The document summarizes performance test results of the IBM LTO-6 tape drive. Key findings include:
1) The LTO-6 tape drive achieves a native data rate of up to 160 MB/s, an increase of 14% over the prior LTO-5 drive.
2) Using new LTO-6 media, the drive provides a native capacity of 2,500GB (6,250GB with 2.5:1 compression), a 67% increase over LTO-5 media.
3) Performance tests show the drive meets or exceeds expectations for data rates when reading and writing both new LTO-6 and prior LTO-5 media formats across a range of compression
This document discusses InnoDB compression at Facebook. It describes how compression saves disk space and reduces I/O, allowing fewer servers. Benchmarks show compressed InnoDB performs as well or better than uncompressed. Facebook improved compression by reducing failures, adding statistics, removing compressed pages from redo logs, and adaptive padding. Future work includes more efficient compression and testing larger pages/other algorithms.
This session covers the engineering strategies and lessons learned at IBM creating industry leading in-memory data warehousing technology for use with both cloud and on-premises software. Along with rich in-memory SQL support for OLAP, data mining, and data warehousing leveraging memory optimized parallel vector processing, we’ll showcase the in-database analytics for R, spatial, and the built-in synchronization with Cloudant JSON NoSQL. We'll take a closer look at the architectural strategy for treating RAM as the new disk (and worth avoiding access to), while dramatically constraining the potential cost pressures of in-memory technology. We’ll describe how we designed for super-simplicity with load-and-go no-tuning technology for any size system, and of course… a demo. Ridiculously easy to use and freakishly fast. Not your grandmother’s IBM database.
The document provides guidelines and commands for analyzing z/VM performance. It discusses processor, storage, paging, and server machine guidelines. Key performance metrics like CPU utilization, response time, and throughput are defined. Commands for monitoring load, queues, paging, storage, and more are listed. REORDMON is introduced as a tool for diagnosing reorder processing overhead.
This document summarizes 11 cool features of Defrag.NSF+ v11, a Domino-specific database defragmentation product. It provides automatic scheduling and tagging of databases for defragmentation. It intelligently switches between file and volume defragmentation and analyzes and consolidates freespace to reduce fragmentation. It also includes automated maintenance of system databases and reporting on database health and optimization.
Designing Information Structures For Performance And Reliabilitybryanrandol
This document discusses optimizing database server performance through hardware, operating system, and database design considerations. It covers topics like CPU performance, memory architecture, disk I/O, and database types like OLTP and OLAP. The document compares GreenPlum and PostgreSQL databases and explains how to tweak PostgreSQL configuration parameters to optimize performance.
Defrag.NSF+ is a Domino-specific database defragmentation tool that runs as a Domino server task. It intelligently auto-switches between file-level and volume-level defragmentation. Key features include automatic scheduling and tagging of databases for defragmentation, analyzing and reducing database fragmentation, and automated maintenance of system databases. Regular defragmentation with Defrag.NSF+ can significantly speed up backup times and improve database performance.
This document discusses InnoDB compression at Facebook. It describes how compression saves disk space and reduces I/O, allowing fewer servers. Benchmarks show compressed InnoDB performs as well or better than uncompressed. Facebook improved compression by reducing failures, adding statistics, removing compressed pages from redo logs, and adaptive padding. Future work includes more efficient compression and testing larger pages/other algorithms.
This session covers the engineering strategies and lessons learned at IBM creating industry leading in-memory data warehousing technology for use with both cloud and on-premises software. Along with rich in-memory SQL support for OLAP, data mining, and data warehousing leveraging memory optimized parallel vector processing, we’ll showcase the in-database analytics for R, spatial, and the built-in synchronization with Cloudant JSON NoSQL. We'll take a closer look at the architectural strategy for treating RAM as the new disk (and worth avoiding access to), while dramatically constraining the potential cost pressures of in-memory technology. We’ll describe how we designed for super-simplicity with load-and-go no-tuning technology for any size system, and of course… a demo. Ridiculously easy to use and freakishly fast. Not your grandmother’s IBM database.
The document provides guidelines and commands for analyzing z/VM performance. It discusses processor, storage, paging, and server machine guidelines. Key performance metrics like CPU utilization, response time, and throughput are defined. Commands for monitoring load, queues, paging, storage, and more are listed. REORDMON is introduced as a tool for diagnosing reorder processing overhead.
This document summarizes 11 cool features of Defrag.NSF+ v11, a Domino-specific database defragmentation product. It provides automatic scheduling and tagging of databases for defragmentation. It intelligently switches between file and volume defragmentation and analyzes and consolidates freespace to reduce fragmentation. It also includes automated maintenance of system databases and reporting on database health and optimization.
Designing Information Structures For Performance And Reliabilitybryanrandol
This document discusses optimizing database server performance through hardware, operating system, and database design considerations. It covers topics like CPU performance, memory architecture, disk I/O, and database types like OLTP and OLAP. The document compares GreenPlum and PostgreSQL databases and explains how to tweak PostgreSQL configuration parameters to optimize performance.
Defrag.NSF+ is a Domino-specific database defragmentation tool that runs as a Domino server task. It intelligently auto-switches between file-level and volume-level defragmentation. Key features include automatic scheduling and tagging of databases for defragmentation, analyzing and reducing database fragmentation, and automated maintenance of system databases. Regular defragmentation with Defrag.NSF+ can significantly speed up backup times and improve database performance.
We4IT lcty 2013 - infra-man - domino run faster We4IT Group
The document discusses optimizing performance for IBM Lotus Domino. It recommends using 64-bit hardware and operating systems to allow Domino to utilize more memory. Transaction logging and separating disks for data, transaction logs, and indexes are also advised. The document provides tips for configuring hardware, operating systems, and Domino server settings to improve performance.
A First Look at the DB2 10 DSNZPARM ChangesWillie Favero
This document discusses changes to DB2 subsystem parameter module (DSNZPARM) in DB2 10. It provides information on DSNZPARM macros, how parameters can be changed through installation panels or dynamically using -SET SYSPARM command, and differences between hidden, opaque and visible parameters. The document also introduces new documentation for opaque parameters and explains how to display current DSNZPARM settings using sample program DSN8ED7.
The document discusses implementing a Parallel Sysplex which couples multiple z/OS systems together using hardware and software services. Key steps include defining coupling facility structures, configuring XCF signalling paths using CTCs or a coupling facility, formatting and configuring sysplex couple data sets, and defining CFRM policies to manage coupling facility resources.
This document provides an overview of NVM compression, a hybrid flash-aware application level compression solution. It discusses the drawbacks of existing row-level compression in MySQL and outlines an architecture for NVM compression that avoids these drawbacks. Key aspects of the NVM compression approach include performing compression only during flush, using sparse addressing to avoid over-provisioning flash space, and adding a new multi-threaded flush framework. Evaluation results and building blocks of the solution are also briefly mentioned.
TokuDB is an ACID/transactional storage engine that makes MySQL even better by increasing performance, adding high compression, and allowing for true schema agility. All of these features are made possible by Tokutek's Fractal Tree indexes.
Ibm db2 10.5 for linux, unix, and windows getting started with db2 installa...bupbechanhgmail
This document provides instructions for installing and configuring IBM DB2 10.5 on Linux and Windows systems. It covers prerequisites such as disk space and memory requirements. It also provides step-by-step instructions for installing DB2 using the Setup wizard on both Windows and Linux. Additional sections describe verifying the installation, configuring licensing, and includes appendices on tasks like uninstalling DB2, checking for updates and applying fix packs.
Best practices for DB2 for z/OS log based recoveryFlorence Dubois
The need to perform a DB2 log-based recovery of multiple objects is a very rare event, but statistically, it is more frequent than a true disaster recovery event (flood, fire, etc). Taking regular backups is necessary but far from sufficient for anything beyond minor application recovery. If not prepared, practiced and optimised, it can lead to extended application service downtimes – possibly many hours to several days. This presentation will provide many hints and tips on how to plan, design intelligently, stress test and optimise DB2 log-based recovery.
Planning and What's New in Windows Server 2008 R2 SP1 for VirtualizationLai Yoong Seng
This document discusses new features in Windows Server 2008 R2 SP1 that are useful for virtualization. It introduces the Microsoft Assessment and Planning Toolkit (MAP) which can be used to assess environments and make recommendations for technologies like server consolidation and virtualization. It then focuses on two new features: Dynamic Memory, which allows VMs to dynamically adjust their memory usage based on demand; and RemoteFX, which enables 3D graphics and remote desktop experiences. It provides examples of how Dynamic Memory works and how it can reduce memory waste. It also outlines the supported guest operating systems and steps for upgrading to take advantage of these features.
A Performance Characterization of Postgres on Different Storage SystemsDong Ye
This document evaluates the performance of PostgreSQL on different storage systems including Fibre Channel, NFS, local storage, and VSAN using benchmarks like TPC-C and TPC-H. It finds that Fibre Channel through RDM/Passthrough and NFS mounted by the hypervisor performed best on the TPC-C benchmark, while local SSD storage was fastest on the TPC-H benchmark. The results provide insights into which storage systems are best suited for different types of PostgreSQL workloads.
The document discusses IBM's POWER7 technology and Power 755 server. It provides details on the POWER7 processor including its 8 cores, 32 threads per chip, and 32MB on-chip memory. It compares POWER7's performance against Intel's Nehalem and Westmere processors, noting POWER7's advantages in core count, cache size, memory bandwidth, and scalability. The Power 755 server is highlighted as delivering high performance for HPC workloads with better performance and efficiency than competitors.
DB2 is a database manager that runs on Linux, Unix, and Windows operating systems. It allows users to catalog databases, start and stop instances, and configure parameters. Key commands for managing DB2 include db2icrt for creating instances, db2idrop for dropping instances, db2ilist for listing instances, and db2set for setting configuration parameters at the global, instance, and node level. The db2set command provides centralized control over environmental variables.
This document discusses best practices for optimizing SQL Server performance. It recommends establishing a baseline, identifying bottlenecks, making one change at a time and measuring the impact. It also provides examples of metrics, tools and techniques to monitor performance at the system, database and query levels. These include Windows Performance Monitor, SQL Server Activity Monitor, Dynamic Management Views and trace flags.
The document is a technical presentation about the IBM System x3850 X5 server from March 2009. It provides an overview of the flagship 4-socket, 4U rack-optimized server including specifications such as support for 4 Intel Xeon 7500-series CPUs, 64 DDR3 DIMMs, 6 PCIe slots, RAID support, and 2 10Gb Ethernet ports. Details are given about optimal configurations for database and virtualization workloads.
Spring 2013 coordinated ims and db2 recoveryJessica Toy
IBM offers several solutions for coordinating IMS and DB2 backup and recovery. Approach 1 uses a single system level backup (SLB) containing both IMS and DB2 volumes, while Approach 2 uses separate SLBs for IMS and DB2 with point-in-time recovery log application. Both approaches minimize downtime and data loss during disaster recovery. FlashCopy is used to quickly create SLBs with minimal host impact. Recovery time objectives (RTOs) and recovery point objectives (RPOs) are improved through frequent SLB creation and transmission to remote sites.
Write Amplification: An Analysis of In-Memory Database Durability TechniquesJaemyung Kim
Write Amplification: An Analysis of In-Memory Database Durability Techniques - 2015 Third International Workshop on In-Memory Data Management and Analytics
Ibm spectrum scale fundamentals workshop for americas part 4 spectrum scale_r...xKinAnx
This document provides information about replication and stretch clusters in IBM Spectrum Scale. It defines replication as synchronously copying file system data across failure groups for redundancy. While replication improves availability, it reduces performance and increases storage usage. Stretch clusters combine two or more clusters to create a single large cluster, typically using replication between sites. Replication policies and failure group configuration are important to ensure effective data duplication.
Btrfs: Design, Implementation and the Current StatusLukáš Czerner
This document provides an overview of the Btrfs file system including its design, implementation, features and current status. Btrfs was created to address limitations in other Linux file systems like ext4 and XFS by providing improved scalability, reliability, and advanced features. It uses a copy-on-write design with checksumming and flexible btrees to store metadata and data. Key current features include snapshots, checksumming, scrubbing, and volume management. The file system is actively developed but aspects like the file system check and raid 5/6 support remain incomplete.
Apache Traffic Server is an open source HTTP proxy and caching server. It provides high performance content delivery through caching, request multiplexing, and connection pooling. The document discusses Traffic Server's history and features, including its multithreaded event-driven architecture, caching capabilities, clustering support, and extensive configuration options. It also addresses how Traffic Server can improve performance and ease operations through automatic restart, plugin extensions, and statistics collection.
Better performance and cost effectiveness empower better results in the cognitive era. For more information, visit: http://www.ibm.com/systems/power/hardware/linux-lc.html
The IBM System Storage® TS3310 Tape Library is a modular, scalable tape library designed to address the tape storage needs of rapidly growing companies that find themselves space- and resource-constrained with tape backup and other tape applications. Learn more: http://ibm.co/ONeH7m
For organizations with rack-mount servers, the IBM 7226 Multimedia Storage Enclosure provides reliable and flexible data backup and security in the data center. With high-speed, serial-attached SCSI (SAS), USB and Fibre Channel (FC) interface options, as well as compatibility with a range of storage devices, the 7226 Multimedia Storage Enclosure can help protect data on critical IBM Power Systems™.
We4IT lcty 2013 - infra-man - domino run faster We4IT Group
The document discusses optimizing performance for IBM Lotus Domino. It recommends using 64-bit hardware and operating systems to allow Domino to utilize more memory. Transaction logging and separating disks for data, transaction logs, and indexes are also advised. The document provides tips for configuring hardware, operating systems, and Domino server settings to improve performance.
A First Look at the DB2 10 DSNZPARM ChangesWillie Favero
This document discusses changes to DB2 subsystem parameter module (DSNZPARM) in DB2 10. It provides information on DSNZPARM macros, how parameters can be changed through installation panels or dynamically using -SET SYSPARM command, and differences between hidden, opaque and visible parameters. The document also introduces new documentation for opaque parameters and explains how to display current DSNZPARM settings using sample program DSN8ED7.
The document discusses implementing a Parallel Sysplex which couples multiple z/OS systems together using hardware and software services. Key steps include defining coupling facility structures, configuring XCF signalling paths using CTCs or a coupling facility, formatting and configuring sysplex couple data sets, and defining CFRM policies to manage coupling facility resources.
This document provides an overview of NVM compression, a hybrid flash-aware application level compression solution. It discusses the drawbacks of existing row-level compression in MySQL and outlines an architecture for NVM compression that avoids these drawbacks. Key aspects of the NVM compression approach include performing compression only during flush, using sparse addressing to avoid over-provisioning flash space, and adding a new multi-threaded flush framework. Evaluation results and building blocks of the solution are also briefly mentioned.
TokuDB is an ACID/transactional storage engine that makes MySQL even better by increasing performance, adding high compression, and allowing for true schema agility. All of these features are made possible by Tokutek's Fractal Tree indexes.
Ibm db2 10.5 for linux, unix, and windows getting started with db2 installa...bupbechanhgmail
This document provides instructions for installing and configuring IBM DB2 10.5 on Linux and Windows systems. It covers prerequisites such as disk space and memory requirements. It also provides step-by-step instructions for installing DB2 using the Setup wizard on both Windows and Linux. Additional sections describe verifying the installation, configuring licensing, and includes appendices on tasks like uninstalling DB2, checking for updates and applying fix packs.
Best practices for DB2 for z/OS log based recoveryFlorence Dubois
The need to perform a DB2 log-based recovery of multiple objects is a very rare event, but statistically, it is more frequent than a true disaster recovery event (flood, fire, etc). Taking regular backups is necessary but far from sufficient for anything beyond minor application recovery. If not prepared, practiced and optimised, it can lead to extended application service downtimes – possibly many hours to several days. This presentation will provide many hints and tips on how to plan, design intelligently, stress test and optimise DB2 log-based recovery.
Planning and What's New in Windows Server 2008 R2 SP1 for VirtualizationLai Yoong Seng
This document discusses new features in Windows Server 2008 R2 SP1 that are useful for virtualization. It introduces the Microsoft Assessment and Planning Toolkit (MAP) which can be used to assess environments and make recommendations for technologies like server consolidation and virtualization. It then focuses on two new features: Dynamic Memory, which allows VMs to dynamically adjust their memory usage based on demand; and RemoteFX, which enables 3D graphics and remote desktop experiences. It provides examples of how Dynamic Memory works and how it can reduce memory waste. It also outlines the supported guest operating systems and steps for upgrading to take advantage of these features.
A Performance Characterization of Postgres on Different Storage SystemsDong Ye
This document evaluates the performance of PostgreSQL on different storage systems including Fibre Channel, NFS, local storage, and VSAN using benchmarks like TPC-C and TPC-H. It finds that Fibre Channel through RDM/Passthrough and NFS mounted by the hypervisor performed best on the TPC-C benchmark, while local SSD storage was fastest on the TPC-H benchmark. The results provide insights into which storage systems are best suited for different types of PostgreSQL workloads.
The document discusses IBM's POWER7 technology and Power 755 server. It provides details on the POWER7 processor including its 8 cores, 32 threads per chip, and 32MB on-chip memory. It compares POWER7's performance against Intel's Nehalem and Westmere processors, noting POWER7's advantages in core count, cache size, memory bandwidth, and scalability. The Power 755 server is highlighted as delivering high performance for HPC workloads with better performance and efficiency than competitors.
DB2 is a database manager that runs on Linux, Unix, and Windows operating systems. It allows users to catalog databases, start and stop instances, and configure parameters. Key commands for managing DB2 include db2icrt for creating instances, db2idrop for dropping instances, db2ilist for listing instances, and db2set for setting configuration parameters at the global, instance, and node level. The db2set command provides centralized control over environmental variables.
This document discusses best practices for optimizing SQL Server performance. It recommends establishing a baseline, identifying bottlenecks, making one change at a time and measuring the impact. It also provides examples of metrics, tools and techniques to monitor performance at the system, database and query levels. These include Windows Performance Monitor, SQL Server Activity Monitor, Dynamic Management Views and trace flags.
The document is a technical presentation about the IBM System x3850 X5 server from March 2009. It provides an overview of the flagship 4-socket, 4U rack-optimized server including specifications such as support for 4 Intel Xeon 7500-series CPUs, 64 DDR3 DIMMs, 6 PCIe slots, RAID support, and 2 10Gb Ethernet ports. Details are given about optimal configurations for database and virtualization workloads.
Spring 2013 coordinated ims and db2 recoveryJessica Toy
IBM offers several solutions for coordinating IMS and DB2 backup and recovery. Approach 1 uses a single system level backup (SLB) containing both IMS and DB2 volumes, while Approach 2 uses separate SLBs for IMS and DB2 with point-in-time recovery log application. Both approaches minimize downtime and data loss during disaster recovery. FlashCopy is used to quickly create SLBs with minimal host impact. Recovery time objectives (RTOs) and recovery point objectives (RPOs) are improved through frequent SLB creation and transmission to remote sites.
Write Amplification: An Analysis of In-Memory Database Durability TechniquesJaemyung Kim
Write Amplification: An Analysis of In-Memory Database Durability Techniques - 2015 Third International Workshop on In-Memory Data Management and Analytics
Ibm spectrum scale fundamentals workshop for americas part 4 spectrum scale_r...xKinAnx
This document provides information about replication and stretch clusters in IBM Spectrum Scale. It defines replication as synchronously copying file system data across failure groups for redundancy. While replication improves availability, it reduces performance and increases storage usage. Stretch clusters combine two or more clusters to create a single large cluster, typically using replication between sites. Replication policies and failure group configuration are important to ensure effective data duplication.
Btrfs: Design, Implementation and the Current StatusLukáš Czerner
This document provides an overview of the Btrfs file system including its design, implementation, features and current status. Btrfs was created to address limitations in other Linux file systems like ext4 and XFS by providing improved scalability, reliability, and advanced features. It uses a copy-on-write design with checksumming and flexible btrees to store metadata and data. Key current features include snapshots, checksumming, scrubbing, and volume management. The file system is actively developed but aspects like the file system check and raid 5/6 support remain incomplete.
Apache Traffic Server is an open source HTTP proxy and caching server. It provides high performance content delivery through caching, request multiplexing, and connection pooling. The document discusses Traffic Server's history and features, including its multithreaded event-driven architecture, caching capabilities, clustering support, and extensive configuration options. It also addresses how Traffic Server can improve performance and ease operations through automatic restart, plugin extensions, and statistics collection.
Better performance and cost effectiveness empower better results in the cognitive era. For more information, visit: http://www.ibm.com/systems/power/hardware/linux-lc.html
The IBM System Storage® TS3310 Tape Library is a modular, scalable tape library designed to address the tape storage needs of rapidly growing companies that find themselves space- and resource-constrained with tape backup and other tape applications. Learn more: http://ibm.co/ONeH7m
For organizations with rack-mount servers, the IBM 7226 Multimedia Storage Enclosure provides reliable and flexible data backup and security in the data center. With high-speed, serial-attached SCSI (SAS), USB and Fibre Channel (FC) interface options, as well as compatibility with a range of storage devices, the 7226 Multimedia Storage Enclosure can help protect data on critical IBM Power Systems™.
The document summarizes performance tests conducted on an IBM FlashSystem 840 storage array.
Key findings from the tests include:
1) The array delivered over 500,000 IOPS in a test with a 100% random workload and 80% reads/20% writes using 4KB blocks.
2) Maximum performance was achieved with a 60-70% read/30-40% write workload mix rather than 100% reads.
3) Updating the array's firmware while under a 100K IOPS workload caused only a brief drop in performance and a slight increase in response times, demonstrating that firmware updates can be performed with minimal impact even under moderate load.
The document provides an overview of IBM Spectrum Archive, which provides a scalable and cost-effective solution for managing the growing storage requirements of big data through the use of IBM LTFS technology and IBM Spectrum Scale. IBM Spectrum Archive utilizes these technologies to extend storage infrastructure to lower-cost tape storage and improve manageability. It is presented as an integrated solution for flash, disk, and tape storage under IBM Spectrum Storage.
Tributary Systems introduces new LTO 6 tape technology for NonStop systems running J-Series operating systems. The LTO 6 tape drives offer higher storage capacity of up to 6.25TB per cartridge, compatibility with LTO 5 and 4 media, and reliability features like error correction. Tributary Systems also continues to enhance its Storage Director backup virtualization solution, which connects any host platform to various storage technologies and applies appropriate data management policies. Storage Director provides benefits like consolidated backup storage, support for tape libraries and deduplication devices, and replication capabilities.
Yesterday's thinking may still believe NVMe (NVM Express) is in transition to a production ready solution. In this session, we will discuss how the evolution of NVMe is ready for production, the history and evolution of NVMe and the Linux stack to address where NVMe has progressed today to become the low latency, highly reliable database key value store mechanism that will drive the future of cloud expansion. Examples of protocol efficiencies and types of storage engines that are optimizing for NVMe will be discussed. Please join us for an exciting session where in-memory computing and persistence have evolved.
Presentation db2 best practices for optimal performancesolarisyougood
This document summarizes best practices for optimizing DB2 performance on various platforms. It discusses sizing workloads based on factors like concurrent users and response time objectives. Guidelines are provided for selecting CPUs, memory, disks and platforms. The document reviews physical database design best practices like choosing a page size and tablespace design. It also discusses index design, compression techniques, and benchmark results showing DB2's high performance.
The document discusses SONAS's performance on the SPECsfs benchmark. It reports that SONAS set a new world record for maximum throughput per file system at 403,000 IOPS. This was achieved using a single file system across 10 interface nodes and 8 storage pods, demonstrating true scale-out without compromising on other factors like capacity or simplicity. Charts show SONAS outperforming other vendors on metrics of performance per file system, aggregate performance, and maintaining high performance as capacity increases.
Learn about IBM Flex System p24L, p260 and p460 Compute Nodes. The IBM Flex System p260 and p460 Compute Nodes are servers based on IBM POWER architecture technologies. These compute nodes run in IBM Flex System Enterprise Chassis units to provide a high-density, high-performance compute node environment, using advanced processing technology. For more information on Pure Systems, visit http://ibm.co/18vDnp6.
Visit the official Scribd Channel of IBM India Smarter Computing at http://bit.ly/VwO86R to get access to more documents.
The document discusses testing done by IBM to evaluate the performance improvements provided by the IBM MAX5 memory expansion technology. The testing showed that by adding 512GB of memory via a MAX5 unit, increasing total memory to 1TB, the following benefits were achieved:
- Response time for business intelligence reports was 1.5-2.8 times faster.
- The cost of producing business intelligence reports could be decreased by 31%-64% over 3 years.
- The throughput of web-facing applications was 2.4-4.9 times greater.
- Read/write response time was decreased by 60%-80%.
This document provides an overview of a training session on storage and the Data Facility Storage Management Subsystem (DFSMS) for z/OS. The training will cover z/OS storage fundamentals, storage systems for z/OS including disk drives, tape drives, and the IBM DS8000 family of storage systems. It will also cover the DFSMS software which manages storage hierarchies and the movement of data between online, nearline, and offline storage devices. Attendees must complete 9 of the 12 listed lectures and all required lab exercises to earn a certificate.
Flash for the Real World – Separate Hype from RealityHitachi Vantara
Join us for a live webcast and hear Hu Yoshida, Chief Technology Officer of Hitachi Data Systems, discuss the real world criteria for making an effective decision when evaluating flash storage. With all the noise in the market it can be difficult to separate fact from fiction in order to evaluate the performance, efficiency and economic trade-offs for flash storage.
Specifically, you’ll learn how to determine if flash storage will help you:
Actually achieve the performance you need as you compare technology options.
Realize efficiency gains that extend beyond the promise of flash performance.
Make the economic case for real-world business decisions before taking the leap.
This white paper talks about multiple performance tuning parameters available for software RAID6 driver in Linux. It helps us understand and optimize these tunables. With improved RAID6 driver architecture in Linux Kernel 3.16.4, it explains how we can achieve similar per data disk performance corresponding to RAID0 driver.
Ceph Day Berlin: Ceph on All Flash Storage - Breaking Performance BarriersCeph Community
The document discusses performance testing done comparing Ceph running on all-flash storage (InfiniFlash OS) versus stock Ceph storage. Testing showed the InfiniFlash OS implementation of Ceph achieved up to 12x better performance for 8K random read workloads and up to 3x better performance for 64K and 256K random read workloads compared to stock Ceph. Performance also scaled linearly as additional InfiniFlash storage nodes were added. The InfiniFlash OS provides an enterprise-hardened version of Ceph optimized for all-flash storage performance.
In this session I will tell you what Hortonworks and IBM Power solutions are and how we can realize significant business value development and prompt use of open innovation in future cognitive utilization. In addition, I will introduce the value added unique to IBM that can be provided by IBM and Hortonworks partnership from the viewpoint of storage, analytics, data science and streaming analysis.
Getting Under the Hood of Kafka Streams: Optimizing Storage Engines to Tune U...HostedbyConfluent
"With a few tweaks under the hood of your Kafka Streams implementation, you greatly improve performance. Sound too good to be true? Well, the secret lies in understanding storage engines.
You may already know that if you're using Kafka streams, you already have a storage engine in place, but do you know what options are available to tune it for optimal performance and scalability?
This presentation will discuss the importance of optimizing and choosing storage engines for Kafka streams applications.
Outline:
- What a storage engine is and how it relates to Kafka stateful streams
- The importance of understanding storage engines for optimal performance and scalability
- Evaluation of Storage Engines - Overview of popular storage engines, including Leveldb, Rocksdb, and Speedb open-source
- Review of the 5 most relevant configurable items and how they affect performance
- Practical ways to optimize and fine tune your storage engine
- Showcase - 2 minutes drop-in replacement demonstration"
This document summarizes tests run on an IBM System x3850 X5 server connected to IBM System Storage DS3400 storage controllers to evaluate how additional memory capacity on the server can benefit SQL Server data warehouse workloads. Six queries were run against a 300GB SQL Server database with 64GB, 128GB, 256GB, and 512GB of server memory. The results showed only minor improvements from 64GB to 128GB, but significant response time reductions of over 50% for most queries when increasing memory to 256GB or 512GB. Additional memory above 128GB provided substantial performance benefits for these data warehouse workloads.
This eight page technical report documents a series of tests that demonstrate the benefits of using large amounts of server memory for a large-scale Decision Support System workload on an eX5 platform
This document discusses the benefits of using Linux on IBM Power systems servers. It claims that Power systems can reduce costs through higher performance, consolidation, and open source software like KVM and OpenStack. It seeks to dispel myths that Power systems are expensive, that virtualization is different, and that the architecture is closed. It provides examples of using Power systems with Linux to gain performance advantages for applications like SAP and databases through higher core counts, memory and bandwidth compared to x86 servers.
Similar to IBM System Storage LTO Ultrium 6 Tape Drive Performance White Paper (20)
This IBM Redpaper provides a brief overview of OpenStack and a basic familiarity of its usage with the IBM XIV Storage System Gen3. The illustration scenario that is presented uses the OpenStack Folsom release implementation IaaS with Ubuntu Linux servers and the IBM Storage Driver for OpenStack. For more information on IBM Storage Systems, visit http://ibm.co/LIg7gk.
Visit http://bit.ly/KWh5Dx to 'Follow' the official Twitter handle of IBM India Smarter Computing.
Learn how all flash needs end to end Storage efficiency. For more information on IBM FlashSystem, visit http://ibm.co/10KodHl.
Visit http://bit.ly/KWh5Dx to 'Follow' the official Twitter handle of IBM India Smarter Computing.
Learn about vSphere Storage API for Array Integration on the IBM Storwize family. IBM Storwize V7000 Unified combines the block storage capabilities of Storwize V7000 with file storage capabilities into a single system for greater ease of management and efficiency. For more information on IBM Storage Systems, visit http://ibm.co/LIg7gk.
Visit http://bit.ly/KWh5Dx to 'Follow' the official Twitter handle of IBM India Smarter Computing.
Learn about IBM FlashSystem 840 and its complete product specification in this Redbook. FlashSystem 840 provides scalable performance for the most demanding enterprise class applications. IBM FlashSystem 840 accelerates response times with IBM MicroLatency to enable faster decision making. For more information on IBM FlashSystem, visit http://ibm.co/10KodHl.
Visit http://on.fb.me/LT4gdu to 'Like' the official Facebook page of IBM India Smarter Computing.
Learn about the IBM System x3250 M5,.The x3250 M5 offers the following energy-efficiency features to save energy, reduce operational costs, increase energy availability, and contribute to a green environment, energy-efficient planar components help lower operational costs. For more information on System x, visit http://ibm.co/Q7m3iQ.
http://www.scribd.com/doc/210746104/IBM-System-x3250-M5
This Redbook talks about the product specification of IBM NeXtScale nx360 M4. The NeXtScale nx360 M4 server provides a dense, flexible solution with a low total cost of ownership (TCO). The half-wide, dual-socket NeXtScale nx360 M4 server is designed for data centers that require high performance but are constrained by floor space. For more information on System x, visit http://ibm.co/Q7m3iQ.
http://www.scribd.com/doc/210745680/IBM-NeXtScale-nx360-M4
The IBM System x3650 M4 HD is a (1) 2-socket 2U rack-optimized server that supports up to 32 internal drives and features an innovative design for optimal performance, uptime, and dense storage. It offers (2) excellent reliability, availability, and serviceability for improved business environments. The server is (3) designed for easy deployment, integration, service, and management.
Here are the product specification for IBM System x3300 M4. This product can be managed remotely.The x3300 M4 server contains IBM IMM2, which provides advanced service-processor control, monitoring, and an alerting function. The IMM2 lights LEDs to help you diagnose the problem, records the error in the event log, and alerts you to the problem. For more information on System x, visit http://ibm.co/Q7m3iQ.
Visit http://on.fb.me/LT4gdu to 'Like' the official Facebook page of IBM India Smarter Computing.
Learn about IBM System x iDataPlex dx360 M4. IBM System x iDataPlex is an innovative data center solution that maximizes performance and optimizes energy and space efficiency. The iDataPlex solution provides customers with outstanding energy and cooling efficiency, multi-rack level manageability, complete flexibility in configuration, and minimal deployment effort. For more information on System x, visit http://ibm.co/Q7m3iQ.
http://www.scribd.com/doc/210744055/IBM-System-x-iDataPlex-dx360-M4
The IBM System x3500 M4 server provides powerful and scalable performance for business applications in an energy efficient tower or rack design. It features the latest Intel Xeon E5-2600 v2 or E5-2600 processors with up to 24 cores, 768GB RAM, 32 hard drives, and 8 PCIe slots. Comprehensive systems management tools and redundant components help ensure high availability, while its small footprint and 80 Plus Platinum power supplies reduce data center costs.
Learn about system specification for IBM System x3550 M4. The x3550 M4 offers numerous features to boost performance, improve scalability, and reduce costs. Improves productivity by offering superior system performance with up to 12-core processors, up to 30 MB of L3 cache, and up to two 8 GT/s QPI interconnect links. For more information on System x, visit http://ibm.co/Q7m3iQ.
Learn about IBM System x3650 M4. The x3650 M4 is an outstanding 2U two-socket business-critical server, offering improved performance and pay-as-you grow flexibility along with new features that improve server management capability. For more information on System x, visit http://ibm.co/Q7m3iQ.
http://www.scribd.com/doc/210741926/IBM-System-x3650-M4
Learn about the product specification of IBM System x3500 M3. System x3500 M3 has an energy-efficient design which works in conjunction with the IMM to govern fan rotation based on the readings that it delivers. This saves money under normal conditions because the fans do not have to spin at high speed. For more information on System x, visit http://ibm.co/Q7m3iQ.
http://www.scribd.com/doc/210741626/IBM-System-x3500-M3
Learn about IBM System x3400 M3. The x3400 M3 offers numerous features to boost performance and reduce costs, x3400 M3 has the ability to grow with your application requirements with these features. Powerful systems management features simplify local and remote management of the x3400 M3. For more information on System x, visit http://ibm.co/Q7m3iQ.
Visit http://on.fb.me/LT4gdu to 'Like' the official Facebook page of IBM India Smarter Computing.
Learn about IBM System 3250 M3 which is a single-socket server that offers new levels of performance and flexibility
to help you respond quickly to changing business demands. Cost-effective and compact, it is well suited to small to mid-sized businesses, as well as large enterprises. For more information on System x, visit http://ibm.co/Q7m3iQ.
http://www.scribd.com/doc/210740347/IBM-System-x3250-M3
Learn about IBM System x3200 M3 and its specifications. The System x3200 M3 features easy installation and management with a rich set of options for hard disk drives and memory. The efficient design helps to save energy and provide a better work environment with less heat and noise. For more information on System x, visit http://ibm.co/Q7m3iQ.
http://www.scribd.com/doc/210739508/IBM-System-x3200-M3
Learn about the configuration of IBM PowerVC. IBM PowerVC is built on OpenStack that controls large pools of server, storage, and networking resources throughout a data center. IBM Power Virtualization Center provides security services that support a secure environment. Installation requires just 20 minutes to get a virtual machine up and running. For more information on Power Systems, visit http://ibm.co/Lx6hfc.
Visit http://on.fb.me/LT4gdu to 'Like' the official Facebook page of IBM India Smarter Computing.
Learn about Ibm POWER7 Virtualization Performance. PowerVM Lx86 is a cross-platform virtualization solution that enables the running of a wide range of x86 Linux applications on Power Systems platforms within a Linux on Power partition without modifications or recompilation of the workloads. For more information on Power Systems, visit http://ibm.co/Lx6hfc.
http://www.scribd.com/doc/210734237/A-Comparison-of-PowerVM-and-Vmware-Virtualization-Performance
This reference architecture document describes deploying the VMware vCloud Enterprise Suite on the IBM PureFlex System hardware platform. Key points:
- The vCloud Suite software provides components for managing and delivering cloud services, while the IBM PureFlex System provides an integrated hardware platform in a single chassis.
- The reference architecture focuses on installing the vCloud Suite management components as virtual machines on an ESXi host to manage consumer resources.
- The IBM PureFlex System provides servers, networking, and storage in a single chassis that can then be easily scaled out. This standardized deployment accelerates provisioning of cloud infrastructure.
- Deployment considerations cover systems management using IBM Flex System Manager, server, networking, storage configurations
Learn how x6: The sixth generation of EXA Technology is fast, agile and Resilient for Emerging Workloads from Alex Yost. Vice President, IBM PureSystems and System x
IBM Systems and Technology Group. x6 drives cloud and big data for enterprises by achieving insight faster thereby outperforming competitors. For more information on System x, visit http://ibm.co/Q7m3iQ.
http://www.scribd.com/doc/210715795/X6-The-sixth-generation-of-EXA-Technology
IBM System Storage LTO Ultrium 6 Tape Drive Performance White Paper
1. IBM® System Storage®
September, 2012
IBM® System Storage® LTO Ultrium 6 Tape Drive
Performance White Paper
By Rogelio Rivera, Tape Performance
Gustavo Vargas, Tape Performance
Marco Vázquez, Tape Performance
IBM Guadalajara
2. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
Table of Contents
Introduction.....................................1
LTO-6 Tape Drive Overview............1
Performance Overview....................2
Performance Evaluation..................2
FC Data rate Performance..............4
SAS Data Rate Performance..........9
Tape Capacity...............................13
Locate and Rewind Performance..14
Load/Unload Performance............15
Speed Matching Performance......16
Conclusions...................................17
The LTO-6 tape drive features
hardware encryption of data, two
interface options, and two form
factors. The data rate improvement
from the previous generation is 14%
and the tape capacity also increases
67% from the previous generation 5.
Introduction
This white paper describes the performance of the IBM® System Storage® LTO
Ultrium 6 tape drive (LTO-6 tape drive) working on open systems environments.
LTO-6 Tape Drive Overview
The sixth generation IBM LTO-6 tape drive offers a great storage capacity and an
excellent performance using technology designed for the mid range open
systems environment that include IBM System p®, IBM System i® and IBM
System x® servers; selected Oracle® and Hewlett Packard® servers
and Intel® servers running supported versions of Microsoft® Windows®
or Linux®.
There are 2 interfaces available for the LTO-6 tape drive:
• 8Gb Fibre Channel (FC-8)
• 6Gb SAS
Both interfaces are available on half and full heights models.
The IBM LTO-6 tape drive offers a native data rate of up to 160 MB/s an increase
of 14% over the previous IBM LTO-5 tape drive.
The IBM LTO-6 tape drive offers a new generation of data cartridge (gen6 media)
that offers a native capacity of 2500GB (6250GB with 2.5:1 compression) 1.7x the
capacity of the previous gen5 tape cartridge generation. This new capacity is
achieved in the same tape length (846m) as the gen5 media due to a higher
density format (gen6 format).
The IBM LTO-6 tape drive will support the Linear Tape File System (LTFS) which
presents the tape storage as a file-based storage system. Additionally the IBM
LTO-6 tape drive is able to read and write previous generation 5 media
and read only generation 4 media to help customers protect their existing
tape investments.
Page 1
3. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
Performance Overview
The key features of the IBM LTO-6 tape drive are designed to improve
performance and capabilities when compared to the IBM LTO-5, IBM LTO-4 and
other vendors tape drives, some of the improvements are:
• Native data rate of up to 160 MB/s
• Native data physical capacity of 2500GB
• The data compression engine improvements- data that compresses with
LTO-5 drives at 2:1 should now compress at 2.5:1
• Faster tape drive mechanism (9 m/s for half height drives) which results
in quicker locate and rewind times
• Support for 8Gb FC and 6Gb SAS connectivity
• SkipSync Function to provide small file backhitchless flush
capability
• Cache buffer: 1024MB (full height models) and 512MB (half height
models)
This white paper examines the performance benchmarks of the IBM LTO-6 tape
drive and associated features. The performance of the different drive host
interface types will also be closely examined.
Performance Evaluation
All the performance benchmarks were run on one or more of the following
systems:
• Server IBM System x3850 running Linux RHEL 6.1
with an LSI® SAS 9212-4i4e HBA and a QLogic®
QLE2562 8Gb FC HBA.
• Server IBM System x3755 running Linux SUSE 11 SP1
with an LSI® SAS 9200-8e HBA and a QLogic® QLE2562 8Gb
FC HBA.
• Server IBM System x3250 running Linux SUSE 11 SP2 with a
QLogic® QLE2562 8Gb FC HBA.
Page 2
4. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
The performance benchmarks used for the tests are a toolbox of in-
house C-based performance measurement tools designed to fully
exercise the host interface and tape drive with the least amount of
overhead. As such, the primary goal of the benchmarks was to provide a
picture of the maximum capabilities of the LTO-6 tape drive. All data
rates/capacity reflect a decimal basis where MB = 1,000,000 bytes and
GB=1,000 MB. Actual tape drive data rate and cartridge capacity might
vary depending on factors such as data compression, server and disk
performance variables.
There are a number of factors that impact performance, especially data
rate at high compression ratios and large block sizes. Server hardware
performance, server slot and operating system/device driver
performance are important factors. Another source of variability in the
data rate performance tests could be due to the firmware used for the
Fibre Channel and SAS connections.
Since this white paper is primarily concerned with looking at LTO-6 tape
drive capabilities, it is beyond the scope of this whitepaper to attempt to
present a complete picture of the relative performance characteristics of
all possible host server/HBA/tape drive combinations. However, when
appropriate an attempt will be made to provide some idea of how
operating system or HBA selection can affect data rate performance.
Page 3
5. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
Fibre Channel Data Rate Performance
The data rate performance of the IBM LTO-6 Fibre Channel full height tape drive
is described by the following set of charts that show how the drive behaves when
writing or reading data that compresses uniformly at ratios 1:1, 2:1, 2.5:1, 3:1 and
maximum (80:1) using differing block sizes and LTO gen6 media.
The IBM LTO-6 tape drive
achieves a native data rate of
160 MB/s with LTO gen6 media.
Higher rates are reached with
compressible data and the new
2.5:1 compression engine.
LTO6 FC FH Write Data Rates Gen6 Media
0
100
200
300
400
500
600
700
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
LTO6 FC FH Read Data Rates Gen6 Media
0
100
200
300
400
500
600
700
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
Page 4
6. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
The following charts show the IBM LTO-6 Fibre Channel full height tape drive
performance when writing or reading LTO gen5 media with data that compresses
uniformly at ratios 1:1, 2:1, 2.5:1, 3:1 and maximum (80:1) using differing block
sizes.
The LTO-6 drive can read and
write LTO gen5 media at the LTO
generation 5 operating point.
With no compressible data a
data rate of 140 MB/s is
achieved.
LTO6 FC FH Write Data Rates Gen5 Media
0
100
200
300
400
500
600
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
LTO6 FC FH Read Data Rates Gen5 Media
0
100
200
300
400
500
600
700
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
Page 5
7. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
The following set of charts show streaming data rate performance of the IBM
LTO-6 Fiber Channel half height tape drive when writing or reading data that
compresses uniformly at ratios 1:1, 2:1, 2.5:1, 3:1 and maximum (80:1) using
differing block sizes and LTO gen6 media.
The half height form factor drive
provides similar streaming
performance as compared with
the full height drive.
LTO6 FC HH Write Data Rates Gen6 Media
0
100
200
300
400
500
600
700
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
LTO6 FC HH Read Data Rates Gen6 Media
0
100
200
300
400
500
600
700
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
Page 6
8. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
The following set of charts describe the IBM LTO-6 Fibre Channel half height tape
drive performance when writing or reading LTO gen5 media with data that
compresses uniformly at ratios 1:1, 2:1, 2.5:1, 3:1 and maximum (80:1) using
differing block sizes.
The half height form factor drive
can read and write LTO Gen5
media with similar rates as
compared with the full height
drive with LTO Gen5 media.
LTO6 FC HH Write Data Rates Gen5 Media
0
100
200
300
400
500
600
700
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
LTO6 FC HH Read Data Rates Gen5 Media
0
100
200
300
400
500
600
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
Page 7
9. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
The above performance benchmarks show IBM LTO-6 Fibre Channel
tape drive performance (full height and half height models) from one or
more of the systems: IBM system x3850 running RHEL 6.1 with a
QLogic® QLE2562 HBA, IBM system x3755 running SUSE 11 SP1 with
a QLogic® QLE2562 HBA and IBM system x3250 running SUSE 11 SP2
with a QLogic® QLE2562 HBA. It must be noted that results from the
IBM system x3850 for Max compression ratio at 512KB block size
dropped about 9% from those shown, however, the rest of the data rates
for all the compression ratios and block sizes were consistent on all
systems. In this case, some aspect of the host-side configuration
decreased the system data rate performance for the specific case
mentioned.
Page 8
10. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
SAS Data Rate Performance
The data rate performance of the IBM LTO-6 SAS full height tape drive is
described by the following set of charts that show how the drive behaves when
writing or reading data that compresses uniformly at ratios 1:1, 2:1, 2.5:1, 3:1 and
maximum (80:1) using differing block sizes and LTO gen6 media.
The SAS data rate performance
capability of the IBM LTO-6 full
height drive meets expectations
at all compression ratios.
LTO6 SAS FH Write Data Rate Gen6 Media
0
50
100
150
200
250
300
350
400
450
500
550
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
LTO6 SAS FH Read Data Rate Gen6 Media
0
50
100
150
200
250
300
350
400
450
500
550
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
Page 9
11. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
The following charts show the IBM LTO-6 SAS full height tape drive performance
when writing or reading LTO gen5 media with data that compresses uniformly at
ratios 1:1, 2:1, 2.5:1, 3:1 and maximum (80:1) using differing block sizes.
The LTO-6 SAS full height drive
meets the LTO gen5 data rates
expectations when using LTO
gen5 media.
LTO6 SAS FH Write Data Rates Gen5 Media
0
50
100
150
200
250
300
350
400
450
500
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
LTO6 SAS FH Read Data Rates Gen5 Media
0
50
100
150
200
250
300
350
400
450
500
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
Page 10
12. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
The following set of charts show streaming data rate performance of the IBM
LTO-6 SAS half height tape drive when writing or reading data that compresses
uniformly at ratios 1:1, 2:1, 2.5:1, 3:1 and maximum (80:1) using differing block
sizes and LTO gen6 media.
The LTO-6 SAS half height form
factor drive provides similar
streaming performance as
compared with the SAS full
height drive.
LTO6 SAS HH Write Data Rates Gen6 Media
0
50
100
150
200
250
300
350
400
450
500
550
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
LTO6 SAS HH Read Data Rates Gen6 Media
0
50
100
150
200
250
300
350
400
450
500
550
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
Page 11
13. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
The following charts show the LTO-6 SAS half height tape drive performance
when writing or reading LTO gen5 media with data that compresses uniformly at
ratios 1:1, 2:1, 2.5:1, 3:1 and maximum (80:1) using differing block sizes.
LTO6 SAS HH Write Data Rates Gen5 Media
0
50
100
150
200
250
300
350
400
450
500
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
LTO6 SAS HH Read Data Rates Gen5 Media
0
50
100
150
200
250
300
350
400
450
500
32 64 128 256 512
Block Size (KB)
Data
Rate
(MB/s)
1:1
2:1
2.5:1
3:1
Max
Page 12
14. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
Tape Capacity
The following chart shows the capacity for LTO gen6 and LTO gen5 media
measured with the LTO-6 tape drive. Tape capacity is obtained by writing 256KB
blocks of uncompressible data until an error code is returned when EOT (End Of
Tape) is reached. The LTO-6 tape drive with gen6 media increases about 67%
over gen5 tape cartridge capacity offering a native physical capacity of 2500GB
while maintaining the 1500 GB expectation with gen5 media.
The LTO-6 tape drive with gen6
media offers a significant
capacity increase over gen5
media.
Tape Capacity
0
500
1000
1500
2000
2500
FC FH FC HH SAS FH SAS HH
Capacity
(GB)
Gen6 Media
Gen5 Media
Page 13
15. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
Locate and Rewind Performance
The following charts show the average time for the LTO-6 drive to locate a
random block id on the tape starting at BOT (Beginning Of Tape), the average
time to locate a random block id starting at some random location on the tape, the
average rewind time as well as maximum times measured for the three
operations. To determine average and maximum times, plenty locate and rewind
operations were made on a completely filled tape.
The full height form factor drives offer faster locate and rewind
performance than the half height form factor drives when using LTO gen6
media. With LTO gen5 media, a similar difference is observed.
The full height form factor drives
offer faster locate and rewind
performance than the half height
form factor drives when using
LTO gen6 media.
LTO6 SAS and FC Locate and Rewind Times
Gen6 Media
91.91
112.1
120.15
49.58
34.28
48.97
103.6
107.9
109.2
56.6
40.7
55.7
92
88.4
97.89
48.8
33.2
48.3
106
120.98
55.6
39.9
55.7
106.81
0
15
30
45
60
75
90
105
120
Avg BOT
Locate
Avg RND
Locate
Avg Rewind Max BOT
Locate
Max RND
Locate
Max Rewind
Seconds
FH FC
HH FC
FH SAS
HH SAS
LTO6 SAS and FC Locate and Rewind Times
Gen5 Media
92.93
92.65
98.53
47.72
38.12
51.01
100.01
100.73
101.66
54.1
38.08
51.94
92.87
93.06
98.12
48
38.26
50.95
99.74
108.06
107.37
53.58
45.41
57.01
0
20
40
60
80
100
120
Avg BOT
Locate
Avg RND
Locate
Avg Rewind Max BOT
Locate
Max RND
Locate
Max Rewind
Seconds
FHFC
HHFC
FHSAS
HHSAS
Page 14
16. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
Load and Unload Performance
The following charts show the tape cartridge load and unload times for the LTO-6
tape drive with LTO gen6 and LTO gen5 media. The performance when loading a
cartridge is slightly better for the full height drives. The time to unload a cartridge
is faster for the full height drives and there is no significant difference regarding
the interface type used.
The IBM LTO-6 tape drive has good
load performance for all form
factors, generation media and
interface types. The time to unload a
cartridge is faster for the full height
form factor drives for both gen6 and
gen5 media.
LTO6 SAS and FC Load and Unload Times
Gen5 Media
11.4
22.26
9.8
17.28
11.48
22.62
9.85
17.35
0
5
10
15
20
25
Load Time Unload Time
Seconds
LTO6 HH FC
LTO6 FHFC
LTO6 HH SAS
LTO6 FHSAS
LTO6 SAS and FC Load and Unload Times
Gen6 Media
21.45
13.35
17.2
11.68
21.34
13.38
17.41
11.52
0
5
10
15
20
25
Load Time Unload Time
Seconds
LTO6 HH FC
LTO6 FH FC
LTO6 HH SAS
LTO6 FH SAS
Page 15
17. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
Speed Matching Performance
The IBM LTO-6 tape drive disposes of 14 defined speeds to match the host data
rate. The implementation of this feature allows to reduce the number of
backhitches when the net host data rate is less than the maximum drive native
data rate, reducing unnecessary tape motion, and may increase overall
performance in certain environments.
To perform this test, host delays are used to vary the speed of the data
from the host to the LTO-6 tape drive. At each host data rate the number
of datasets handled by each speed is recorded (A dataset is the unit of
writing for the drive on the physical tape side and consists of collection of
logical blocks, file marks, ECC, and other format attributes). The Y-axis
shows both the effective data rate and the number of datasets handled
by each speed for the host data rate (X-axis). The chart shows how the
drive selects the appropriate speed from the available 14 speeds to
match the host data rate and most of the datasets written are handled by
the suitable speed.
LTO6 FC FH Write Speed Matching
0
500
1000
1500
2000
2500
3000
3500
4000
32 48 65 82 95 111 124 137 149 159
Host Data Rate (MB/s)
Dataset
Count
0
20
40
60
80
100
120
140
160
180
s1
s2
s3
s4
s5
s6
s7
s8
s9
s10
s11
s12
s13
s14
DR
Speed matching helps to
improve the overall tape drive
data rate at lower host data
rates.
Page 16
18. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
Conclusions
Since the introduction of the first IBM LTO tape drive, every following generation
has incorporated new features and performance improvements to respond
storage needs, now the IBM LTO-6 tape drive in conjunction with the new LTO
gen6 media represents an efficient solution for todays growing storage
demands. Native capacity increase from 1500GB (gen5 media) to
2500GB (gen6 media) and even more with data that is compressible
(6250GB with 2.5:1 compression) while maintaining the same tape
length (846m). This capacity increase does not impact locate/rewind
performance and the faster mechanism for the half height form factors
improves performance for tape motion operations (locate/rewind times)
from previous half height drive generations.
In addition, the IBM LTO-6 tape drive continuous to support media
partitioning, encryption of data and WORM media.
All the improvements make the IBM LTO-6 tape drive suited for handling
backup, restore, and archive of higher data capacities at higher data
transfer rates as compared with the previous LTO-5 tape drive.
Page 17
20. IBM® System Storage® LTO Ultrium 6 Tape Drive
Tape Drive Performance White Paper
THE INFORMATION PROVIDED IN THIS DOCUMENT IS
DISTRIBUTED "AS IS" WITHOUT ANY WARRANTY, EITHER
EXPRESS OR IMPLIED. IBM EXPRESSLY DISCLAIMS ANY
WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR
PURPOSE OR NON-INFRINGEMENT. IBM shall have no responsibility
to update this information. IBM products are warranted according to the
terms and conditions of the agreements (e.g., IBM Customer Agreement,
Statement of Limited Warranty, International Program License
Agreement, etc.) under which they are provided. IBM is not responsible
for the performance or interoperability of any non-IBM products
discussed herein.
The provision of the information contained herein is not intended to, and
does not, grant any right or license under any IBM patents or copyrights.
Inquiries regarding patent or copyright licenses should be made, in
writing, to:
IBM Director of Licensing
IBM Corporation
North Castle Drive
Armonk, NY 10504-1785
U.S.A.
Page 19