| There are three major classifications of the solutions in the VSE. These are: VSE Partitioning Solutions: These are the different types of partitioning available within the Virtual Server Environment. VSE Utility Pricing Solutions: These are the solutions that allow customers to acquire sufficient system resources for peak loads but pay for them when only they are needed. VSE Management and High Availability Tools: These include workload management, configuration, capacity planning, high availability, and disaster tolerance solutions.
This book covers each of these classes of solutions. Now let's describe what tools are included in each of these areas and provide some context for the remainder of the book. VSE Partitioning Solutions When you think of "virtualization," partitioning is probably what you are thinking about. Partitioning allows you to run multiple workloads on a system while isolating them from each other by virtualizing the resources so the workloads can't impact each other. Another big advantage of this is that it gives you much better control over how resources are allocated to the various workloads. Figure 1-3 shows the four partitioning solutions available with the Virtual Server Environment. These are: nPartitions (nPars): These are fully electrically isolated hardware-based partitions. Some of the compelling benefits here are the fact that hardware failures are isolated to the partition they occur in. You can also perform hardware maintenance activities in one partition while other partitions are running. Virtual Partitions (vPars): These allow users to run multiple HP-UX instances on a single system or nPar. This provides finer granularity than nPars and resources can be reallocated between partitions while they are running. The vPars product is available on both HP 9000 and Integrity platforms and is implemented by assigning hardware components to each partition. Integrity Virtual Machines: These are similar to vPars in that you can run multiple independent instances of the OS on a single system or nPar. Integrity Virtual Machines (VMs) provide fully virtualized hardware which makes it possible to have the partitions go to sub-CPU granularity, and they can also share physical I/O cards between separate OS instances. In addition, VMs can be moved from one system to another, system completely transparently to the applications running inside the VM. Secure Resource Partitions: These allow you to run multiple workloads in a single instance of HP-UX and isolate the workloads from each other. Isolation includes resources and communications. Each partition gets its own network interface, CPU, and memory resources even though they are running within a single copy of HP-UX.
Figure 1-3. The HP Partitioning Continuum for HP 9000 and Integrity Platforms 
As you can tell from Figure 1-3, you can also stack these solutions on top of one another. You can run vPars, Integrity VM, or Secure Resource Partitions inside of nPartitions. You can also run Secure Resource Partitions inside vPars or Integrity VM. Many customers have found that deploying multiple types of partitions on a single server provides a good combination of isolation and flexibility. Part 1 of this book covers all of these partitioning solutions in full detail. VSE Utility Pricing Solutions HP's Utility Pricing Solutions for the HP 9000 and Integrity family of servers provide the ability to acquire hardware that is preconfigured for future requirements, but they give you flexibility about how and when you pay for those resources. note These Utility Pricing Solutions provide a very compelling opportunity to save money. So much so that we make the statement that you should not be purchasing or leasing servers that don't have one or more of these solutions. We describe why in Part 2 of this book.
There are Utility Pricing solutions for both capital purchase and leased systems. The solutions for capital purchase are: Instant Capacity: This provides the ability to acquire a system with inactive capacity. Most of the cost of these inactive components is deferred until they are needed. This program supports both the activation of CPUs and/or whole cells with the memory attached. The instant capacity program was designed to simplify system upgrades by allowing a customer to acquire systems with ample spare capacity to support future growth without paying for all of that capacity up front. Temporary Instant Capacity: This provides the ability to activate instant capacity CPUs for short periods of time. This gives customers the ability to respond to short-term peaks in demand by activating additional capacity.
Customers who choose leased servers can take advantage of a Pay-per-use lease. This is what you might think of as a true utility model, where the lease payment varies each month based on the actual utilization of the server in the previous month. These programs are described in Part 2 of this book. VSE Management and High Availability Tools The unprecedented level of adaptability provided by the Virtual Server Environment makes it possible to set up servers with multiple workloads that are isolated from each other while sharing resources. In addition, resources can be reallocated among workloads as required in real time and workloads can be migrated between servers. This makes it much easier to increase the utilization of servers. What's more, all of this can be done transparently to users. However, this level of virtualization and flexibility introduces new challenges in the area of management. To address this management challenge, HP has introduced the VSE management suite of tools. This is an integrated suite of management tools that will cover the primary management functions in a datacenter, including: Workload provisioning: Tools will be provided that will simplify the creation of new partitions and virtual machines. Visualization and configuration: These are the traditional configuration and monitoring capabilities that are provided with each partitioning product. However, there will be one significant differencestarting with the first release of the HP Integrity Essentials Virtualization Manager, they are all integrated into a single common user interface so that it will be easier to identify what workloads are running in each partition on each server in the datacenter. Workload management: HP has been the leader in workload management since its introduction of the HP-UX Workload Manager product in April 2000. The VSE management suite will have links into this product but will have even tighter integration with the new Global Workload Manager product. Capacity planning: Because of the unique level of flexibility available in the VSE, it is necessary that capacity planning tools be knowledgeable about the types of flexing technologies available on the servers that are being monitored. To address this, HP has introduced the HP Integrity Essentials Capacity Advisor product as part of the VSE management suite. High availability and disaster tolerance: Another unique capability in the VSE is the ability to move workloads from one system to another. This can be done with Integrity VM where the "machine" itself can be rehosted, including all its resources and addresses. In addition, HP has the industry-leading Serviceguard suite of solutions for high availability and disaster tolerance. Having these products integrated into the VSE management suite makes it much easier to see what workloads are running where and to move workloads for convenience or in response to a failure.
All of these tools are being tightly integrated with each other and with the HP Systems Insight Manager so that the user experience will be one where you don't need to know what tool to use to accomplish a task. Not all of these features will be fully supported in the first release, but these capabilities will be filled out over time. We describe the VSE Management tools in detail in Part 3 of this book. Integration of VSE Solutions So far we have described a number of very useful tools for creating and managing an adaptive enterprise. What really sets these solutions apart is that they are integrated with each other. This makes it possible to stack various types of partitions on a system to run multiple workloads, provision the system with idle capacity, and automate the activation and reallocation of resources among all these things when workloads need them. This is the realization of the promise of an adaptive enterprise. | 