|
|
1.1 ! root 1: .\" Copyright (c) 1983 Regents of the University of California. ! 2: .\" All rights reserved. The Berkeley software License Agreement ! 3: .\" specifies the terms and conditions for redistribution. ! 4: .\" ! 5: .\" @(#)tunefs.8 6.5 (Berkeley) 6/24/90 ! 6: .\" ! 7: .TH TUNEFS 8 "June 24, 1990" ! 8: .UC 5 ! 9: .SH NAME ! 10: tunefs \- tune up an existing file system ! 11: .SH SYNOPSIS ! 12: .B tunefs ! 13: .I tuneup-options ! 14: .IR special | filesys ! 15: .SH DESCRIPTION ! 16: .I Tunefs ! 17: is designed to change the dynamic parameters of a file system ! 18: which affect the layout policies. ! 19: The parameters which are to be changed are indicated by the flags ! 20: given below: ! 21: .IP "\fB\-a\fP maxcontig" ! 22: .br ! 23: This specifies the maximum number of contiguous blocks that will ! 24: be laid out before forcing a rotational delay (see \-d below). ! 25: The default value is one, since most device drivers require ! 26: an interrupt per disk transfer. ! 27: Device drivers that can chain several buffers together in a single ! 28: transfer should set this to the maximum chain length. ! 29: .IP "\fB\-d\fP rotdelay" ! 30: .br ! 31: This specifies the expected time (in milliseconds) ! 32: to service a transfer completion ! 33: interrupt and initiate a new transfer on the same disk. ! 34: It is used to decide how much rotational spacing to place between ! 35: successive blocks in a file. ! 36: .IP "\fB\-e\fP maxbpg" ! 37: .br ! 38: This indicates the maximum number of blocks any single file can ! 39: allocate out of a cylinder group before it is forced to begin ! 40: allocating blocks from another cylinder group. ! 41: Typically this value is set to about one quarter of the total blocks ! 42: in a cylinder group. ! 43: The intent is to prevent any single file from using up all the ! 44: blocks in a single cylinder group, ! 45: thus degrading access times for all files subsequently allocated ! 46: in that cylinder group. ! 47: The effect of this limit is to cause big files to do long seeks ! 48: more frequently than if they were allowed to allocate all the blocks ! 49: in a cylinder group before seeking elsewhere. ! 50: For file systems with exclusively large files, ! 51: this parameter should be set higher. ! 52: .IP "\fB\-m\fP minfree" ! 53: .br ! 54: This value specifies the percentage of space held back ! 55: from normal users; the minimum free space threshold. ! 56: The default value used is 10%. ! 57: This value can be set to zero, however up to a factor of three ! 58: in throughput will be lost over the performance obtained at a 10% ! 59: threshold. ! 60: Note that if the value is raised above the current usage level, ! 61: users will be unable to allocate files until enough files have ! 62: been deleted to get under the higher threshold. ! 63: .IP "\fB\-o\fP optimization preference" ! 64: .br ! 65: The file system can either try to minimize the time spent ! 66: allocating blocks, or it can attempt minimize the space ! 67: fragmentation on the disk. ! 68: If the value of minfree (see above) is less than 10%, ! 69: then the file system should optimize for space to avoid ! 70: running out of full sized blocks. ! 71: For values of minfree greater than or equal to 10%, ! 72: fragmentation is unlikely to be problematical, and ! 73: the file system can be optimized for time. ! 74: .SH "SEE ALSO" ! 75: fs(5), ! 76: newfs(8), ! 77: mkfs(8) ! 78: .PP ! 79: M. McKusick, W. Joy, S. Leffler, R. Fabry, ! 80: ``A Fast File System for UNIX'', ! 81: \fIACM Transactions on Computer Systems 2\fP, 3. ! 82: pp 181-197, August 1984. ! 83: (reprinted in the System Manager's Manual, SMM:14) ! 84: .SH BUGS ! 85: This program should work on mounted and active file systems. ! 86: Because the super-block is not kept in the buffer cache, ! 87: the changes will only take effect if the program ! 88: is run on dismounted file systems. ! 89: To change the root file system, the system must be rebooted ! 90: after the file system is tuned. ! 91: .PP ! 92: You can tune a file system, but you can't tune a fish.
This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.