HP-UX 11i v3 Native Multi-Pathing for Mass Storage (August 2012)

Note

For information on determining or configuring the optimized/un-

optimized controller or target port, see the disk device’s user

documentation. Based on this information you can set the preferred_path

and the pref_tport attributes appropriately on the host system.

• Weighted round-robin (weighted_rr) – This policy distributes the I/O load across all active

lunpaths in a round-robin manner and according to the weight assigned to each lunpath. A number

of I/O operations corresponding to the weight of the lunpath is transferred on the same lunpath

before another lunpath is selected. Lunpaths with a weight of 0 are excluded from I/O transfer.

The default disk device load balancing policy is round-robin, even on servers supporting hard

partitioning. However, the administrator can change it to any other policy.

Note

For devices implementing the T10 ALUA (Asymmetric Logical Unit Access)

standard (see the SPC-3 section 5.8), the SCSI stack on HP-UX 11i v3

automatically detects the set of lunpaths to the optimized target port groups

and uses only these lunpaths for I/O transfer according to the I/O load

balancing policy set for the disk device. Lunpaths to un-optimized target

port groups are put in standby state and are not used for I/O transfer. You

can disable this behavior by setting the alua_enabled attribute to false

for the disk device. In this case, the support of the ALUA standard is

ignored by the SCSI stack and all lunpaths are used for I/O transfer.

The administrator has flexibility in setting the scope of load balancing policies for disks. You can set a

load balancing policy for all disk devices, or for a set of disk devices meeting some specific criteria,

such as vendor id, product id, and firmware revision, or for a specific disk device.

Policies supported for tapes, changer, and pass-through devices

Because of the serial nature of the I/O flow to tape and changer devices, only one path selection

mechanism is supported by

estape and eschgr drivers: path lock down (path_lockdown). When

opening a tape or changer LUN DSF, the lunpath onto which all the I/O operations are sent can be

selected by the mass storage subsystem or set by the administrator. By default, the mass storage

subsystem internally selects the optimal lunpath. The administrator can manually set the lunpath with

the lpt_to_lockdown attribute. The lpt_to_lockdown attribute must only be set when the device

is in the UNOPEN state.

Devices claimed by the pass-through driver (

esctl) also use path lockdown as the I/O load balance

policy.

High availability of SCSI devices

To provide applications with continuous access to the LUNs, the mass storage subsystem has built-in

LUN access high availability.

By default, without requiring special upper layers, the mass storage subsystem increases the

availability of LUN devices to applications by doing lunpath failovers, detecting and removing failing

lunpaths from the application I/O traffic, by monitoring offline lunpaths, and automatically making

lunpaths that come online available to application I/O traffic.

The SCSI subsystem monitors SCSI components, such as SCSI controllers, target ports, lunpaths and

LUNs. As failing components are identified, the mass storage subsystem reports errors by logging