POWER5 & APV: Perfect Together
With the arrival of the 20th anniversary of AIX taking place this year, it's indeed time for a celebration. Along with the evolution of AIX, with the introduction of IBM's POWER5 architecture, the combination of both has evolved so that the midrange platforms now feature mainframe-type capabilities (reliability, virtualization and performance). IBM has made substantial improvements throughout the years on their IBM proprietary RISC based hardware, where additional mainframe-type components are actually needed today to utilize the new architecture. Systems like the hardware management console (HMC) and the Hypervisor (software which runs on hardware machines and manages one or more operating systems) are important elements of the POWER architecture and are necessary to build systems and take advantage of systems such as Advanced Power Virtualization (APV) (Figure 1).
Though clearly the vast majority of businesses are still running AIX on the System p5 platform, perhaps the most impressive part of the POWER5 architecture is that it has been optimized to run Linux. Unlike other RISC-based hardware, IBM has fully implemented most of the functionality of the POWER5 architecture into its Linux support. This is largely due to the recent developments of the Linux 2.6 kernel, which has brought Linux into the Enterprise. IBM added its own code to SUSE and Red Hat variants to provide support to the POWER5 architecture. The POWER5 architecture clearly is one of the most powerful midrange systems available today. The flexibility that logical partitioning offers to the midrange environment is a "killer app" type attribute, which provides companies the capability to rapidly add and change environments. APV helps decrease TCO, while allowing the use of shared I/O resources where applicable. A major feature available to System p platform through POWER5 technology is micro-partitioning (a feature of APV), which enables the creation of multiple virtual partitions within a single processor, each one tailored to the resource requirements of a particular application based on business needs and priorities. Through micro-partitioning of its LPARs, it allows users to take advantage of unused clock cycles in an attempt to mirror the mainframe world by utilizing as much CPU capacity as possible. Without partitioning, processing resources are typically underutilized. SMP partitioning today traditionally requires allocation of one or more entire microprocessors to each partition supported. Micro-partitions can be tailored to the demands of individual applications, in increments of as little as 1/10th of a processor. The results: increased productive use of system resources, higher system productivity and lower TCO. We'll discuss this more in detail later in this article.
The flexibility that IBM provides which allows different types of operating systems to run on System p platforms (Linux, AIX, i5/OS partitions) provides companies with even more consolidation options. IBM has also published their roadmap of systems that clearly define the next steps through the next several years (Figure 2), which is extremely important to partners who need to know the direction that their technology company is taking. The POWER5 Architecture is clearly a winner.
An exploration of APV is necessary, but first I have to introduce the Hypervisor.
The technology behind the virtualization of IBM System p platforms is provided by a kind of firmware known as the POWER Hypervisor. The Hypervisor supports partitioning and dynamic resource movement across multiple operating systems, without which there would be no APV. It's the foundation of the IBM virtualization engine, which is implemented as part of the overall POWER5 architecture.
The POWER Hypervisor supports many advanced functions when compared to previous versions found in POWER4 processor-based systems. This includes sharing of processors, virtual I/O, and high-speed communications among partitions. The way it does this is by using a virtual LAN, while it enables multiple operating systems to run on the single system, including AIX, Linux, and i5/OS. With support for dynamic resource movement across multiple environments, clients can move processors, memory, and I/O between partitions on the system as workloads are moved between partitions.
The POWER5 processor supports special machine instructions, which are exclusively used by the Hypervisor. If an operating system instance in a partition requires access to hardware, it must first invoke the Hypervisor using Hypervisor calls. The POWER Hypervisor also allows privileged access to the operating system for dedicated hardware facilities and includes protection for those facilities in the processor and memory locations. Without the Hypervisor, there is no APV. Like any virtual engine, the hypervisor does come with an overhead, but not nearly as much as systems such as VMWARE. Its benefits clearly outweighs the overhead and allows one to take full advantage of the POWER5 architecture.
comments powered by