HP Mainframe Connectivity Design Guide

Multiple allegiance
MA enables multiple LPARs to issue multiple I/O requests in parallel to a disk volume. MA is a
standard part of the PAV feature in the P9500 storage system.
Hyper PAV
Hyper PAVs build on the PAV and MA functions. The difference is that instead of fixed alias
addresses configured for each base address, Hyper PAV allows a pool of alias addresses to be
used and reused to access base addresses. This means that fewer alias addresses need to be
configured because each alias address can be assigned dynamically to a base address as needed.
The IOS in the mainframe manages the Hyper PAV pool, assigning (or binding) an alias address
to a base address for the duration of a single I/O command. After the I/O completes, the alias
address is released, and it can then be bound to another base address when needed.
Table 66 (page 107) lists the P9500 CUs that support Hyper PAV.
Table 66 P9500 PAV support
Number of PAVs per base addressCU typeStorage system
255
1
2107P9500
1
The number required for your workload is typically much less.
P9500 configuration rules
Table 67 (page 107) lists the P9500 storage system/mainframe configuration rules.
Table 67 P9500 configuration rules
Description
Rule
nl
number
The P9500 storage system must be configured to emulate an IBM CU for mainframe implementations.
For a list of IBM CUs, see Table 61 (page 105).
1
The P9500 storage system must be configured to emulate IBM disk volumes. For a list of IBM volume
types, see Table 62 (page 105).
2
A maximum of 256 disk volumes per CU/LCU is supported.3
The storage system is logically partitioned into uniquely addressable LCUs. Each image appears as an
independent LCU; however, all LCUs share a common set of hardware facilities. Each LCU is configured
with a maximum of 256 volumes.
4
Each volume in the LCU is assigned a unique two-digit hexadecimal unit address in the range x'00' to
x'FF'.
5
A combination of the LCU number and the unit/device address uniquely identifies a particular volume
in the storage system.
6
The LCU is called the CUADD in the mainframe configuration file. The CUADD is a two-digit hexadecimal
value in the range x'00' though x'FF'. For example, LCU x'01' is the same as CUADD x'01'.
7
A unique 4-digit hexadecimal device address in the range x'0000' to x'FFFF' is assigned to each volume
in the mainframe configuration. This device address is also referred to as a UCB.
8
A unique 4-digit hexadecimal CUNUMBER in the range x'0000' to x'FFFF' is assigned to each LCU in
the mainframe configuration. Typically, the starting device address or UCB is assigned as the CU number.
9
A UNITADD is the starting address of the range of devices configured in the LCU. For example, if the
LCU has 256 devices configured with the device addresses x'00' through x'FF', then x'00' is the
UNITADD.
10
The UNITADD parameter contains two fields: the first field is UNITADD, and the second field is a decimal
value describing the range of addresses in the LCU. For example, an LCU with 256 devices and
configured with the device addresses x'00' through x'FF' has a UNITADD parameter of 00,256.
11
P9500 configuration rules 107