HP-UX vPars and Integrity VM V6.3 Administrator Guide

6.4.1.7 Dynamic addition of storage adapters
Starting with V6.3, vPars and Integrity VM storage adapters can be dynamically added to a running
vPar or VM guest. This is in addition to the existing ability to add new LUNs behind an existing
virtual adapter. This capability is available with both HPVM AVIO Storage adapters and HPVM
NPIV Storage adapters. In the case of AVIO Storage adapters, the feature allows addition of
storage capacity without guest downtime. With NPIV Storage adapters, it allows online addition
of storage capacity and online addition of redundant paths to an existing NPIV LUNs.
For more information about dynamic addition of IO devices to vPars and VM guests, see
Section 13.17 (page 247).
6.4.1.8 Notifying guest OS of changes in guest storage configuration
With HP-UX 11i v3, the AVIO storage vPar and VM guest driver can receive events asynchronously
from the VSP whenever the underlying storage, such as LUN or target changes state, for example,
when a new LUN or target is added or deleted or when the size of a LUN changes.
The asynchronous event generation occurs in addition to any notifications issued using the SCSI
programming model, such as CHECK CONDITION on a subsequent I/O. When the AVIO storage
driver on the vPar or VM guest detects the events, it takes appropriate actions, such as discovering
the new targets. For example, if new targets are added using the hpvmmodify -a command,
then the vPar or VM guest driver automatically detects the new device without the manual scan.
The vPar or VM guest automatically detects any modification to the underlying backing storage.
To avoid damage to vPar or VM guest, you must change the underlying backing storage on an
existing vPar or VM guest device when it is not running. If the change is to a running vPar or VM
guest, the administrator must ensure that the change will not adversely affect the health of the
running environment. Although, HP-UX vPars and Integrity VM does check to determine if the device
is in use, those checks are not reliable, because the vPar or VM guest might or might not be using
the device at the time it is checked.
Backing storage can be adversely affected if the actual storage or access path is modified directly
by an HP-UX server command, for example, by removing a file backing store or unmounting the
file system. If the devices being changed are a result of some SAN reconfiguration, you must run
the ioscan command on the VSP before attempting the change with the hpvmmodify command.
If the backing storage is changed by remapping a different wwid to an existing dsf using scsimgr
replace_wwid -D dsf, you must run the hpvmdevmgmt -I command. If the backing storage
is SAN presented as a different device and the change is done using io_redirect_dsf -d
old_dsf -n new_dsf, the vPar or VM guest must be modified using the hpvmmodify command
to reference the new disk in place of the old disk.
NOTE: When a SLVM LV is configured as a backing store for a vPar or VM guest, any changes
made to the LV size, will be automatically propagated to the vPar or VM guest only if it is running
on the cluster node configured as the server for the volume group to which the LV belongs.
Restarting the guest reflects the modified LV size.
6.4.1.9 Virtual storage setup time
Some virtual devices take longer to set up than others. Whole disks are very easy to set up because
they require nothing more than a character device file. This is usually created automatically when
the VSP system is booted.
Logical volume creation is simple. Logical volumes are used widely on HP-UX systems. The Veritas
Enterprise Administrator, the HP-UX Logical Volume Manager commands, or the SMH can be used
to create logical volumes. With experience, you can use logical volume commands quickly.
Creating files for virtual devices is not hard, but takes time. Files are usually placed on top of
logical volumes, so you might have to create a logical volume first.
74 Storage devices