diff options
Diffstat (limited to 'share/man/man4/vinum.4')
| -rw-r--r-- | share/man/man4/vinum.4 | 120 |
1 files changed, 57 insertions, 63 deletions
diff --git a/share/man/man4/vinum.4 b/share/man/man4/vinum.4 index 4555b17..52407df 100644 --- a/share/man/man4/vinum.4 +++ b/share/man/man4/vinum.4 @@ -45,6 +45,7 @@ .Nd Logical Volume Manager .Sh SYNOPSIS .Cd "modload /lkm/vinum_mod.o" +.Cd "modload /lkm/vinum_mod.raid5.o" .Sh DESCRIPTION .Nm is a logical volume manager inspired by, but not derived from, the Veritas @@ -116,20 +117,20 @@ SD2: box The subdisks of a striped plex must all be the same size. .It \fIRAID-5 plexes\fP\| require at least three equal-sized subdisks. They -resemble striped plexes, except that in each slice, one subdisk stores parity -information. This subdisk changes in each slice: in the first slice, it is the +resemble striped plexes, except that in each stripe, one subdisk stores parity +information. This subdisk changes in each stripe: in the first stripe, it is the first subdisk, in the second it is the second subdisk, etc. In the event of a single disk failure, .Nm will recover the data based on the information stored on the remaining subdisks. This mapping is particularly suited to read-intensive access. The subdisks of a -RAID5 plex must all be the same size. +RAID-5 plex must all be the same size. .\" Make sure to flush! .El .It .Nm Drives -are the lowest level of the storage hierarchy. They represent either complete -disks or disk partitions. +are the lowest level of the storage hierarchy. They represent disk special +devices. .It .Nm offers automatic startup. Unlike UNIX file systems, @@ -150,29 +151,45 @@ necessary to match the LKM to the version of the operating system. Failure to do so will cause .Nm to issue an error message and terminate. +.Pp +.Nm +is currently available in two versions: a freely available version which does +not contain RAID-5 functionality, and a full version including RAID-5 +functionality, which is available from Cybernet Systems Inc. .Sh RUNNING VINUM -The +The freely available version of the .Nm LKM is called -.Pa /lkm/vinum_mod.o . -To start it, load the module and read in the configuration: +.Pa /lkm/vinum_mod.o , +and the RAID-5 version is +.Pa /lkm/vinum_mod.raid5.o . +To load the module: .Bd -unfilled -offset indent # modload /lkm/vinum_mod.o -# vinum read /dev/sd1h .Ed -.sp -The first command loads the LKM, and the second reads an existing +.Pp +At this point, .Nm -configuration from the device -.Ar /dev/sd1h . -You should omit this command if there is no configuration on disk. The name of -the disk slice (in this example -.Ar /dev/sd1h ) -can be any of the slices used by +is loaded but has not been configured. In an existing installation, the +following command reads the configuration from disk +.Ar /dev/da1h . +.Bd -unfilled -offset indent +# vinum read /dev/da1h +.Ed +.sp +The name of the disk device (in this example +.Ar /dev/da1h ) +can be any of the devices used by .Nm vinum . These commands are normally embedded in the startup file .Pa /etc/rc . .Pp +See +.Xr vinum 8 +for information on how to create a +.Nm +configuration. +.Pp To unload the LKM, first find the .Ar Id field in @@ -199,19 +216,19 @@ Use the utility to configure and start .Nm objects. -.Ss IOCTL CALLS +.Sh IOCTL CALLS .Pa ioctl calls are intended for the use of the .Nm configuration program only. The are described in the header file .Pa /sys/sys/vinumio.h .Ss DISK LABELS -Conventional disk partitions have a +Conventional disk special devices have a .Em disk label -in the second sector of the partitions. See +in the second sector of the device. See .Xr disklabel 5 -for more details. This disk label describes the layout of the slices within the -partition. +for more details. This disk label describes the layout of the partitions within +the device. .Nm does not subdivide volumes, so volumes do not contain a physical disk label. For convenience, @@ -241,22 +258,21 @@ three partitions, a, b and c, all the same except for the fstype, for example: .Nm ignores the DIOCWDINFO and DIOCSDINFO ioctls, since there is nothing to change. As a result, any attempt to modify the disk label will be silently ignored. -.Ss MAKING FILE SYSTEMS +.Sh MAKING FILE SYSTEMS Since .Nm -volumes do not contain slices, the names do not need to conform to the standard -rules for naming disk slices. For a physical disk slice, the last letter of the -device name specifies the slice identifier (a to h). +volumes do not contain partitions, the names do not need to conform to the +standard rules for naming disk partitions. For a physical disk partition, the +last letter of the device name specifies the partition identifier (a to h). .Nm volumes need not conform to this convention, but if they do not, .Nm newfs -will complain that it cannot determine the slice. To solve this problem, use -the +will complain that it cannot determine the partition. To solve this problem, +use the .Fl v flag to .Nm newfs . -.Ss OBJECT NAMING -.Pp +.Sh OBJECT NAMING .Nm assigns default names to plexes and subdisks, although they may be overridden. We do not recommend overriding the default names. Experience with the @@ -274,8 +290,8 @@ name from which they are derived. .Bl -bullet .It When -.Nm -starts, it creates a directory +.Nm vinum(8) +creates or deletes objects, it creates a directory .Ar /dev/vinum , in which it makes device entries for each volume it finds. It also creates subdirectories, @@ -288,22 +304,18 @@ creates two more directories, and .Ar /dev/vinum/drive , in which it stores hierarchical information for volumes and drives. -.Pp -.Nm -assigns names for plexes and subdisks automatically. They are derived from the -names of the object to which they are attached. .It Unlike .Nm UNIX drives, .Nm -volumes are not subdivided into slices, and thus do not contain a partition -(slice) table. Unfortunately, this confuses a number of utilities, notably +volumes are not subdivided into partitions, and thus do not contain a disk +label. Unfortunately, this confuses a number of utilities, notably .Nm newfs , which normally tries to interpret the last letter of a .Nm -volume name as a slice identifier. If you use a volume name which does not end -in the letters +volume name as a partition identifier. If you use a volume name which does not +end in the letters .Ar a to .Ar c , @@ -314,7 +326,7 @@ flag to in order to tell it to ignore this convention. .\" .It -Plexes do not need to be assigned names manually. By default, a plex name is +Plexes do not need to be assigned explicit names. By default, a plex name is the name of the volume followed by the letters \f(CW.p\fR and the number of the plex. For example, the plexes of volume .Ar vol3 @@ -324,7 +336,7 @@ are called and so on. These names can be overridden, but it is not recommended. .br .It -Like plexes, subdisks are assigned names automatically, and manual naming is +Like plexes, subdisks are assigned names automatically, and explicit naming is discouraged. A subdisk name is the name of the plex followed by the letters \f(CW.s\fR and a number identifying the subdisk. For example, the subdisks of plex @@ -499,10 +511,6 @@ crwxr-xr-- 1 root wheel 91, 0 Mar 30 16:08 rtinyvol crwxr-xr-- 1 root wheel 91, 4 Mar 30 16:08 rvol5 .Ed .Pp -See -.Xr vinum 4 -for a description of the minor number format. -.Pp In the case of unattached plexes and subdisks, the naming is reversed. Subdisks are named after the disk on which they are located, and plexes are named after the subdisk. @@ -789,13 +797,13 @@ drive_up#up and running .Sh BUGS AND OMISSIONS Many. .Nm vinum -is currently in alpha test. Please report any bugs not in the list below to +is currently in beta test. Please report any bugs not in the list below to .Ar <grog@lemis.com> . .sp The following functions are known to be deficient or not implemented: .Bl -bullet .It -It is necessary to initialize RAID5 plexes. Failure to do so will not impede +It is necessary to initialize RAID-5 plexes. Failure to do so will not impede normal operation, but it will cause complete corruption if one of the disks should fail. I don't know any good way to enforce this initialization (or the even slower alternative of rebuilding the parity blocks). If anybody has a good @@ -803,27 +811,13 @@ idea, I'd be grateful for input. .It Detection of differences between the version of the kernel and the LKM is not yet implemented. -.It -Detaching plexes and subdisks has not yet been implemented. -.It -Reintegration of failed disks has not yet been implemented. -.It -.Nm -requires a special version of -.Ar newfs , -which has not yet been committed. The current version places some restrictions -on volume names\(emsee above. -.It -Anonymous plexes and subdisks (which are not associated with a volume or plex -respectively) are currently not supported. This also means that detaching an -object is not supported. .El .Sh AUTHOR Greg Lehey .Pa <grog@lemis.com> . .Sh HISTORY .Nm vinum -first appeared in FreeBSD 2.2.6. +first appeared in FreeBSD 3.0. .Sh SEE ALSO .Xr vinum 8 , .Xr disklabel 5 , |
