Welcome to the IBM Power virtualization management e-learning module. Power virtualization management, or PowerVM(R), is a combination of hardware, firmware, and software that provides efficient CPU, network, and disk virtualization for IBM POWER5(TM), POWER6(R), and POWER7(R) servers.
In this e-Learning module, we will review the overall benefits provided by PowerVM server virtualization. You will be introduced to PowerVM architecture, PowerVM management options, get an overview of PowerVM virtualization technologies, and be introduced to the three different PowerVM editions. Users should have a basic understanding of virtualization before beginning. For a basic understanding of virtualization, refer to the virtualization basics topic in the IBM information center.
IBM Power Systems servers combined with PowerVM technology are designed to help you consolidate multiple workloads onto fewer systems, which increases server utilization, provides flexibility, and reduces cost. IBM PowerVM technology is one of the industry leading virtualization solutions for AIX(R), IBM i, and Linux(R) operating system platforms, providing a secure and scalable virtualization environment.
Business responsiveness and operational speed is increased by dynamically re-allocating resources to applications as needed to better match changing business needs and handle unexpected changes in demand. IT infrastructure workloads become independent of hardware resources, thereby enabling you to make business driven policies to deliver resources based on time, cost and service level requirements.
Let us look at the major architectural components of PowerVM.
PowerVM is built on the advanced reliability, availability, serviceability features, as well as the leadership performance of POWER system platforms and virtualization technologies enabled by the POWER Hypervisor(TM) and the Virtual I/O Server.
The POWER Hypervisor (show/highlight hypervisor layer in picture) is a basic component of system firmware that is considered the foundation of IBM PowerVM. Combined with features designed into the IBM POWER processors, the POWER Hypervisor enables logical partitioning and other virtualization technologies to support multiple environments running on a single system. The POWER Hypervisor provides an abstraction between the physical hardware resources and the logical partitions that use them. The POWER Hypervisor is always installed and activated, regardless of system configuration.
The Virtual I/O Server, (show/highlight VIOS representation in picture) known as VIOS, is a software appliance that enables sharing of physical I/O resources among supported AIX, Linux and IBM i logical partitions (show AIX, Linux, and IBM i LPARs in picture). The VIOS technology facilitates consolidation and more efficient utilization of physical storage and network devices. Dual VIOS configurations can be deployed to provide maximum availability for client partitions.
You will notice in discussing PowerVM technology that there are some terminology differences from IBM x86 servers. For example, on x86 systems, an isolated partition running a workload is called a virtual machine. On Power systems, that is referred to as a logical partition, or LPAR. For information on these differences, refer to the PowerVM Getting Started Guide, chapter 1, section 1.
The three main POWER management consoles that are available are the Hardware Management Console, or HMC, the Integrated Virtualization Manager or IVM, and the IBM Flex System Manager(TM), known as FSM.
The HMC is a hardware appliance that provides a centralized point of hardware management of one or more Power Systems from entry level servers to high-end systems. The HMC interface helps you to create, manage, and modify logical partitions, configure hardware, and manage service calls.
Management of entry level systems is also possible with IVM which is a function of the VIOS. IVM is a simplified virtualization management solution that is integrated within the VIOS andmanages a single stand-alone server. IVM simplifies LPAR creation and management for entry Power Systems(TM) servers and blades.
The IVM is confined to the system the VIOS is installed on, while the HMC can manage multiple systems.
The IBM FSM is software for IBM PureFlex(TM) System solutions for managing compute nodes, storage, networking, and virtualization across multiple chassis, all within a single interface.
PowerVM features virtualization technologies for logical partitioning, shared input/output, and memory sharing. Additional optional virtualization features are also available such as live partition mobility, and partition Suspend/Resume functionality.
Logical partitioning is a virtualization technology enabled by the POWER Hypervisor which enables you to run separate workloads in separate partitions on the same physical server. Each partition has its own operating environment such as AIX, Linux, or IBM i and server resources.
The number of LPARs that can be created depends on the system. Typically, partitions are used for different purposes, such as database operations, client/server operations, web server operations, test environments, and production environments. Each partition can communicate with the other partitions as if each partition were a separate machine.
An LPAR interprets its virtual processor as an existing physical server. There can be more virtual processors assigned than actual cores available, enabling LPARs to share the same core. From an operating system perspective, a virtual processor cannot be distinguished from a physical processor.
The configuration details of an LPAR, including the operating system type, processor, memory and I/O resources, is called the partition profile. This profile becomes a record on the management console that specifies the possible configuration for the partition. When you activate the LPAR, the managed system attempts to start the logical partition using the configuration information in the partition profile. The IVM console does not support partition profiles.
Dynamic logical partitioning, known as DLPAR, is the capability of an existing LPAR to be reconfigured on the fly without having to shut down the operating system and interrupt the current operation. Processors, memory, I/O components, and other LPAR resources, can be modified while the LPAR is still running applications. DLPAR might be used for removal or deallocation operations such as removing memory, virtual adapters, or to allow a processor to be taken offline so it can be used for new workloads on the machine.
LPARs might be assigned entire processors for dedicated use, or partial processor units from a shared processor pool. In dedicated mode, physical processors are assigned completely to one LPAR. The dedicated partition maintains absolute priority for dedicated CPU cycles.
Micro-Partitioning technology coupled with the Power Hypervisor facilitates the sharing of processors between logical partitions. In shared mode, logical partitions use virtual processors to access fractions of physical processors or processing units. The number of processing units defined for a shared processor partition is known as entitled capacity.
With POWER technology you can configure dedicated partitions to become processor donors for idle processors they own, allowing for the donation of spare CPU cycles from dedicated processor partitions to a shared processor pool.
Shared processor partitions can be either capped or uncapped. The mode determines the maximum processing capacity that might be allocated from the shared processor pool. In capped mode the processing capacity can never exceed the entitled capacity of the shared processor partition. Uncapped mode enables processing capacity to exceed the entitled capacity when resources are available.
In busy shared processor pools where the shared processor partitions have used all of their entitled capacity, the POWER Hypervisor will allocate resources to LPARs based on their assigned weight value. The higher the number, the better the chances the LPAR will get spare cycles.
The VIOS allows virtualization of network and physical storage resources to enable sharing of these resources among multiple client logical partitions. Power VM technology uses a Shared Ethernet Adapter, or SEA, to connect a physical Ethernet network to a virtual Ethernet network. This enables several client partitions to share one physical adapter.
Virtual local area networks, or VLANs, are used to divide networks into smaller, more manageable segments. When an LPAR needs to participate in a VLAN that is external to the hypervisor, access to the external network is provided by the SEA. You can add, remove, or modify the existing set of VLANs for a virtual Ethernet adapter that is assigned to an active partition by using the PowerVM HMC, or one of the other available PowerVM management consoles, either IVM or FSM.
There are two different methods by which virtualized storage devices can be accessed by the client partitions - either by Virtual SCSI, or by Virtual Fibre Channel devices. Virtual SCSI provides standard SCSI compliant access to disk devices, optical devices and tape devices. Virtual Fibre Channel devices use NPIV to provide access to Fibre Channel attached disk and tape libraries.
Virtual I/O servers also provide access to Shared storage pools. Shared storage pools simplify the aggregation of large numbers of disks across multiple Virtual I/O Servers, improve the utilization of the available storage, and simplify administration tasks.
When using shared storage pools, the Virtual I/O Server provides storage through logical unitsthat are assigned to client partitions. A logical unit is a file backed storage device that resides in the cluster filesystem in the shared storage pool. It appears as a virtual SCSI disk in the client partition.
Dual Virtual I/O Servers can be configured to provide multi-path access to client virtual storage. This allows VIOS maintenance to be performed without affecting the virtual I/O clients.
Logical Partitions can be configured to use dedicate memory or use a flexible and dynamic shared memory model enabled by IBM Active Memory(TM) Sharing technology, or AMS. With a pure dedicated memory model, a fixed amount of physical memory is allocated to a logical partition. It is the task of the system administrator to optimize available memory distribution among partitions by issuing a dynamic memory reconfiguration as needed.
With a shared memory model, a dedicated amount of physical memory is not assigned to each partition. Instead, the POWER Hypervisor automatically determines the optimal distribution of the physical memory in the shared memory pool to partitions, and adjusts the memory assignment based on partition load.
When a shared memory partition needs more memory than the current amount of unused memory in the shared memory pool, the hypervisor stores a portion of the memory in auxiliary storage. Access to the auxiliary storage is provided by a Virtual I/O Server.
AMS technology is available with the IBM PowerVM Enterprise Edition hardware feature, which also includes the license for the VIOS software.
There are several additional, optional PowerVM features available depending upon which PowerVM version is being used.
PowerVM Live Partition Mobility, or LPM, allows you to move an active, running LPAR from one system to another without shutting down or disrupting operations. This includes the partition operating system and any running applications.
PowerVM technology also includes support for suspend and resume operations which enables the state of a running LPAR to be stopped and saved, and at a later time restarted from exactly the same point of execution. This is useful for performance and energy management. When a suspended LPAR is in a standby, hibernated state, all of its resources can be used by other partitions.
Active Memory Expansion, or AME, is technology that allows the effective maximum memory capacity to be much larger than the true physical memory maximum. AME uses the CPU resources of a partition to compress or decompress the memory contents of the partition, enabling memory expansion up to 100%. A partition is therefore able to do significantly more work or support more users with the same physical amount of memory. Similarly, it can enable a server to run more partitions and do more work for the same physical amount of memory.
Virtualization technology is offered in three editions on Power Systems, with each suited for various purposes. These are the PowerVM Express Edition, PowerVM Standard Edition, and PowerVM Enterprise Edition. This set of comprehensive systems technologies and services are designed to enable you to combine and manage resources using a consolidated, logical view.
In this e-Learning module, you have learned about the benefits of PowerVM server virtualization, Power VM architecture, PowerVM Management Options, gotten and overview of PowerVM Virtualization technologies, and review the 3 different PowerVM editions.
For more information about PowerVM technology, see the IBM Redbooks(R) and information center topics.