Extracting Resource Allocation Data for Chargeback in a HP Virtual Server Environment for HP Integrity Servers

cannot effectively use the remainder of the time for any other purpose. The user might prefer to be

billed for utilized court time in increments of minutes, but the club allocates time in hourly increments.

Monthly member charge based

on;

• Minutes on a court (utilized) ?

• Minutes allocated (courts

blocked in hour increments)?

• Minimum hours of access

guaranteed in membership?

• Combination?

Figure 5 Tennis club analogy

The next example is specific to servers managed with VSE technology and is depicted in Figure 6.

Using the Virtual Partition (vPar) technology, the CPU cores of a server can be allocated dynamically

to different vPars running different operating systems (OSs) and associated applications. The cores

are allocated in whole-core increments. The OS and applications in a vPar utilize a percentage of the

cores of the vPar. This utilization value is typically less than 100% of the core allocated to the vPar.

Utilization is the amount actually used by the user, but the core allocation is the resource provided by

IT. As with unused minutes of reserved time on a tennis court not being usable by any other members,

the unused processing of vPar cores is not usable by any other vPar. In addition, vPars can be

combined with VSE Global Workload Manager (gWLM) technology. When used with gWLM, the

vPar can have a guaranteed number of cores (referred to as “owned”) that exceeds both utilization

and allocation. The number of guaranteed cores requires additional capacity sizing by IT

independent of usage of the vPar. This sizing relates directly to IT’s cost of providing the vPar to the

user.