Use the rmvdisk command to delete a
volume. This command cannot be
used for high availability volumes. Use the rmvolume command
for high availability volumes.
Syntax
rmvdisk { [ -removehostmappings ] | [ -force ] } { vdisk_id | vdisk_name }
Parameters
- -force
- (Optional) Deletes the specified volume, even if mappings still
exist between this volume and one or more hosts. This parameter deletes
any host-to-volume mappings and any FlashCopy mappings that
exist for this volume. If the -force deletion
of a volume causes dependent mappings to be stopped, any target volumes
for those mappings that are in Metro Mirror or Global Mirror relationships
are also stopped. The dependent mappings can be identified by
using the lsvdiskdependentmaps command on the volume
that you want to delete.
Note: Using the -force parameter
might result in a data loss. Use it only under the direction of your
product support information, or if you are willing to accept the risk
of data loss in the volume to which the drive belongs.
- -removehostmappings
- (Optional) Removes all host mappings for the specified volume
before deleting the volume itself.
Note: Using the -removehostmappings parameter
might result in a data loss. Use it only under the direction of your
product support information, or if you are willing to accept the risk
of data loss in the volume to which the drive belongs.
- vdisk_id | vdisk_name
- Specifies the name of the volume to delete, either by ID or by
name.
Note: To deactivate
compression, use the rmvdiskcopy to delete the
last compressed volume copy for an I/O group.
Description
This
command deletes an existing managed mode volume or an existing image
mode volume. The extents that made up this volume are returned to
the pool of free extents that are available on the managed disk group,
if the volume is in managed mode.
Remember: If
you issue this command, any data that was on the volume is lost. Before
you issue this command, ensure that the volume (and any data that
resides on it) is no longer required.
This
command is unsuccessful if:
- Volume protection is enabled (using the chsystem command)
- The volume being removed has received I/O within the defined volume
protection time period
With an active-active relationship,
either or both of the master and auxiliary volumes can provide the
information for host systems to read through the master volume ID.
To remove the auxiliary volume from the relationship, delete the relationship
so hosts can access the master copy.
Remember: Any FlashCopy mappings
with the specified volume as their source volume are deleted when
you specify this command.
Deleting a managed mode volume
When
you use this command to delete a managed mode volume, all the data
on the volume is deleted. The extents that make up the volume are
returned to the pool of free extents that are available in the managed
disk group.
If host mappings
exist for the volume, or if any FlashCopy mappings
would be affected, the deletion fails. You can use the -force parameter
to force the deletion. If you use the -force parameter,
mappings that have the volume as source or target are deleted, other
mappings in a cascade might be stopped, and then the volume is deleted.
The -force parameter also deletes any Metro Mirror
or Global Mirror relationships that exist for the specified volume
(and any information that is not staged in the fast write cache).
If
the volume is in the process of migrating to an image mode volume
(using the migratetoimage command), the deletion
fails unless you use the -force parameter. If
you use the -force parameter, the migration is
halted and then the volume is deleted. Before you issue this command,
ensure that the volume (and any data that resides on it) is no longer
required.
Deleting an image mode volume
If
the volume is mirrored and one or both copies is in image mode, you
must first wait for all fast-write data to be moved to the controller
logical unit. This ensures that the data on the controller is consistent
with the data on the image mode volume before the volume is deleted.
This process can take several minutes to complete, and is indicated
by the fast_write_state state of the volume being empty. If
the -force parameter is specified, the fast-write
data is discarded and the volume is deleted immediately; the data
on the controller logical unit is left inconsistent and unusable. If
the copies are not synchronized, you must use the -force parameter.
If
you run the command while data is in the cache, the system attempts
to move the data out of the cache; this process can time out, however.
If
there are any virtual medium errors on the volume, the command fails.
You can force the deletion by using the
-force parameter;
however, this can cause data integrity
problems.
Note: A virtual
medium error occurs when you copy data from one disk (the source)
to another (the target). Reading the source indicates that there is
a medium error. At that moment, you must have two identical copies
of data and you must then simulate a medium error on the target disk.
You can simulate a medium error on the target disk by creating a virtual
medium error on the target disk.
If FlashCopy mappings
or host mappings exist for the volume, the deletion fails unless
you use the -force parameter. If you use the -force parameter,
mappings are deleted and the volume is deleted. If there is any data
that is not staged in the fast write cache for this volume, the deletion
of the volume fails. When the -force parameter
is specified, any data that is not staged in the fast write cache
is deleted. Deleting an image mode volume causes the managed disk
that is associated with the volume to be removed from the managed
disk group. The mode of the managed disk is returned to unmanaged.
If the relationship
is in consistent_copying or consistent_stopped state, and the change
volume is being used by a Global Mirror relationship using multicycling
mode, the relationship moves to inconsistent_copying or inconsistent_stopped
state.
Note: If the relationship is part of a consistency group entire
group is affected by this state transition.
The secondary volume
becomes corrupt, and inaccessible for host input/output I/O data if:
- A changed volume is part of an idling relationship
- The changed volume is being used for secondary protection
- The background copy process is still migrating the change volume
data to the secondary volume
You must issue
recovervdisk to
gain access to the volume contents once more. If a change volume
was part of an
idling relationship and being
used for Global Mirror relationship using multicycling mode, and the
relationship was deleted but the background copy process continued
and is still migrating data to the secondary volume then the secondary
volume also becomes corrupt. In any of these cases, this
recovervdisk fails
without
-force being specified.
Note: - The -force parameter must be used if rmvdisk is
specified and rejected if the volume is a change volume for a relationship.
- If the volume is a change volume for a relationship, specifying rmvdisk with -force removes
the change volume from the relationship.
An invocation example
rmvdisk -force vdisk5
The
resulting output:
No feedback
An
invocation example
rmvdisk -removehostmappings vdisk3
The
resulting output:
No feedback