Shell Commands › Volume Management › create (Volume Management)--Create a New Volume for an Application

create (Volume Management)--Create a New Volume for an Application

Create a new application or global volume

Syntax

vol create <name> (mkfs=follow these steps: | fs=follow these steps:) [ <prop>=<val> ]* [ -n ] [--prefill] [ --batch ] 

Options

<name>

Name of the volume to create. The format of <name> is the following: <application>:[<class>.]<volume>, _GLOBAL:<volume>

or _GLOBAL_RO:<volume>.

follow these steps:

File system type (for example, ext3)

<prop>=<val>

Volume parameters. The following parameters may be set:

size

Volume size, units may be specified (for example, 128M[B], 2G[B]). If units are not specified, it is assumed that the value is in MB. Default is 1G

comment

Volume comment. This parameter is not supported if the 'link' parameter is specified.

fs_options

Additional file system options used to format the volume (See notes below for specifics)

mirrored

Volume is mirrored

link

Name of application user or global volume to which the volume is to be linked. The format of the link is the following:

<application>:<volume>

or _GLOBAL:<volume>.

parN.*

Partitioning parameters. (See notes below for specifics)

-n

Skip verification that volume reference is in the class descriptor if adding an application singleton class volume.

--prefill

Allocate ALL blocks of volume.

--batch

Display output in UDL format.

Notes:

• Only grid maintainers may create global read-only volumes or symbolic links to global read-only volumes (that is, GLOBAL_RO:<volume>).
• The following file systems are currently supported: ext2, ext3, fat16, fat32, ntfs, ntfs03, ntfs08, reiserfs, swap, ext3-snapshot, ufssol, zfs, and raw. Although ntfs and ntfs03 are supported filesystem names on the command line, the resulting volume is formatted as ntfs08.

  The following describes the format of the partitioning parameters:

  parN.size

  Specifies the size for partition N. Default: (disk_size - total_size_of_partitions_with_size_specified) / number_of_partitions_with_no_size_specified.

  parN.fs

  Specifies the filesystem type for partition N. Supports the same values as the fs property and also adds lvm and Extended. If the value is Extended, the partition will be of type extended and can hold logical volumes, no fs, fs_options, vg or lv properties can be user for this partition. To define logical partitions for an extended partition, use the parN second level property. fs=ext3-snapshot is now deprecated, the preferred way is to use fs=lvm. Default: raw

  parN.fs_options

  Options to pass to mkfs when creating the filesystem on the partition. Default: none

  parN.parN

  If a partition is specified as fs=Extended, this second level property can be used to create logical partitions on that extended partition.Any 2nd level property can be used as a 3rd level property for a logical partition. Example par3.fs=Extended par3.par1.fs=ext3.

  parN.vg

  The partition is created with fs=lvm, this can be used to specify the volume group for that lvm partition. If no vg is specified, a random string will be used.

  parN.<lv_name>.size

  Size in [G,M,T] Bytes, default is (partition_size - total_size_of_logical_volumes_with_size_specified) / number_of_logical_volumes_with_no_size_specified

  parN.<lv_name>.fs

  Specifies the filesystem type for this logical volume. Allowed values are any fs handled by the linux Filer.

  parN.<lv_name>.fs_options

  Options to pass to mkfs when creating the filesystem on the logical volume. The value specified for the fs_options parameter is a space separated list, is file system specific, and is valid only with the mkfs parameter. The following are the options available for each supported filesystem:

  ext2 | ext3 | swap | reiserfs

  <any option> - any option understood by mkfs.<filesystem>. The options will be passed directly to the mkfs tool for the chosen filesystem.

  ext3-snapshot

  vol_group_name - string specifying the name of the LVM volume group to create on the volume. If not specified, a volume group name is randomly generated.

  data_percentage - percentage of the volume that is used to store data, remaining portion of the volume is for snapshots. If not specified, 80% of the volume is used for data.

  <any option> - any option understood by mkfs.ext3. The options will be passed directly to the mkfs tool and will be used for the data volume.

  zfs

  pool_name - the name of the pool - defaults to volume name if this is not set

  mountpoint - mount path of the pool root (or "legacy" or "none", that is, mountpoint=legacy) - defaults to /pool_name

  autoreplace - on/off - controls automatic device replacement - defaults to off

  delegation - on/off - controls whether a non-privileged user is granted access based on the dataset permissions defined on the dataset - defaults to on

  failmode - wait/continue/panic - defaults to wait

  version - 1-10 - defaults to 10 (current)

  ntfs, ntfs03, ntfs08

  volume_label - the volume label for the dst volume. If empty, the vol_name property value is used instead.

  active - create the new partition as active (bootable). Valid values are yes and no. If omitted, the default value of no is used during format while the value defaults to the src volume type during fscopy

Examples

vol create my-app:data fs=ext3

create 1G application user volume

vol create my-app:data fs=ext3 fs_options="-b 4096 -j -J size=180 -m 0"

create 1G application user volume specifying block size, journal size and reserved blocks.

vol create my-app:DBASE.boot fs=ext3 size=256M -n

create 256M application singleton class boot volume and don't check that volume reference exists in the class descriptor of DBASE.

vol create my-app:data fs=ext3-snapshot size=1G fs_options="vol_group_name=VGVOL data_percentage=75 -b 2048"

create 1G application user volume of type ext3-snapshot, with block size set to 2048 for the fs on the data volume.

vol create my-app:data size=5G par1.fs=ext2 par1.size=100M par2.fs=lvm par2.vg=VOLGR00 par2.root_lv.size=2GB par2.root_lv.fs=ext3 par2.home_lv.fs=ext3

Create 5GB disk with 2 primary partitions, first one is of size 100MB and ext2 formated, second one uses the rest of the disk space and has lvm on it with two volumes - one for / (using 2GB) and one for /home (using the rest of the partition space) - this is a typical setup that distro installers do by default

vol create my-app:data size=10G par1.fs=ext2 par2.fs=ext3 par3.fs=ext3 par4.fs=ext3

Create 10GB disk with 4 partitions

vol create my-app:data size=10G par1.fs=ext2 par2.fs=ext3 par3.fs=ext3 par4.fs=Extended par4.par1.fs=ext3

Same thing as above, but the last partition is logical (it will actually create 5 partitions, 3 primary, 1 extended and 1logical which resides on the extended)

This section contains the following topics:

volume create Command Output Specification