Enterprise-class Server Research Articles

Enabling Genomics Pipelines in Commodity Personal Computers With Flash Storage.

Analysis of a patient's genomics data is the first step toward precision medicine. Such analyses are performed on expensive enterprise-class server machines because input data sets are large, and the intermediate data structures are even larger (TB-size) and require random accesses. We present a general method to perform a specific genomics problem, mutation detection, on a cheap commodity personal computer (PC) with a small amount of DRAM. We construct and access large histograms of k-mers efficiently on external storage (SSDs) and apply our technique to a state-of-the-art reference-free genomics algorithm, SMUFIN, to create SMUFIN-F. We show that on two PCs, SMUFIN-F can achieve the same throughput at only one third (36%) the hardware cost and half (45%) the energy compared to SMUFIN on an enterprise-class server. To the best of our knowledge, SMUFIN-F is the first reference-free system that can detect somatic mutations on commodity PCs for whole human genomes. We believe our technique should apply to other k-mer or n-gram-based algorithms.

Shuttle: Facilitating Inter-Application Interactions for OS-Level Virtualization

OS-level virtualization generates a minimal start-up and run-time overhead on the host OS and thus suits applications that require both good isolation and high efficiency. However, multiple-member applications required for forming a system may need to occasionally communicate across this isolation barrier to cooperate with each other while they are separated in different VMs to isolate intrusion or fault. Such application scenarios are often critical to enterprise-class servers, HPC clusters and intrusion/fault-tolerant systems, etc. We make the first effort to support the inter-application interactions in an OS-level virtualization system without causing a significant compromise on VM isolation. We identify all interactive operations that impact inter-application interactions, including inter-process communications, application invocations, resource name transfers, and application dependencies. We propose Shuttle, a novel approach for facilitating inter-application interactions within and across OS-level virtual machines. Our results demonstrate that Shuttle can correctly address all necessary inter-application interactions while providing good isolation capability for all sample applications on different versions of Windows OS.

Resilient cloud computing

IBM SmartCloud® Enterprise+ (SCE+) is IBM's premier cloud computing offering for enterprise customers. The SCE+ cloud is designed to be a resilient, highly available system with no single point of failure. SCE+ uses an integrated set of enterprise-class servers, network elements, and storage components that have internal redundancy, RAS (reliability, availability, and serviceability) features, and high mean time between failures. SCE+ is also a managed cloud. The SCE+ management system utilizes IBM service-management and platform-management tools. To help ensure that the customer's virtual servers are running and avoid system failures, we employ several techniques. For example, virtual servers are automatically restarted upon crash or upon hosting physical-server failures. In addition, network elements are interconnected to allow alternate network traversal paths, and data are mirrored to offset storage failure, while each of these server, storage, and network elements also have redundant internal configurations. In addition to providing guest-level availability, important customer workloads, such as the enterprise resource planning applications and databases, require highly available clusters. More generally, we describe approaches used to achieve resiliency in SCE+.

Facilitating inter-application interactions for OS-level virtualization

OS-level virtualization generates a minimal start-up and run-time overhead on the host OS and thus suits applications that require both good isolation and high efficiency. However, multiple-member applications required for forming a system may need to occasionally communicate across this isolation barrier to cooperate with each other while they are separated in different VMs to isolate intrusion or fault. Such application scenarios are often critical to enterprise-class servers, HPC clusters and intrusion/fault-tolerant systems, etc. We make the first effort to support the inter-application interactions in an OS-level virtualization system without causing a significant compromise on VM isolation. We identify all interactive operations that impact inter-application interactions, including inter-process communications, application invocations, resource name transfers and application dependencies. We propose Shuttle, a novel approach for facilitating inter-application interactions within and across OS-level virtual machines. Our results demonstrate that Shuttle can correctly address all necessary inter-application interactions while providing good isolation capability to all sample applications on different versions of Windows OS.

Electronic packaging of the IBM System z196 enterprise-class server processor cage

In this paper, we describe the first- and second-level system packaging structure of the IBM zEnterprise® 196 (z196) enterprise-class server. The design point required a more than 50% overall increase in system performance (in millions of instructions per second) in comparison to its predecessor. This resulted in a new system design that includes, among other things, increased input/output bandwidth, more processors with higher frequencies, and increased current demand of more than 2,000 A for the six processor chips and two cache chips per multichip module. To achieve these targets, we implemented several new packaging technologies. The z196 enterprise-class server uses a new differential memory interface between the processor chips and custom-designed server memory modules. The electrical power delivery system design follows a substantially new approach using Vicor Factor Power® blocks, which results in higher packaging integration density and minimized package electrical losses. The power noise decoupling strategy was changed because of the availability of deep-trench technology on the new processor chip generation.

Performance analysis of idle programs

This paper presents an approach for performance analysis of modern enterprise-class server applications. In our experience, performance bottlenecks in these applications differ qualitatively from bottlenecks in smaller, stand-alone systems. Small applications and benchmarks often suffer from CPU-intensive hot spots. In contrast, enterprise-class multi-tier applications often suffer from problems that manifest not as hot spots, but as idle time , indicating a lack of forward motion. Many factors can contribute to undesirable idle time, including locking problems, excessive system-level activities like garbage collection, various resource constraints, and problems driving load. We present the design and methodology for WAIT, a tool to diagnosis the root cause of idle time in server applications. Given lightweight samples of Java activity on a single tier, the tool can often pinpoint the primary bottleneck on a multi-tier system. The methodology centers on an informative abstraction of the states of idleness observed in a running program. This abstraction allows the tool to distinguish, for example, between hold-ups on a database machine, insufficient load, lock contention in application code, and a conventional bottleneck due to a hot method. To compute the abstraction, we present a simple expert system based on an extensible set of declarative rules. WAIT can be deployed on the fly, without modifying or even restarting the application. Many groups in IBM have applied the tool to diagnosis performance problems in commercial systems, and we present a number of examples as case studies.

Packaging design challenges of the IBM System z10 Enterprise Class server

This paper describes the system packaging and technologies of the IBM System z10™ high-end Enterprise Class server. This machine exceeds the multiprocessor performance of the previous system by 50%. A new generation of the IBM Elastic Interface was developed in order to maintain the increased interconnect signal speed of up to 2.93 Gb/s. Power control and power delivery to the multicore processors were a special challenge for the server packaging because of the high currents and the high number of voltage domains.

Redefining Broadcast Transmission Systems: A Software-Based Approach

With the continued advances of information technology into content management and production workflows, broadcast facilities are becoming largely information technology (IT)-centric operations, leveraging proven enterprise-class server and network technologies. A major exception to this trend, however, is the broadcast transmission process (master control and playout), which remains heavily dependent on specialized broadcast hardware. In this process, transmission of a moderately complex television channel can involve a dozen or more specialized hardware components linked by an automation system. In many cases, the costs associated with this traditional approach to broadcast transmission can be prohibitive for internet protocol television (IPTV) and other new operators intending to set up and manage their own playout facilities. Furthermore, to compete head-on with mainstream broadcast channels, new service providers need to match the quality standard and aesthetics that an increasingly discerning audience has come to expect. The ability to move the transmission process to an IT platform opens the door to an entirely new business model for broadcasters. With a fully-featured broadcast transmission chain provided by a software solution operating on industry-standard IT servers rather than specialized broadcast hardware, it is now possible to build and operate a broadcast transmission center as an extension of an IT-centric enterprise.

High-speed interconnect and packaging design of the IBM System z9 processor cage

This paper describes the system packaging and technologies of the IBM System z9TM enterprise-class server. The central electronic complex of the system consists of four nodes, each housing a multichip module (MCM) with 16 chips consuming up to 1,200 W. The z9TM server doubles the multiprocessor performance of the System z990 by increasing the central processing unit (CPU) configuration and using an internally developed elastic interface to increase interconnect speed on all high-speed buses. In contrast to all previous zSeries® designs, which were running at half of the processor speed, the packaging interconnects on the multichip module run at the same speed as the processor (1.72 GHz). High frequencies and massively parallel connectivity lead to a raw packaging bandwidth of up to 1,764 GB/s between processors and cache within a single frame for a fully configured four-node z9 system.

Parallel optical interconnects for enterprise class server clusters: needs and technology solutions

We examine the status of enterprise-class server clusters and the communication issues that need to be addressed in future systems. With increasing system performance, new approaches beyond traditional copper-only communication solutions have to be examined. Parallel optics is an attractive solution to overcome copper's shortcomings, but traditional approaches to parallel optics have had their own limitations. We describe a new approach to parallel optics - dense parallel optics - and its relevance to enterprise servers. After discussing system communication needs, we examine dense parallel optics from both the passive cabling and active component views. We explain how dense parallel optics offers a unique way to address the performance, cost, reliability, and scalability of server systems. We also discuss how current approaches to dense parallel optics afford system opportunities beyond simple data transport.