpref algorithm is described in comments: /* * Find a cylinder to place a directory. * * The policy implemented by this algorithm is to select from * among those cylinder groups with above the average number of * free inodes, the one with the smallest number of directories. */ A new directory is allocated in a different cylinder groups than its parent directory resulting in a directory tree that is spreaded across all the cylinder groups. This spreading out results in a non-optimal access to the directories and files. When we have a small filesystem it is not a problem but when the filesystem is big then perfomance degradation becomes very apparent. What I mean by a big file system ? 1. A big filesystem is a filesystem which occupy 20-30 or more percent of total drive space, i.e. first and last cylinder are physically located relatively far from each other. 2. It has a relatively large number of cylinder groups, for example more cylinder groups than 50% of the buffers in the buffer cache. The first results in long access times, while the second results in many buffers being used by metadata operations. Such operations use cylinder group blocks and on-disk inode blocks. The cylinder group block (fs->fs_cblkno) contains struct cg, inode and block bit maps. It is 2k in size for the default filesystem parameters. If new and parent directories are located in different cylinder groups then the system performs more input/output operations and uses more buffers. On filesystems with many cylinder groups, lots of cache buffers are used for metadata operations. My solution for this problem is very simple. I allocate many directories in one cylinder group. I also do some things, so that the new allocation method does not cause excessive fragmentation and all directory inodes will not be located at a location far from its file's inodes and data. The algorithm is: /* * Find a cylinder group to place a directory. * * The policy implemented by this algorithm is to allocate a * directory inode in the same cylinder group as its parent * directory, but also to reserve space for its files inodes * and data. Restrict the number of directories which may be * allocated one after another in the same cylinder group * without intervening allocation of files. * * If we allocate a first level directory then force allocation * in another cylinder group. */ My early versions of dirpref give me a good results for a wide range of file operations and different filesystem capacities except one case: those applications that create their entire directory structure first and only later fill this structure with files. My solution for such and similar cases is to limit a number of directories which may be created one after another in the same cylinder group without intervening file creations. For this purpose, I allocate an array of counters at mount time. This array is linked to the superblock fs->fs_contigdirs[cg]. Each time a directory is created the counter increases and each time a file is created the counter decreases. A 60Gb filesystem with 8mb/cg requires 10kb of memory for the counters array. The maxcontigdirs is a maximum number of directories which may be created without an intervening file creation. I found in my tests that the best performance occurs when I restrict the number of directories in one cylinder group such that all its files may be located in the same cylinder group. There may be some deterioration in performance if all the file inodes are in the same cylinder group as its containing directory, but their data partially resides in a different cylinder group. The maxcontigdirs value is calculated to try to prevent this condition. Since there is no way to know how many files and directories will be allocated later I added two optimization parameters in superblock/tunefs. They are: int32_t fs_avgfilesize; /* expected average file size */ int32_t fs_avgfpdir; /* expected # of files per directory */ These parameters have reasonable defaults but may be tweeked for special uses of a filesystem. They are only necessary in rare cases like better tuning a filesystem being used to store a squid cache. I have been using this algorithm for about 3 months. I have done a lot of testing on filesystems with different capacities, average filesize, average number of files per directory, and so on. I think this algorithm has no negative impact on filesystem perfomance. It works better than the default one in all cases. The new dirpref will greatly improve untarring/removing/coping of big directories, decrease load on cvs servers and much more. The new dirpref doesn't speedup a compilation process, but also doesn't slow it down. Obtained from: Grigoriy Orlov <gluk@ptci.ru> ===== ===== iedowse 2001/04/23 17:37:17 PDT Pre-dirpref versions of fsck may zero out the new superblock fields fs_contigdirs, fs_avgfilesize and fs_avgfpdir. This could cause panics if these fields were zeroed while a filesystem was mounted read-only, and then remounted read-write. Add code to ffs_reload() which copies the fs_contigdirs pointer from the previous superblock, and reinitialises fs_avgf* if necessary. Reviewed by: mckusick ===== ===== nik 2001/04/10 03:36:44 PDT Add information about the new options to newfs and tunefs which set the expected average file size and number of files per directory. Could do with some fleshing out. =====
2001-09-06 06:16:00 +04:00
.Op Fl g Ar avgfilesize
.Op Fl h Ar avgfpdir
.Op Fl i Ar bytes-per-inode
.Op Fl k Ar skew
.Op Fl l Ar interleave
.Op Fl m Ar free-space
.Op Fl n Ar rotational-positions
.Op Fl o Ar optimization
.Op Fl p Ar track-spares
.Op Fl r Ar revolutions
.Op Fl s Ar size
.Op Fl t Ar ntracks
.Op Fl u Ar nsectors
.Op Fl x Ar sectors
1993-03-21 12:45:37 +03:00
.Ar special
.Nm mount_mfs
.Op Fl N
.Op Fl T Ar disk-type
.Op Fl a Ar maxcontig
.Op Fl b Ar block-size
.Op Fl c Ar cpg
.Op Fl d Ar rotdelay
.Op Fl e Ar maxbpg
.Op Fl f Ar frag-size
.Op Fl i Ar bytes-per-inode
.Op Fl m Ar free-space
.Op Fl n Ar rotational-positions
.Op Fl o Ar options
.Op Fl s Ar size
1993-03-21 12:45:37 +03:00
.Ar special node
.Sh DESCRIPTION
.Nm
is used to initialize and clear file systems before first use.
1998-04-29 13:49:10 +04:00
Before running
.Nm
1993-03-21 12:45:37 +03:00
or
.Nm mount_mfs ,
1998-04-29 13:49:10 +04:00
the disk must be labeled using
1993-03-21 12:45:37 +03:00
.Xr disklabel 8 .
.Nm
1993-03-21 12:45:37 +03:00
builds a file system on the specified special device
basing its defaults on the information in the disk label.
Typically the defaults are reasonable, however
.Nm
1993-03-21 12:45:37 +03:00
has numerous options to allow the defaults to be selectively overridden.
.Pp
.Nm mount_mfs
1993-03-21 12:45:37 +03:00
is used to build a file system in virtual memory and then mount it
on a specified node.
.Nm mount_mfs
1993-03-21 12:45:37 +03:00
exits and the contents of the file system are lost
when the file system is unmounted.
If
.Nm mount_mfs
1993-03-21 12:45:37 +03:00
is sent a signal while running,
for example during system shutdown,
it will attempt to unmount its
corresponding file system.
The parameters to
.Nm mount_mfs
1993-03-21 12:45:37 +03:00
are the same as those to
.Nm "" .
If the
.Fl T
flag is specified (see below), the special file is unused.
Otherwise, it is only used to read the disk label which provides
1993-03-21 12:45:37 +03:00
a set of configuration parameters for the memory based file system.
2001-06-07 18:52:50 +04:00
The special file is typically that of the first swap area, since
1998-04-29 13:49:10 +04:00
that is where the file system will be backed up when free memory
gets low and the memory supporting the file system has to be paged.
If the keyword ``swap'' is used instead of a special file name,
default configuration parameters will be used.
(This option is useful when trying to use
.Nm mount_mfs
1998-04-29 13:49:10 +04:00
on a machine without any disks).
1993-03-21 12:45:37 +03:00
.Pp
Options with numeric arguments may contain an optional (case-insensitive)
suffix:
.Bl -tag -width 3n -offset indent -compact
.It b
Bytes; causes no modification. (Default)
.It k
Kilo; multiply the argument by 1024
.It m
Mega; multiply the argument by 1048576
.It g
Giga; multiply the argument by 1073741824
.El
.Pp
1993-03-21 12:45:37 +03:00
The following options define the general layout policies.
.Bl -tag -width Fl
.It Fl B Ar byte-order
Specify the metadata byte order of the file system to be created.
1998-04-29 13:49:10 +04:00
Valid byte orders are `be' and `le'.
If no byte order is specified, the file system is created in host
1998-04-29 13:49:10 +04:00
byte order.
.It Fl F
Create a file system image in the regular file referenced by
.Ar special .
The file system size needs to be specified with
.Dq Fl s Ar size .
1993-03-21 12:45:37 +03:00
.It Fl N
Causes the file system parameters to be printed out
without really creating the file system.
.It Fl O
1998-04-29 13:49:10 +04:00
Creates a
.Bx 4.3
format file system.
This option is primarily used to build root file systems
that can be understood by older boot ROMs.
.It Fl T Ar disk-type
Uses information for the specified disk from
.Pa /etc/disktab
instead of trying to get the information from a disklabel.
.It Fl Z
Pre-zeros the file system image created with
.Fl F .
This is necessary if the image is to be used by
.Xr vnd 4
(which doesn't support file systems with
.Sq holes ) .
1993-03-21 12:45:37 +03:00
.It Fl a Ar maxcontig
This specifies the maximum number of contiguous blocks that will be
laid out before forcing a rotational delay (see the
1993-03-21 12:45:37 +03:00
.Fl d
option).
The default value is 8.
1993-03-21 12:45:37 +03:00
See
.Xr tunefs 8
for more details on how to set this option.
.It Fl b Ar block-size
1998-04-29 13:49:10 +04:00
The block size of the file system, in bytes.
.It Fl c Ar cpg
The number of cylinders per cylinder group in a file system.
The default value is 16.
1993-03-21 12:45:37 +03:00
.It Fl d Ar rotdelay
This specifies the expected time (in milliseconds) to service a transfer
completion interrupt and initiate a new transfer on the same disk.
The default is 0 milliseconds.
1993-03-21 12:45:37 +03:00
See
.Xr tunefs 8
for more details on how to set this option.
.ne 1i
1993-03-21 12:45:37 +03:00
.It Fl e Ar maxbpg
This indicates the maximum number of blocks any single file can
allocate out of a cylinder group before it is forced to begin
allocating blocks from another cylinder group.
The default is about one quarter of the total blocks in a cylinder group.
See
.Xr tunefs 8
for more details on how to set this option.
.It Fl f Ar frag-size
The fragment size of the file system in bytes.
Incorporate the enhanced ffs_dirpref() by Grigoriy Orlov, as found in FreeBSD (three commits; the initial work, man page updates, and a fix to ffs_reload()), with the following differences: - Be consistent between newfs(8) and tunefs(8) as to the options which set and control the tuning parameters for this work (avgfilesize & avgfpdir) - Use u_int16_t instead of u_int8_t to keep track of the number of contiguous directories (suggested by Chuck Silvers) - Work within our FFS_EI framework - Ensure that fs->fs_maxclusters and fs->fs_contigdirs don't point to the same area of memory The new algorithm has a marked performance increase, especially when performing tasks such as untarring pkgsrc.tar.gz, etc. The original FreeBSD commit messages are attached: ===== mckusick 2001/04/10 01:39:00 PDT Directory layout preference improvements from Grigoriy Orlov <gluk@ptci.ru>. His description of the problem and solution follow. My own tests show speedups on typical filesystem intensive workloads of 5% to 12% which is very impressive considering the small amount of code change involved. ------ One day I noticed that some file operations run much faster on small file systems then on big ones. I've looked at the ffs algorithms, thought about them, and redesigned the dirpref algorithm. First I want to describe the results of my tests. These results are old and I have improved the algorithm after these tests were done. Nevertheless they show how big the perfomance speedup may be. I have done two file/directory intensive tests on a two OpenBSD systems with old and new dirpref algorithm. The first test is "tar -xzf ports.tar.gz", the second is "rm -rf ports". The ports.tar.gz file is the ports collection from the OpenBSD 2.8 release. It contains 6596 directories and 13868 files. The test systems are: 1. Celeron-450, 128Mb, two IDE drives, the system at wd0, file system for test is at wd1. Size of test file system is 8 Gb, number of cg=991, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=35 2. PIII-600, 128Mb, two IBM DTLA-307045 IDE drives at i815e, the system at wd0, file system for test is at wd1. Size of test file system is 40 Gb, number of cg=5324, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=50 You can get more info about the test systems and methods at: http://www.ptci.ru/gluk/dirpref/old/dirpref.html Test Results tar -xzf ports.tar.gz rm -rf ports mode old dirpref new dirpref speedup old dirprefnew dirpref speedup First system normal 667 472 1.41 477 331 1.44 async 285 144 1.98 130 14 9.29 sync 768 616 1.25 477 334 1.43 softdep 413 252 1.64 241 38 6.34 Second system normal 329 81 4.06 263.5 93.5 2.81 async 302 25.7 11.75 112 2.26 49.56 sync 281 57.0 4.93 263 90.5 2.9 softdep 341 40.6 8.4 284 4.76 59.66 "old dirpref" and "new dirpref" columns give a test time in seconds. speedup - speed increasement in times, ie. old dirpref / new dirpref. ------ Algorithm description The old dirpref algorithm is described in comments: /* * Find a cylinder to place a directory. * * The policy implemented by this algorithm is to select from * among those cylinder groups with above the average number of * free inodes, the one with the smallest number of directories. */ A new directory is allocated in a different cylinder groups than its parent directory resulting in a directory tree that is spreaded across all the cylinder groups. This spreading out results in a non-optimal access to the directories and files. When we have a small filesystem it is not a problem but when the filesystem is big then perfomance degradation becomes very apparent. What I mean by a big file system ? 1. A big filesystem is a filesystem which occupy 20-30 or more percent of total drive space, i.e. first and last cylinder are physically located relatively far from each other. 2. It has a relatively large number of cylinder groups, for example more cylinder groups than 50% of the buffers in the buffer cache. The first results in long access times, while the second results in many buffers being used by metadata operations. Such operations use cylinder group blocks and on-disk inode blocks. The cylinder group block (fs->fs_cblkno) contains struct cg, inode and block bit maps. It is 2k in size for the default filesystem parameters. If new and parent directories are located in different cylinder groups then the system performs more input/output operations and uses more buffers. On filesystems with many cylinder groups, lots of cache buffers are used for metadata operations. My solution for this problem is very simple. I allocate many directories in one cylinder group. I also do some things, so that the new allocation method does not cause excessive fragmentation and all directory inodes will not be located at a location far from its file's inodes and data. The algorithm is: /* * Find a cylinder group to place a directory. * * The policy implemented by this algorithm is to allocate a * directory inode in the same cylinder group as its parent * directory, but also to reserve space for its files inodes * and data. Restrict the number of directories which may be * allocated one after another in the same cylinder group * without intervening allocation of files. * * If we allocate a first level directory then force allocation * in another cylinder group. */ My early versions of dirpref give me a good results for a wide range of file operations and different filesystem capacities except one case: those applications that create their entire directory structure first and only later fill this structure with files. My solution for such and similar cases is to limit a number of directories which may be created one after another in the same cylinder group without intervening file creations. For this purpose, I allocate an array of counters at mount time. This array is linked to the superblock fs->fs_contigdirs[cg]. Each time a directory is created the counter increases and each time a file is created the counter decreases. A 60Gb filesystem with 8mb/cg requires 10kb of memory for the counters array. The maxcontigdirs is a maximum number of directories which may be created without an intervening file creation. I found in my tests that the best performance occurs when I restrict the number of directories in one cylinder group such that all its files may be located in the same cylinder group. There may be some deterioration in performance if all the file inodes are in the same cylinder group as its containing directory, but their data partially resides in a different cylinder group. The maxcontigdirs value is calculated to try to prevent this condition. Since there is no way to know how many files and directories will be allocated later I added two optimization parameters in superblock/tunefs. They are: int32_t fs_avgfilesize; /* expected average file size */ int32_t fs_avgfpdir; /* expected # of files per directory */ These parameters have reasonable defaults but may be tweeked for special uses of a filesystem. They are only necessary in rare cases like better tuning a filesystem being used to store a squid cache. I have been using this algorithm for about 3 months. I have done a lot of testing on filesystems with different capacities, average filesize, average number of files per directory, and so on. I think this algorithm has no negative impact on filesystem perfomance. It works better than the default one in all cases. The new dirpref will greatly improve untarring/removing/coping of big directories, decrease load on cvs servers and much more. The new dirpref doesn't speedup a compilation process, but also doesn't slow it down. Obtained from: Grigoriy Orlov <gluk@ptci.ru> ===== ===== iedowse 2001/04/23 17:37:17 PDT Pre-dirpref versions of fsck may zero out the new superblock fields fs_contigdirs, fs_avgfilesize and fs_avgfpdir. This could cause panics if these fields were zeroed while a filesystem was mounted read-only, and then remounted read-write. Add code to ffs_reload() which copies the fs_contigdirs pointer from the previous superblock, and reinitialises fs_avgf* if necessary. Reviewed by: mckusick ===== ===== nik 2001/04/10 03:36:44 PDT Add information about the new options to newfs and tunefs which set the expected average file size and number of files per directory. Could do with some fleshing out. =====
2001-09-06 06:16:00 +04:00
.It Fl g Ar avgfilesize
The expected average file size for the file system.
.It Fl h Ar avgfpdir
The expected average number of files per directory on the file system.
.It Fl i Ar bytes-per-inode
1993-03-21 12:45:37 +03:00
This specifies the density of inodes in the file system.
The default is to create an inode for each 4096 bytes of data space.
1993-03-21 12:45:37 +03:00
If fewer inodes are desired, a larger number should be used;
to create more inodes a smaller number should be given.
.It Fl m Ar free-space
The percentage of space reserved from normal users; the minimum free
space threshold.
The default value used is 5%.
See
.Xr tunefs 8
for more details on how to set this option.
.It Fl n Ar rotational-positions
Determines how many rotational time slots there are in
one revolution of the disk.
.It Fl o Ar optimization
Optimization preference; either
.Dq space
or
.Dq time .
The file system can either be instructed to try to minimize the time spent
allocating blocks, or to try to minimize the space fragmentation on the disk.
If the value of minfree (see above) is less than 5%,
the default is to optimize for space;
if the value of minfree is greater than or equal to 5%,
the default is to optimize for time.
See
.Xr tunefs 8
for more details on how to set this option.
1993-03-21 12:45:37 +03:00
.It Fl s Ar size
The size of the file system in sectors.
An
.Sq s
suffix will be interpreted as the number of sectors (the default).
All other suffixes are interpreted as per other numeric arguments,
except that the number is converted into sectors by dividing by the
sector size (as specified by
.Fl S Ar secsize )
after suffix interpretation.
.El
1993-03-21 12:45:37 +03:00
.Pp
1998-04-29 13:49:10 +04:00
The following options override the standard sizes for the disk geometry.
1993-03-21 12:45:37 +03:00
Their default values are taken from the disk label.
Changing these defaults is useful only when using
.Nm
to build a file system whose raw image will eventually be used on a
different type of disk than the one on which it is initially created
(for example on a write-once disk).
Note that changing any of these values from their defaults will make
1998-04-29 13:49:10 +04:00
it impossible for
.Xr fsck_ffs 8
1997-03-08 16:47:08 +03:00
to find the alternative superblocks if the standard superblock is lost.
.Bl -tag -width Fl
1993-03-21 12:45:37 +03:00
.It Fl S Ar sector-size
The size of a sector in bytes (almost never anything but 512).
Defaults to 512.
.It Fl k Ar skew
Sector \&0 skew, per track.
Used to describe perturbations in the media format to compensate for
a slow controller.
Track skew is the offset of sector 0 on track N relative to sector 0
on track N-1 on the same cylinder.
.It Fl l Ar interleave
Hardware sector interleave.
Used to describe perturbations in the media format to compensate for
a slow controller.
Interleave is physical sector interleave on each track,
specified as the denominator of the ratio:
.Dl sectors read/sectors passed over
Thus an interleave of 1/1 implies contiguous layout, while 1/2 implies
logical sector 0 is separated by one sector from logical sector 1.
.It Fl p Ar track-spares
Spare sectors per track.
Spare sectors (bad sector replacements) are physical sectors that occupy
space at the end of each track.
They are not counted as part of the sectors per track.
1993-03-21 12:45:37 +03:00
.Pq Fl u
since they are not available to the file system for data allocation.
.It Fl r Ar revolutions
The speed of the disk in revolutions per minute.
.ne 1i
.It Fl t Ar ntracks
The number of tracks per cylinder available for data allocation by the file
system.
.It Fl u Ar nsectors
The number of sectors per track available for data allocation by the file
system.
This does not include sectors reserved at the end of each track for bad
block replacement (see the
.Fl p
1998-04-29 13:49:10 +04:00
option).
.It Fl x Ar spare-sectors-per-cylinder
Spare sectors (bad sector replacements) are physical sectors that occupy
space at the end of the last track in the cylinder.
They are deducted from the sectors per track.
1993-03-21 12:45:37 +03:00
.Pq Fl u
of the last track of each cylinder since they are not available to the file
system for data allocation.
1993-03-21 12:45:37 +03:00
.El
.Pp
The options to the
.Nm mount_mfs
command are as described for the
.Nm
command, except for the
.Fl o
option.
.Pp
That option is as follows:
.Bl -tag -width indent
.It Fl o
Options are specified with a
.Fl o
flag followed by a comma separated string of options.
See the
1993-03-21 12:45:37 +03:00
.Xr mount 8
man page for possible options and their meanings.
1993-03-21 12:45:37 +03:00
.El
.Sh NOTES
1998-05-07 07:40:48 +04:00
If the file system will be exported over NFS, the
.Xr fsirand 8
utility should be run after
.Nm
to improve security.
.Pp
The owner and group ids of the root node of the new file system
are set to the effective uid and gid of the user initializing
the file system.
.Pp
For the
.Nm
command to succeed,
the disklabel should first be updated such that the fstype field for the
partition is set to
.Bx 4.2 .
1993-03-21 12:45:37 +03:00
.Sh SEE ALSO
.Xr disktab 5 ,
.Xr fs 5 ,
.Xr dumpfs 8 ,
1993-03-21 12:45:37 +03:00
.Xr disklabel 8 ,
.Xr diskpart 8 ,
.\" .Xr format 8 ,
1998-05-07 07:40:48 +04:00
.Xr fsck_ffs 8 ,
.Xr fsirand 8 ,
.Xr mount 8 ,
1993-03-21 12:45:37 +03:00
.Xr tunefs 8
.Rs
.%A M. McKusick
.%A W. Joy
.%A S. Leffler
.%A R. Fabry
.%T A Fast File System for UNIX ,
.%J ACM Transactions on Computer Systems 2
.%V 3
.%P pp 181-197
.%D August 1984
.%O (reprinted in the BSD System Manager's Manual)
.Re
.Sh HISTORY
The
.Nm
command appeared in
.Bx 4.2 .