xlv_assemble - initialize logical volume objects from disk labels
xlv_assemble [ -h name ] [ -alfnq ] [ -r root ] [ -s list ] [ -tvKP ]
xlv_assemble scans all the disks attached to the local system for logical
volume labels. It assembles all the available logical volumes and
generates a configuration data structure. xlv_assemble also creates the
device nodes for all XLV volumes in /dev/xlv and /dev/rxlv. The kernel
is then activated with the newly created configuration data structure.
If necessary, xlv_assemble also asks the xlv_plexd(1M) daemon to perform
any necessary plex revives.
xlv_assemble is automatically run on system startup from a script in the
/etc/init.d/xlv directory. By default, it is also automatically run
after you run xlv_make(1M).
xlv_assemble supports the following options:
-h name Use name as the local nodename. Every logical volume label
contains a system nodename. See the -l option below.
This option has no effect if name is the local nodename.
-a Assemble all logical volumes. See the -l option below.
-f Do a full revive instead of an incremental revive. A full
revive means that all volumes that are not marked clean will be
revived. The normal operation is an incremental revive. In
this case, xlv_assemble only revives those volumes that are not
clean and that have not already been assembled.
-l Only assemble logical volumes that are owned by the local
system. Local logical volumes have the local nodename in their
logical volume labels. This is the default option.
When volumes are created via xlv_make(1M) the local system
nodename is recorded in the XLV disk label in each disk's
volume header (file named xlvlab). When xlv_assemble runs and
notices that the nodename in the XLV disk label does not match
the local nodename, the XLV device nodes that are created have
the nodename in the XLV disk label, followed by a dot,
prepended to the volume names.
Use the -h option to change the local nodename for this instant
Use xlv_mgr(1M) command change nodename to change the nodename
associated with the disk.
-n Scan all disks for logical volume labels, but don't save the
logical volume configuration and don't activate the kernel with
-q Proceed quietly and don't display status messages after putting
together the logical volume configuration.
-r root Use root as the root directory. This is used in the miniroot
when / is mounted as /root.
-s list Selectively assemble the comma separated list of volume names.
This provides a means of incrementally assembling additional
configurations. The logical volumes specified in list are
assembled regardless of whether or not the labels contain a
local system nodename.
-t Display terse status messages naming the logical volumes
-v Display verbose status messages about the logical volumes
-K Don't activate the kernel with this logical volume
-P Don't initiate plex revives on the logical volumes configured.
The default behavior of xlv_assemble can be changed with the environment
variable XLV_ASSEMBLE_ARGS. The environment variable is parsed before
the command line options.
xlv_assemble changes volume element states. The volume element states
empty The volume element has no user data. When a volume element is
initially created, it is in this state.
stale The volume element had user data but the data is now out-ofdate.
The volume element will be revived from an active
volume element. When xlv_assemble is initially invoked on a
mirrored volume element, one volume element is picked,
depending volume element state and timestamp, to be the
active volume element and the other volume elements are
marked stale. A plex revival is initiated to make volume
active The volume element is online and up-to-date.
clean The volume element has been shutdown cleanly. All other
volume elements spanning this address space, if any, have
offline The volume element was active but XLV encounter I/O errors on
this volume element so the volume element is changed to the
offline state. No further I/O is performed on that volume
incomplete The volume element is missing some disk parts.
xlv_labd(1M), xlv_make(1M), xlv_mgr(1M), xlv_plexd(1M), xlv_shutdown(1M),
You must be root to run xlv_assemble.
PPPPaaaaggggeeee 3333 [ Back ]