xref: /freebsd/share/man/man4/ccd.4 (revision 51a9219f5780e61e1437d25220bf8750d9df7f8b)
1.\"	$NetBSD: ccd.4,v 1.5 1995/10/09 06:09:09 thorpej Exp $
2.\"
3.\" Copyright (c) 1994 Jason Downs.
4.\" Copyright (c) 1994, 1995 Jason R. Thorpe.
5.\" All rights reserved.
6.\"
7.\" Redistribution and use in source and binary forms, with or without
8.\" modification, are permitted provided that the following conditions
9.\" are met:
10.\" 1. Redistributions of source code must retain the above copyright
11.\"    notice, this list of conditions and the following disclaimer.
12.\" 2. Redistributions in binary form must reproduce the above copyright
13.\"    notice, this list of conditions and the following disclaimer in the
14.\"    documentation and/or other materials provided with the distribution.
15.\" 3. All advertising materials mentioning features or use of this software
16.\"    must display the following acknowledgement:
17.\"	This product includes software developed for the NetBSD Project
18.\"	by Jason Downs and Jason R. Thorpe.
19.\" 4. Neither the name of the author nor the names of its contributors
20.\"    may be used to endorse or promote products derived from this software
21.\"    without specific prior written permission.
22.\"
23.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
24.\" IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
25.\" OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
26.\" IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
27.\" INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
28.\" BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
29.\" LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
30.\" AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
31.\" OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
32.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
33.\" SUCH DAMAGE.
34.\"
35.\" $FreeBSD$
36.\"
37.Dd August 9, 1995
38.Dt CCD 4
39.Os
40.Sh NAME
41.Nm ccd
42.Nd Concatenated Disk driver
43.Sh SYNOPSIS
44.Cd "device ccd"
45.Sh DESCRIPTION
46The
47.Nm
48driver provides the capability of combining one or more disks/partitions
49into one virtual disk.
50.Pp
51This document assumes that you are familiar with how to generate kernels,
52how to properly configure disks and devices in a kernel
53configuration file, and how to partition disks.
54.Pp
55Note that the
56.Dq raw
57partitions of the disks
58.Em should not
59be combined.
60The kernel will only allow component partitions of type
61.Dv FS_BSDFFS (type
62.Dq Li 4.2BSD
63as shown by
64.Xr disklabel 8 ) .
65.Pp
66In order to compile in support for the
67.Nm ,
68you must add a line similar
69to the following to your kernel configuration file:
70.Pp
71.Dl "device	ccd		# concatenated disk devices"
72.Pp
73As of the
74.Fx 3.0
75release, you do not need to
76configure your kernel with
77.Nm
78but may instead use it as a kernel loadable
79module.
80Simply running
81.Xr ccdconfig 8
82will load the module into the kernel.
83.Pp
84A
85.Nm
86may be either serially concatenated or interleaved.
87To serially
88concatenate the partitions, specify the interleave factor of 0.
89Note that mirroring may not be used with an interleave factor of 0.
90.Pp
91There is a run-time utility that is used for configuring
92.Nm Ns s .
93See
94.Xr ccdconfig 8
95for more information.
96.Ss The Interleave Factor
97If a
98.Nm
99is interleaved correctly, a
100.Dq striping
101effect is achieved, which can increase sequential read/write
102performance.
103The interleave factor is expressed in units of
104.Dv DEV_BSIZE
105(usually 512 bytes).
106For large writes, the optimum interleave factor
107is typically the size of a track, while for large reads, it is about a
108quarter of a track.
109(Note that this changes greatly depending on the
110number and speed of disks.)
111For instance, with eight 7,200 RPM drives
112on two Fast-Wide SCSI buses, this translates to about 128 for writes
113and 32 for reads.
114A larger interleave tends to work better when the
115disk is taking a multitasking load by localizing the file I/O from
116any given process onto a single disk.
117You lose sequential performance when
118you do this, but sequential performance is not usually an issue with a
119multitasking load.
120.Pp
121An interleave factor must be specified when using a mirroring configuration,
122even when you have only two disks (i.e. the layout winds up being the same
123no matter what the interleave factor).
124The interleave factor will determine
125how I/O is broken up, however, and a value 128 or greater is recommended.
126.Pp
127.Nm
128has an option for a parity disk, but does not currently implement it.
129.Pp
130The best performance is achieved if all component disks have the same
131geometry and size.
132Optimum striping cannot occur with different
133disk types.
134.Pp
135For random-access oriented workloads, such as news servers, a larger
136interleave factor (e.g., 65,536) is more desirable.
137Note that there
138is not much
139.Nm
140can do to speed up applications that are seek-time limited.
141Larger
142interleave factors will at least reduce the chance of having to seek
143two disk-heads to read one directory or a file.
144.Ss Disk Mirroring
145You can configure the
146.Nm
147to
148.Dq mirror
149any even number of disks.
150See
151.Xr ccdconfig 8
152for how to specify the necessary flags.
153For example, if you have a
154.Nm
155configuration specifying four disks, the first two disks will be mirrored with
156the second two disks.
157A write will be run to both sides of
158the mirror.
159A read will be run to either side of the mirror depending
160on what the driver believes to be most optimal.
161If the read fails,
162the driver will automatically attempt to read the same sector from the
163other side of the mirror.
164Currently
165.Nm
166uses a dual seek zone model to optimize reads for a multi-tasking load
167rather then a sequential load.
168.Pp
169In an event of a disk
170failure, you can use
171.Xr dd 1
172to recover the failed disk.
173.Pp
174Note that a one-disk
175.Nm
176is not the same as the original partition.
177In particular, this means
178if you have a filesystem on a two-disk mirrored
179.Nm
180and one of the disks fail, you cannot mount and use the remaining
181partition as itself; you have to configure it as a one-disk
182.Nm .
183You cannot replace a disk in a mirrored
184.Nm
185partition without first backing up the partition, then replacing the disk,
186then restoring the partition.
187.Sh WARNINGS
188If just one (or more) of the disks in a
189.Nm
190fails, the entire
191filesystem will be lost unless you are mirroring the disks.
192.Pp
193If one of the disks in a mirror is lost, you should still
194be able to back up your data.
195If a write error occurs, however, data
196read from that sector may be non-deterministic.
197It may return the data
198prior to the write or it may return the data that was written.
199When a
200write error occurs, you should recover and regenerate the data as soon
201as possible.
202.Pp
203Changing the interleave or other parameters for a
204.Nm
205disk usually destroys whatever data previously existed on that disk.
206.Sh FILES
207.Bl -tag -width ".Pa /dev/ccd*"
208.It Pa /dev/ccd*
209.Nm
210device special files
211.El
212.Sh HISTORY
213The concatenated disk driver was originally written at the University of
214Utah.
215.Sh SEE ALSO
216.Xr dd 1 ,
217.Xr ccdconfig 8 ,
218.Xr config 8 ,
219.Xr disklabel 8 ,
220.Xr fsck 8 ,
221.Xr MAKEDEV 8 ,
222.Xr mount 8 ,
223.Xr newfs 8 ,
224.Xr vinum 8
225