Parallel Sysplex Research Articles

System Recovery Boost on IBM z15

System Recovery Boost on the IBM z15 server expedites planned operating system shutdown, either planned or unplanned operating system initial program load (IPL), middleware and workload restart and recovery, and the client workload execution that follows, to accelerate service restoration around downtime. It does this by providing limited-duration “boost periods” that deliver significant usable additional processor capacity and parallelism. On subcapacity machine models, it provides a boost in processor speed by running the general-purpose processors at full-capacity speed, for the boosting LPARs only, and only during the boost periods. It makes all available processing capacity defined to the boosting images available to process any kind of work, “blurring” general-purpose processor and specialty processor capacity together during the boost period. System Recovery Boost also expedites and parallelizes processor reconfiguration actions that may be part of the client's overall restart and recovery process, as orchestrated by Geographically Dispersed Parallel Sysplex (GDPS) automation. Optionally, System Recovery Boost provides the ability to add additional processor capacity from the client's unused “dark cores” via activation of a new type of temporary capacity record. All of this can be accomplished without increasing the client's IBM software billing costs or the processor consumption associated with the client's workload during these boost periods.

IBM z14 Parallel Sysplex and coupling technology

IBM z14 Parallel Sysplex and coupling technology

The IBM z Systems Coupling Facility exploitation of storage-class memory

This paper describes the new support in the IBM zEnterprise® EC12 and BC12 mainframes, as well as the IBM z13™, which allows Coupling Facility partitions to use storage-class memory. This use of storage-class memory improves the scalability and availability of the implementation of the IBM WebSphere® MQ shared message queues in a Parallel Sysplex® configuration. By providing a new level in the storage hierarchy in the Coupling Facility to which MQ messages may be offloaded and from which they may be later retrieved, storage-class memory augments capacity and scale within the MQ shared message queue environment. This in turn provides availability benefits for the Parallel Sysplex by allowing shared queue message content to remain available and grow as needed to accommodate interruptions in service on the part of one or more of the MQ queue managers.

A contrast between mainframe Parallel Sysplex availability and selected distributed computing availability solutions

Critical applications worldwide are deployed on mainframes, and a chief reason for this is the mainframe infrastructure for high availability. In this paper, we describe the history and motivation behind GDPS®, the IBM Geographically Dispersed Parallel Sysplex, and we contrast GDPS with availability solutions that have been deployed on distributed computing platforms at Yahoo!, Amazon, and Google.

IBM Parallel Sysplex design for the IBM z196 system

The coupling adapter hub of an IBM System z® server is a key component for the IBM System z Parallel Sysplex®. The hub is built to exchange messages between systems in a highly efficient manner. This paper describes the latest generation of high-fanout and low-latency coupling adapter cards, the associated firmware, and a new protocol. As in the z10® system, there is a long-range and a short-distance card. The coupling adapter for zEnterprise® 196 (z196) is based on the z10 infrastructure (InfiniBand® link layer), with the internal transport engine for message handling completely redesigned to support the new protocol and improve connectivity, latency, and throughput. In addition to enabling the new adapter's functionality, the Parallel Sysplex support firmware has several significant enhancements in a number of functional areas. Connectivity and utilization are improved through the ability to define more channels and more concurrent connections (message buffer sets) for each channel. Through a combination of hardware and firmware protocols, response time for messages at a short distance is significantly improved. Finally, new methods are presented that support efficient presilicon and postsilicon functional and performance verification.

IBM System z10 I/O subsystem

The performance, reliability, and functionality of a large server are greatly influenced by the design characteristics of its I/O subsystem. The critical components of the IBM System z10™ I/O subsystem have, therefore, been significantly improved in terms of performance, capability, and cost. The first-order network has been redesigned from the long-evolved enhanced self-timed interface (eSTI) links to utilize InfiniBand™ links. A redesign of the host logic of I/O chips and the fiberoptic interfaces within the links made it possible to introduce InfiniBand-based IBM Parallel Sysplex® links. A broad range of legacy I/O channels have been carried forward to connect through InfiniBand, and a foundation has been laid for new channel types of improved functionality and performance. The first such hardware channel to be introduced is the next generation of Ethernet-virtualization data routers. A new and methodical recovery structure has been designed to ensure consistent, extensive support of reliability, availability, and serviceability. A building-block-oriented design process has been developed to enable the innovations that made these advances possible. Finally, a new performance verification methodology has been introduced to ensure that the system and subsystem designs are balanced to make effective use of the increased capacity.

IBM Parallel Sysplex clustering: Technology options for continuous availability

IBM Parallel Sysplex® is a clustering technology that was designed to address specific client business objectives for IBM mainframe System z® servers. The nondisruptive addition of scalable processing capacity and improved application availability with respect to unplanned and planned outages were two key design objectives. This paper focuses on the evolving technology options that support the business objective of continuous availability. Key technology options are discussed relative to their importance to achieving continuous availability. Best practices for effectively using these technology options to improve availability, based on extensive client experiences, are recommended.

Coupling I/O channels for the IBM eServer z900: Reengineering required

The IBM eServer z900 introduces new Parallel Sysplex® coupling channels that satisfy evolving requirements in a way that minimizes product and development costs. Their design also provides backward compatibility with earlier S/390® models, spans all three coupling channel design points, and anticipates future end-of-life technology issues. The original intersystem channel (ISC) design was improved, and new features added, but the core chips were retained. This paper describes the efforts that led to the improved design.

Dense Wavelength Division Multiplexing in Parallel Sysplex and Storage Area Networks

Get PDF Email Share Share with Facebook Tweet This Post on reddit Share with LinkedIn Add to CiteULike Add to Mendeley Add to BibSonomy Get Citation Copy Citation Text Casimer DeCusatis, "Dense Wavelength Division Multiplexing in Parallel Sysplex and Storage Area Networks," Optics & Photonics News 11(12), 18-19 (2000) Export Citation BibTex Endnote (RIS) HTML Plain Text Citation alert Save article

The Integrated Cluster Bus for the IBM S/390 Parallel Sysplex

IBM has developed a new S/390® Parallel Sysplex® coupling interface for the G5 server called the Integrated Cluster Bus (ICB). This interface improves the coupling efficiency by greatly reducing message-passing latency. Using the transport layer of the S/390 self-timed interface (STI) introduced in the G3 server, ICB adds channel function to the hub chip to allow a more direct interconnection between S/390 servers. This new channel has the same function as the present intersystem channel (ISC), but because it is integrated into the hub chip and therefore requires no additional components, its reliability is much better than that of the ISC. Since the ISC transmits data at a peak rate of 106 MB/s over distances exceeding ten kilometers and the ICB transmits data at a peak rate of 333 MB/s at distances of ten meters, the ISC is still required for the more geographically dispersed Parallel Sysplexes, whereas the ICB is well suited to the machine room, where multiple servers ca n be interconnected by ten-meter cables. This paper describes the design approach for the ICB. It describes the fundamental message-passing requirements of the Parallel Sysplex and how they are implemented in very complex yet compact hardware in the server's hub chip.

Integrated Cluster Bus performance for the IBM S/390 Parallel Sysplex

As the speed of the S/390® processors has increased at a rapid rate over the years, it has become extremely challenging to be able to connect them in a Parallel Sysplex® without compromising system performance. The cost of synchronous accesses to the coupling facility will build up progressively unless coupling link technology keeps pace with the development of the processors. Beginning with the G5 processors, IBM has introduced the ICB coupling link technology to enhance the performance of closely integrated processor clusters. With a high data transmission rate complemented by reduced hardware and microcode path lengths, the ICB provides excellent coupling efficiency. This paper emphasizes the need for the ICB and discusses the superior performance delivered by this latest coupling link technology.

Fiber optic interconnects for the IBM S/390 Parallel Enterprise Server G5

Fiber optic interconnections (“interconnects”) have become an increasingly important part of System/390® over the years. With the introduction of the Generation 5 Parallel Enterprise Server in May 1998, a number of important new fiber optic features and enhancements to existing features were made available for applications in input/output device attachment, networking, and parallel sysplex coupling. In this paper, we describe the main applications of fiber optic data links on System/390, with emphasis on the new features announced with the G5. In particular, we describe the fiber connection (FICON™) links, the Gigabit Ethernet interface for the open systems adapter (OSA Express), the Fiber Transport Services fiber quick connect system featuring a new type of small-form-factor fiber optic connector, and the Geographically Dispersed Parallel Sysplex using the 9729 Optical Wavelength Division Multiplexer. We also describe the use of optical mode-conditioning patch cables in FIC ON, Gigabit Ethernet, and Parallel Sysplex® data links as a way to reuse an installed multimode cable infrastructure at higher data rates, and toovercome distance-limiting phenomena such as differential mode delay associated with modal power distribution. Technical data is presented for each of these applications, and we discuss how the new fiber optic interconnects fit into the overall IBM strategy for future large systems.

IBM S/390 Parallel Enterprise Server G5 fault tolerance: A historical perspective

Fault tolerance in IBM S/390® systems during the 1980s and 1990s had three distinct phases, each characterized by a different uptime improvement rate. Early TCM-technology mainframes delivered excellent data integrity, instantaneous error detection, and positive fault isolation, but had limited on-line repair. Later TCM mainframes introduced capabilities for providing a high degree of transparent recovery, failure masking, and on-line repair. New challenges accompanied the introduction of CMOS technology. A significant reduction in parts count greatly improved intrinsic failure rates, but dense packaging disallowed on-line CPU repair. In addition, characteristics of the microprocessor technology posed difficulties for traditional in-line error checking. As a result, system fault-tolerant design, particularly in CPUs and memory, underwent another evolution from G1 to G5. G5 implements an innovative design for a high-performance, fault-tolerant single-chip microprocessor. Dynamic CPU sparing delivers a transparent concurrent repair mechanism. A new internal channel provides a high-performance, highly available Parallel Sysplex® in a single mainframe. G5 is both the culmination of decades of innovation and careful implementation, and the highest achievement of S/390 fault-tolerant design.

Shared-storage clusters

Clusters represent a collection of interconnected computers that collaborate on executing an application and present themselves as one unified computing resource. They are becoming an important segment in the computer industry. The two main flavors of cluster architectures are shared-storage and shared-nothing. This article presents host and I/O implementation details, and performance trade-offs that need to be enforced due to sharing data in shared-storage clusters. Sharing data requires the need for global concurrency and coherency protocols to maintain consistency of the database, and enforce data consistency in the local node’s buffers, respectively. Various shared-storage architectures will be investigated, including the Virtual Shared and Shared-Intermediate Memory models. This article also presents few selected shared-storage clusters, including the DEC VAXCluster, IBM parallel Sysplex and Compaq/Tandem ServerNet.

Geographically dispersed Parallel Sysplex architecture using optical wavelength-division multiplexing

Since its introduction in 1994, the Parallel Sysplex architecture for mainframe computer systems has attracted a great deal of interest because of its high performance, continuous availability, scalability, and lower cost of ownership. Recently, it has become possible to interconnect the building blocks of a Parallel Sysplex (host processor, coupling facility, sysplex timer, and disk storage) at extended distances using fiber optic data links. In particular, optical wavelength-division multiplexing (WDM) has been an enabling technology for the extension of this architecture to a geographically dispersed Parallel Sysplex (GDPS). We describe the use of WDM in GDPS configurations, including distance limitations and typical performance considerations. Experimental results of testing a GDPS with WDM channel extensions up to 40 km are presented, and some test results on the IBM 9729 Optical Wavelength Division Multiplexer used in these configurations are also discussed.

Parallel Sysplex: A Scalable, Highly Available, High Performance Commercial System

The IBM System/390 Parallel Sysplex allows a parallel server to grow from a single system to a configuration of 32 systems, and appear as a single image to the end user and applications. It can provide capacity growth for today's largest commercial workloads by enabling a workload to be spread transparently across a collection of systems with shared access to data. By the way of its parallel architecture and operating system support, the Parallel Sysplex offers near-linear scalability and continuous availability for customers' mission-critical applications. Parallel Sysplex optimizes responsiveness and reliability by distributing workloads across all of the processors in the Sysplex. If one or more processors fail, the workload is redistributed across the remaining processors. Because all of the processors have access to all of the data, it can provide a computing environment with near-continuous availability. Empirical performance data are used to demonstrate that Parallel Sysplex provides better scalability than a tightly-coupled multiprocessing system.

Cluster architectures and S/390 Parallel Sysplex scalability

Supporting high transaction rates and high availability for on-line transaction processing and emerging applications requires systems consisting of multiple computing nodes. We outline various cluster architectures and describe the factors that motivate the S/390® Parallel Sysplex™ architecture and its resulting advantages. We quantify the scalability of the S/390 Parallel Sysplex and show that the transaction rate supported is close to linear as nodes are added to the system. The key facet of the S/390 Parallel Sysplex architecture is the coupling facility. The coupling facility provides for very efficient intertransaction concurrency control, buffer cache coherency control, and shared buffer management, among other functions, that lead to the excellent scalability achieved. It also provides for effective dynamic load balancing, high data buffer hit ratios, and load balancing after a failure.

Preface

S/390* has evolved dramatically and fundamentally from wide-ranging and powerful mainframes operating singly or in combination, to a far more wide-ranging and powerful cluster of massively parallel processors designed to handle networks and client/server systems, as well as the earlier mainframe systems. This evolution to the S/390 Parallel Sysplex* cluster has made it practical for applications and whole businesses to migrate gracefully from stand-alone environments to network and client/server environments.

S/390 cluster technology: Parallel Sysplex

This paper describes a clustered multiprocessor system developed for the general-purpose, large-scale commercial marketplace. The system (S/390e Parallel Sysplex™) is based on an architecture designed to combine the benefits of full data sharing and parallel processing in a highly scalable clustered computing environment. The Parallel Sysplex offers significant advantages in the areas of cost, performance range, and availability.

The importance of systems management for a Parallel Sysplex

IBM S/390® Parallel Sysplex™, a multisystem parallel processing environment, provides benefits in terms of reliability, availability, and the total cost of computing. These benefits, however, bring about a systems management challenge because of the increased number of address spaces that need to be managed. This paper describes how CICSPlex® System Manager (SM) for MVS/ESA™ provides simplified systems management of multiple CICS® regions within a sysplex environment. CICSPlex SM provides workload-sensitive balancing of CICS transactions, a single-system image for CICS operations and monitoring, and general-purpose thresholds for resource conditions within CICS. It also describes how CICSPlex SM is integrated with the MVS workload manager, the MVS automatic restart manager, and automation products such as NetView® and the NetView resource object data model (RODM).