With Remote Direct Memory Access (RDMA), you can make a subset of a host's memory directly available to a remote host. RDMA is available on standard Ethernet-based networks by using the RDMA over Converged Ethernet (RoCE) interface. The RoCE network protocol is an industry-standard initiative by the InfiniBand Trade Association. This IBM® Redpaper publication describes how to set up RoCE to use within an IBM Spectrum® Scale cluster and IBM Elastic Storage® Systems (ESSs). This book is targeted at technical professionals (consultants, technical support staff, IT Architects, and IT Specialists) who are responsible for delivering cost-effective storage solutions with IBM Spectrum Scale and IBM ESSs.

This IBM® Redpaper publication focuses on data orchestration in enterprise data pipelines. It provides details about data orchestration and how to address typical challenges that customers face when dealing with large and ever-growing amounts of data for data analytics. While the amount of data increases steadily, artificial intelligence (AI) workloads must speed up to deliver insights and business value in a timely manner. This paper provides a solution that addresses these needs: Data Accelerator for AI and Analytics (DAAA). A proof of concept (PoC) is described in detail. This paper focuses on the functions that are provided by the Data Accelerator for AI and Analytics solution, which simplifies the daily work of data scientists and system administrators. This solution helps increase the efficiency of storage systems and data processing to obtain results faster while eliminating unnecessary data copies and associated data management.

This IBM® Redpaper publication describes the architecture, installation procedure, and results for running a typical training application that works on an automotive data set in an orchestrated and secured environment that provides horizontal scalability of GPU resources across physical node boundaries for deep neural network (DNN) workloads. This paper is mostly relevant for systems engineers, system administrators, or system architects that are responsible for data center infrastructure management and typical day-to-day operations such as system monitoring, operational control, asset management, and security audits. This paper also describes IBM Spectrum® LSF® as a workload manager and IBM Spectrum Discover as a metadata search engine to find the right data for an inference job and automate the data science workflow. With the help of this solution, the data location, which may be on different storage systems, and time of availability for the AI job can be fully abstracted, which provides valuable information for data scientists.

This IBM® Redbooks® publication provides guidance about how to configure, monitor, and manage your IBM System Storage® DS8800 and DS8700 storage systems to achieve optimum performance. It describes the DS8800 and DS8700 performance features and characteristics, including IBM System Storage Easy Tier® and DS8000® I/O Priority Manager. It also describes how they can be used with the various server platforms that attach to the storage system. Then, in separate chapters, we detail specific performance recommendations and discussions that apply for each server environment, as well as for database and DS8000 Copy Services environments. We also outline the various tools available for monitoring and measuring I/O performance for different server environments, as well as describe how to monitor the performance of the entire DS8000 storage system. This book is intended for individuals who want to maximize the performance of their DS8800 and DS8700 storage systems and investigate the planning and monitoring tools that are available. The IBM System Storage DS8800 and DS8700 storage system features, as described in this book, are available for the DS8700 with Licensed Machine Code (LMC) level 6.6.2x.xxx or higher and the DS8800 with Licensed Machine Code (LMC) level 7.6.2x.xxx or higher. For information about optimizing performance with the previous DS8000 models, DS8100 and DS8300, see the following IBM Redbooks publication: DS8000 Performance Monitoring and Tuning, SG24-7146.