DOCSLIB.ORG

Improving Block-Level Efficiency with Scsi-Mq

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Improving Block-Level Efficiency with Scsi-Mq Improving Block Level Efficiency with scsi-mq Blake Caldwell NCCS/ORNL March 4th, 2015 ORNL is managed by UT-Battelle for the US Department of Energy Block Layer Problems • Global lock of request queue per block device • Cache coherency traffic – If servicing part of a request on multiple cores, the lock must be obtained on the new core and invalidated on the old core • Interrupt locality – Hardware interrupt may occur on wrong core, requiring sending soft-interrupt to proper core 2 Improving Block Level Efficiency Linux Block Layer • Designed with rotational media in mind – Time spent in the queue allows sequential request reordering – a very good thing – Completion latencies 10ms to 100ms • Single request queue – Staging area for merging, reordering, scheduling • Drivers are presented with the same interface for each block device 3 Improving Block Level Efficiency blk-mq (multi-queue) • Rewrite of the Linux block layer (since kernel 3.13) • Two levels of queues 1) Per-core submission queues 2) 1 more more hardware dispatch queues with affinity to NUMA nodes/CPU’s (device-driver specific) • IO scheduling within software queues – Inserted in FIFO order, then interleaved to hardware queues • Tags IOs that are reused for lookup on completion 4 Improving Block Level Efficiency Core 0 Core 1 Core 2 Core 3 Core 0 Core 1 Core 2 Core 3 Application Application blk-mq (Core 0, 1, 2) (Core 0, 1, 2) single queue libaio libaio ost_io ost_io Page Page Cache Cache Software Submission queues Hardware Dispatch Queues Block Device Driver Block Device Driver (SRP Initiator) (SRP Initiator) Hardware Parallelism Block Device Block Device (SRP Target) (SRP Target) Disk/SSD Disk/SSD 5 Improving Block Level Efficiency IO Device Affinity LNET To Storage Controller HCA 1 HCA 2 PCIe PCIe QPI CPU 0 CPU 1 NUMA node 0 NUMA node 1 OSS Node 6 Improving Block Level Efficiency Controller Caching (direct) OSS HCA RP0 RP1 RP0 RP1 ddn1 ddn2 Lun 0 Lun 1 Lun 3 Lun 4 7 Improving Block Level Efficiency Controller Caching (indirect) OSS HCA RP0 RP1 RP0 RP1 ddn1 ddn2 Lun 0 Lun 1 Lun 3 Lun 4 8 Improving Block Level Efficiency Evaluation Setup • Linux 3.18 • OSS – blk-mq (3.13) – Dual Ivy Bridge E5-2650 (2 – scsi-mq (3.17) NUMA nodes) – ib-srp multichannel (3.18) – 64GB – dm-multipath support (4.0) – Dual-port QDR IB to array • Lustre 2.7.0 rc1 • Storage Array – ldiskfs patches rebased to 3.18 – Dual DDN 10k controllers – 8GB write-back cache – RAID 6 (8+2) LUNs – SATA disk 9 Improving Block Level Efficiency Evaluation Goals • Block level testing: adapt tests done with null-blk device to a real storage device with scsi-mq – Does NUMA affinity awareness lead to efficiency • Increased bandwidth • Decreased request latency • Multipath performance • Explore benefits for filesystems – Ready for use with fabric attached storage? – Are there performance benefits? 10 Improving Block Level Efficiency Evaluation Parameters • Affinity – IB MSI-x interrupts spread among cores on NUMA node 1 – Fio threads bound to NUMA node 1 (closest to HCA) – Block device rq_affinity=2 – completion happens on submitting core • Block tuning – Noop scheduler – max_sectors_kb = max_hw_sectors_kb • IB-srp module – 16 channels – Max_sect 8192 – Max_cmd_per_lun 62 – Queue size 127 (fixed by hardware) 11 Improving Block Level Efficiency Throughput (direct path) thread/LUN 350000 300000 250000 200000 150000 2.6.32-431 Write IOPS Write 3.18 mq 100000 3.18 mq+mc 50000 0 1 2 4 6 LUNS 12 Improving Block Level Efficiency Throughput (indirect path) thread/LUN 350000 300000 250000 200000 150000 2.6.32-431 Write IOPS IOPS Write 3.18 mq 100000 3.18 mq+mc 50000 0 1 2 4 6 LUNS 13 Improving Block Level Efficiency CPU Usage (direct path) thread/LUN 90 80 70 60 50 usr 2.6.32-431 sys 2.6.32-431 CPU % 40 usr 3.18 mq 30 sys 3.18 mq 20 usr 3.18 mq+mc sys 3.18 mq+mc 10 0 1 2 4 6 8 LUNS 14 Improving Block Level Efficiency Throughput (direct path) 1 LUN 600000 500000 400000 300000 2.6.32-431 Write IOPS IOPS Write 200000 3.18 mq 3.18 mq+mc 100000 0 1 2 4 8 Threads 15 Improving Block Level Efficiency Read Throughput (direct path) 1LUN 350000 300000 250000 200000 2.6 sq 3.18 mq Read IOPS 150000 3.18 mq+mc 100000 50000 0 0 1 2 3 4 5 6 7 8 9 Threads 16 Improving Block Level Efficiency dm-multipath support • Evaluated on Linux 4.0rc1 with patches for dm blk-mq support – 4k IO, libaio, iodepth=127, SRP multi-channel support enabled Multipath Direct (no IO FWD) indirect (IO FWD) 393.2 MB/s 849.5 MB/s 854.8 MB/s 100658 IOPs 217468 IOPs 218829 IOPs 17 Improving Block Level Efficiency Request latency • Average for 100 4k IOs, fio with synchronous IO engine kernel Multipath Direct Indirect (no IO FWD) (IO FWD) 2.6.32-431.17.1 130.50 119.92 169.22 4.0rc1 (blk-mq) 130.56 103.58 142.84 Improvement -0.04% 13.6% 15.6% 18 Improving Block Level Efficiency Lustre Profiling • FlameGraph of kernel code using Perf, 100Hz, Linux 3.18, Lustre 2.7.0rc1, 1MB writes to a single OST 19 Improving Block Level Efficiency Lustre Applications • Metadata IO – Improve single request latency – Is bandwidth necessary during flushing metadata to MDT? • Object IO – Scheduling • Request size • Request tagging 20 Improving Block Level Efficiency Future Directions • Lustre with Linux 4.0 • Testing with hardware capable of 600k+ 4k IOPs – Random write performance for multiple thread/LUN • Evaluate multiple threads/LUN sequential writes • Read and random tests needs further investigation 21 Improving Block Level Efficiency Conclusion • scsi-mq has potential to lower CPU usage even with rotational media • scsi-mq has lower IO completion latency • Further evaluation needed of device drivers that support multiple hardware dispatch queues 22 Improving Block Level Efficiency Thank You Blake Caldwell <[email protected]> 23 Improving Block Level Efficiency .
Load more

Recommended publications
  • SUSE BU Presentation Template 2014
    TUT7317 A Practical Deep Dive for Running High-End, Enterprise Applications on SUSE Linux Holger Zecha Senior Architect REALTECH AG [email protected] Table of Content • About REALTECH • About this session • Design principles • Different layers which need to be considered 2 Table of Content • About REALTECH • About this session • Design principles • Different layers which need to be considered 3 About REALTECH 1/2 REALTECH Software REALTECH Consulting . Business Service Management . SAP Mobile . Service Operations Management . Cloud Computing . Configuration Management and CMDB . SAP HANA . IT Infrastructure Management . SAP Solution Manager . Change Management for SAP . IT Technology . Virtualization . IT Infrastructure 4 About REALTECH 2/2 5 Our Customers Manufacturing IT services Healthcare Media Utilities Consumer Automotive Logistics products Finance Retail REALTECH Consulting GmbH 6 Table of Content • About REALTECH • About this session • Design principles • Different layers which need to be considered 7 The Inspiration for this Session • Several performance workshops at customers • Performance escalations at customer who migrated from UNIX (AIX, Solaris, HP-UX) to Linux • Presenting the experiences made at these customers in this session • Preventing the audience from performance degradation caused from: – Significant design mistakes – Wrong architecture assumptions – Having no architecture at all 8 Performance Optimization The False Estimation Upgrading server with CPUs that are 12.5% faster does not improve application
    [Show full text]
  • On the Performance Variation in Modern Storage Stacks
    On the Performance Variation in Modern Storage Stacks Zhen Cao1, Vasily Tarasov2, Hari Prasath Raman1, Dean Hildebrand2, and Erez Zadok1 1Stony Brook University and 2IBM Research—Almaden Appears in the proceedings of the 15th USENIX Conference on File and Storage Technologies (FAST’17) Abstract tions on different machines have to compete for heavily shared resources, such as network switches [9]. Ensuring stable performance for storage stacks is im- In this paper we focus on characterizing and analyz- portant, especially with the growth in popularity of ing performance variations arising from benchmarking hosted services where customers expect QoS guaran- a typical modern storage stack that consists of a file tees. The same requirement arises from benchmarking system, a block layer, and storage hardware. Storage settings as well. One would expect that repeated, care- stacks have been proven to be a critical contributor to fully controlled experiments might yield nearly identi- performance variation [18, 33, 40]. Furthermore, among cal performance results—but we found otherwise. We all system components, the storage stack is the corner- therefore undertook a study to characterize the amount stone of data-intensive applications, which become in- of variability in benchmarking modern storage stacks. In creasingly more important in the big data era [8, 21]. this paper we report on the techniques used and the re- Although our main focus here is reporting and analyz- sults of this study. We conducted many experiments us- ing the variations in benchmarking processes, we believe ing several popular workloads, file systems, and storage that our observations pave the way for understanding sta- devices—and varied many parameters across the entire bility issues in production systems.
    [Show full text]
  • Vytvoření Softwareově Definovaného Úložiště Pro Potřeby Ukládání a Sdílení Dat V Rámci Instituce a Jeho Zálohování Do Datových Úložišť Cesnet
    Slezská univerzita v Opavě Centrum informačních technologií Vysoká škola báňská – Technická univerzita Ostrava Fakulta elektrotechniky a informatiky Technická zpráva k projektu 529R1/2014 Vytvoření softwareově definovaného úložiště pro potřeby ukládání a sdílení dat v rámci instituce a jeho zálohování do datových úložišť Cesnet Řešitel: Ing. Jiří Sléžka Spoluřešitelé: Mgr. Jan Nosek, Ing. Pavel Nevlud, Ing. Marek Dvorský, Ph.D., Ing. Jiří Vychodil, Ing. Lukáš Kapičák Leden 2016 Obsah 1 Popis projektu...................................................................................................................................1 2 Cíle projektu.....................................................................................................................................1 3 Volba HW a SW komponent.............................................................................................................1 3.1 Hardware...................................................................................................................................1 3.2 GlusterFS..................................................................................................................................2 3.2.1 Replicated GlusterFS Volume...........................................................................................2 3.2.2 Distributed GlusterFS Volume..........................................................................................3 3.2.3 Striped Glusterfs Volume..................................................................................................3
    [Show full text]
  • Buffered FUSE: Optimising the Android IO Stack for User-Level Filesystem
    Int. J. Embedded Systems, Vol. 6, Nos. 2/3, 2014 95 Buffered FUSE: optimising the Android IO stack for user-level filesystem Sooman Jeong and Youjip Won* Hanyang University, #507-2, Annex of Engineering Center, 17 Haengdang-dong, Sungdong-gu, Seoul, 133-791, South Korea E-mail: [email protected] E-mail: [email protected] *Corresponding author Abstract: In this work, we optimise the Android IO stack for user-level filesystem. Android imposes user-level filesystem over native filesystem partition to provide flexibility in managing the internal storage space and to maintain host compatibility. The overhead of user-level filesystem is prohibitively large and the native storage bandwidth is significantly under-utilised. We overhauled the FUSE layer in the Android platform and propose buffered FUSE (bFUSE) to address the overhead of user-level filesystem. The key technical ingredients of buffered FUSE are: 1) extended FUSE IO size; 2) internal user-level write buffer; 3) independent management thread which performs time-driven FUSE buffer synchronisation. With buffered FUSE, we examined the performances of five different filesystems and three disk scheduling algorithms in a combinatorial manner. With bFUSE on XFS filesystem using the deadline scheduling, we achieved the IO performance improvements of 470% and 419% in Android ICS and JB, respectively, over the existing smartphone device. Keywords: Android; storage; user-level filesystem; FUSE; write buffer; embedded systems. Reference to this paper should be made as follows: Jeong, S. and Won, Y. (2014) ‘Buffered FUSE: optimising the Android IO stack for user-level filesystem’, Int. J. Embedded Systems, Vol. 6, Nos.
    [Show full text]
  • Redhat Virtualization Tuning and Optimization Guide
    Red Hat Enterprise Linux 7 Virtualization Tuning and Optimization Guide Using KVM performance features for host systems and virtualized guests on RHEL Last Updated: 2020-09-10 Red Hat Enterprise Linux 7 Virtualization Tuning and Optimization Guide Using KVM performance features for host systems and virtualized guests on RHEL Jiri Herrmann Red Hat Customer Content Services [email protected] Yehuda Zimmerman Red Hat Customer Content Services [email protected] Dayle Parker Red Hat Customer Content Services Scott Radvan Red Hat Customer Content Services Red Hat Subject Matter Experts Legal Notice Copyright © 2019 Red Hat, Inc. This document is licensed by Red Hat under the Creative Commons Attribution-ShareAlike 3.0 Unported License. If you distribute this document, or a modified version of it, you must provide attribution to Red Hat, Inc. and provide a link to the original. If the document is modified, all Red Hat trademarks must be removed. Red Hat, as the licensor of this document, waives the right to enforce, and agrees not to assert, Section 4d of CC-BY-SA to the fullest extent permitted by applicable law. Red Hat, Red Hat Enterprise Linux, the Shadowman logo, the Red Hat logo, JBoss, OpenShift, Fedora, the Infinity logo, and RHCE are trademarks of Red Hat, Inc., registered in the United States and other countries. Linux ® is the registered trademark of Linus Torvalds in the United States and other countries. Java ® is a registered trademark of Oracle and/or its affiliates. XFS ® is a trademark of Silicon Graphics International Corp. or its subsidiaries in the United States and/or other countries.
    [Show full text]
  • Virtualization Best Practices
    SUSE Linux Enterprise Server 12 SP4 Virtualization Best Practices SUSE Linux Enterprise Server 12 SP4 Publication Date: September 24, 2021 Contents 1 Virtualization Scenarios 2 2 Before You Apply Modifications 2 3 Recommendations 3 4 VM Host Server Configuration and Resource Allocation 3 5 VM Guest Images 26 6 VM Guest Configuration 37 7 VM Guest-Specific Configurations and Settings 43 8 Hypervisors Compared to Containers 46 9 Xen: Converting a Paravirtual (PV) Guest to a Fully Virtual (FV/HVM) Guest 50 10 External References 53 1 SLES 12 SP4 1 Virtualization Scenarios Virtualization oers a lot of capabilities to your environment. It can be used in multiple scenarios. For more details refer to Book “Virtualization Guide”, Chapter 1 “Virtualization Technology”, Section 1.2 “Virtualization Capabilities” and Book “Virtualization Guide”, Chapter 1 “Virtualization Technology”, Section 1.3 “Virtualization Benefits”. This best practice guide will provide advice for making the right choice in your environment. It will recommend or discourage the usage of options depending on your workload. Fixing conguration issues and performing tuning tasks will increase the performance of VM Guest's near to bare metal. 2 Before You Apply Modifications 2.1 Back Up First Changing the conguration of the VM Guest or the VM Host Server can lead to data loss or an unstable state. It is really important that you do backups of les, data, images, etc. before making any changes. Without backups you cannot restore the original state after a data loss or a misconguration. Do not perform tests or experiments on production systems. 2.2 Test Your Workloads The eciency of a virtualization environment depends on many factors.
    [Show full text]
  • One Big PDF Volume
    Proceedings of the Linux Symposium July 14–16th, 2014 Ottawa, Ontario Canada Contents Btr-Diff: An Innovative Approach to Differentiate BtrFs Snapshots 7 N. Mandliwala, S. Pimpale, N. P. Singh, G. Phatangare Leveraging MPST in Linux with Application Guidance to Achieve Power and Performance Goals 13 M.R. Jantz, K.A. Doshi, P.A. Kulkarni, H. Yun CPU Time Jitter Based Non-Physical True Random Number Generator 23 S. Müller Policy-extendable LMK filter framework for embedded system 49 K. Baik, J. Kim, D. Kim Scalable Tools for Non-Intrusive Performance Debugging of Parallel Linux Workloads 63 R. Schöne, J. Schuchart, T. Ilsche, D. Hackenberg Veloces: An Efficient I/O Scheduler for Solid State Devices 77 V.R. Damle, A.N. Palnitkar, S.D. Rangnekar, O.D. Pawar, S.A. Pimpale, N.O. Mandliwala Dmdedup: Device Mapper Target for Data Deduplication 83 V. Tarasov, D. Jain, G. Kuenning, S. Mandal, K. Palanisami, P. Shilane, S. Trehan, E. Zadok SkyPat: C++ Performance Analysis and Testing Framework 97 P.H. Chang, K.H. Kuo, D.Y. Tsai, K. Chen, Luba W.L. Tang Computationally Efficient Multiplexing of Events on Hardware Counters 101 R.V. Lim, D. Carrillo-Cisneros, W. Alkowaileet, I.D. Scherson The maxwell(8) random number generator 111 S. Harris Popcorn: a replicated-kernel OS based on Linux 123 A. Barbalace, B. Ravindran, D. Katz Conference Organizers Andrew J. Hutton, Linux Symposium Emilie Moreau, Linux Symposium Proceedings Committee Ralph Siemsen With thanks to John W. Lockhart, Red Hat Robyn Bergeron Authors retain copyright to all submitted papers, but have granted unlimited redistribution rights to all as a condition of submission.
    [Show full text]
  • Xen Cloud Platform Administrator's Guide Release 0.1 0.1
    Xen Cloud Platform Administrator's Guide Release 0.1 0.1 Published October 2009 1.0 Edition 1 Xen Cloud Platform Administrator's Guide Xen Cloud Platform Administrator's Guide: Release 0.1 Published October 2009 Copyright © 2008 Xen®, Xen.org®, Xen Cloud Platform™, and logos are either registered trademarks or trademarks of Xen.org Systems, Inc. in the United States and/or other countries. Other company or product names are for informational purposes only and may be trademarks of their respective owners. This product contains an embodiment of the following patent pending intellectual property of Xen.org Systems, Inc.: 1. United States Non-Provisional Utility Patent Application Serial Number 11/487,945, filed on July 17, 2006, and entitled “Using Writeable Page Tables for Memory Address Translation in a Hypervisor Environment”. 2. United States Non-Provisional Utility Patent Application Serial Number 11/879,338, filed on July 17, 2007, and entitled “Track- ing Current Time on Multiprocessor Hosts and Virtual Machines”. 2 Contents 1. Document Overview ........................................................................................... 9 How this Guide relates to other documentation .................................................. 9 2. Xen Cloud Platform hosts and resource pools ................................................ 11 Hosts and resource pools overview ................................................................. 11 Requirements for creating resource pools ........................................................ 11
    [Show full text]
  • Synergistically Coupling of Solid State Drives and Hard Disks for Qos-Aware Virtual Memory Ke Liu Wayne State University
    Wayne State University DigitalCommons@WayneState Wayne State University Theses 1-1-2013 Synergistically Coupling Of Solid State Drives And Hard Disks For Qos-Aware Virtual Memory Ke Liu Wayne State University, Follow this and additional works at: http://digitalcommons.wayne.edu/oa_theses Recommended Citation Liu, Ke, "Synergistically Coupling Of Solid State Drives And Hard Disks For Qos-Aware Virtual Memory" (2013). Wayne State University Theses. Paper 236. This Open Access Thesis is brought to you for free and open access by DigitalCommons@WayneState. It has been accepted for inclusion in Wayne State University Theses by an authorized administrator of DigitalCommons@WayneState. SYNERGISTICALLY COUPLING OF SOLID STATE DRIVES AND HARD DISKS FOR QOS-AWARE VIRTUAL MEMORY by KE LIU THESIS Submitted to the Graduate School of Wayne State University, Detroit, Michigan in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE 2013 MAJOR: COMPUTER ENGINEERING Approved by: Advisor Date ACKNOWLEDGEMENTS The two years study at Wayne State University is very interesting and full of challenges. This thesis could not be achieved without help from many people in my life. I am very grateful to my advisor, Dr. Song Jiang, for his guidance, support, encouragement all the time. His guidance and extensive knowledge help me get through all the hard times of this work. Thanks to Dr. Kei Davis, who provided me an internship and a fantastic trip at Los Alamos National Laboratory, as his insightful comments on my work help it improved a lot. I would also like to thank the rest of my thesis committee, Prof. Cheng-zhong Xu and Prof.
    [Show full text]
  • Automatic I/O Scheduler Selection Through Online Workload Analysis
    Automatic I/O scheduler selection through online workload analysis Ramon Nou1, Jacobo Giralt1, Toni Cortes1;2 Barcelona Supercomputing Center (Spain)1 - Technical University of Catalonia(UPC)2 [email protected], [email protected], [email protected] Abstract workloads. The I/O scheduler is assigned statically on a per- I/O performance is a bottleneck for many workloads. The disk basis and is normally predefined by the Linux distribution. I/O scheduler plays an important role in it. It is typically On the CPU side, the additional transistors given by the configured once by the administrator and there is no selection advance of the scale of integration are currently addressed that suits the system at every time. Every I/O scheduler has a mainly to increase the number of cores in the chip. Perfor- different behavior depending on the workload and the device. mance achieved by CPUs keeps getting higher, and so does We present a method to select automatically the most suitable the number of CPUs in a system. A similar progression can be I/O scheduler for the ongoing workload. This selection is found for disk capacity, but not for I/O performance. This per- done online, using a workload analysis method with small I/O formance gap is widening with multicores, since the available traces, finding common I/O patterns. Our dynamic mechanism storage bandwidth must be shared among several cores. The adapts automatically to one of the best schedulers, sometimes effects of this performance gap are typically reduced using achieving improvements on I/O performance for heteroge- more storage devices in parallel (like RAID systems), which neous workloads beyond those of any fixed configuration leads to an increased performance in terms of the number of (up to 5%).
    [Show full text]
  • Tuning Linux OS on IBM System P the POWER of Innovation June 2007
    Front cover Tuning Linux OS on System p The POWER of Innovation Linux kernel tuning DB2, WebSphere, and Oracle tuning SystemTap examples Stephen Hochstetler Kraml Liu Andre Metelo Prasanna S Panchamukhi John Zucker ibm.com/redbooks International Technical Support Organization Tuning Linux OS on IBM System p The POWER of Innovation June 2007 SG24-7338-00 Note: Before using this information and the product it supports, read the information in “Notices” on page ix. First Edition (June 2007) This edition applies to Red Hat Enterprise Linux AS 4 and SUSE Linux Enterprise Server 10. © Copyright International Business Machines Corporation 2007. All rights reserved. Note to U.S. Government Users Restricted Rights -- Use, duplication or disclosure restricted by GSA ADP Schedule Contract with IBM Corp. Contents Notices . ix Trademarks . x Preface . xi The team that wrote this book . xi Become a published author . xiv Comments welcome. xv Chapter 1. Introduction to the book. 1 1.1 Operating an efficient server: four phase . 2 1.2 Performance tuning guidelines . 3 1.3 Using information from the System p Performance Lab . 3 1.4 Linux Technology Center . 5 1.5 Understanding the organization of this book . 5 Chapter 2. IBM System p hardware . 7 2.1 System p advantages . 8 2.1.1 Performance . 8 2.1.2 Scalability . 9 2.1.3 Reliability, availability, and serviceability. 11 2.1.4 Manageability . 13 2.1.5 Virtualization . 14 2.2 System p product line introduction . 15 2.2.1 POWER servers . 16 2.2.2 PowerPC servers and blades . 23 2.2.3 IntelliStation.
    [Show full text]
  • Linux on System Z - Disk I/O Performance – Part 1
    Webcast from Boeblingen, Germany Linux on System z - Disk I/O Performance – Part 1 Mustafa Mešanović , IBM R&D Germany, System Performance Analyst September 18/19, 2013 © 2013 IBM Corporation 2013-09-18/19 Linux on System z – Disk I/O Performance – Part 1 Trademarks IBM, the IBM logo, and ibm.com are trademarks or registered trademarks of International Business Machines Corp., registered in many jurisdictions worldwide. Other product and service names might be trademarks of IBM or other companies. A current list of IBM trademarks is available on the Web at “Copyright and trademark information” at www.ibm.com/legal/copytrade.shtml. Linux is a registered trademark of Linus Torvalds in the United States, other countries, or both. Oracle and Java are registered trademarks of Oracle and/or its affiliates in the United States, other countries, or both. Other product and service names might be trademarks of IBM or other companies. © 2013 IBM Corporation 2 Source: If applicable, describe source origin 2013-09-18/19 Linux on System z – Disk I/O Performance – Part 1 Agenda . Terms and characteristics in the disk I/O area – I/O scheduler, page cache, direct I/O... Storage server setup – common parts, storage pool striping... Disk I/O configurations for FICON/ECKD and FCP/SCSI – and its possible advantages and bottlenecks . Summary / Conclusion 3 © 2013 IBM Corporation 2013-09-18/19 Linux on System z – Disk I/O Performance – Part 1 Linux Kernel Components involved in Disk I/O Application program Issuing reads and writes Virtual File System (VFS)
    [Show full text]