TACC Stats I/O Performance Monitoring for the Intransigent
|
|
|
- Brendan Bruce
- 10 years ago
- Views:
Transcription
1 TACC Stats I/O Performance Monitoring for the Intransigent John Hammond, TACC IASDS11
2 Collaborators Bill Barth, TACC Consumer, first user John McCalpin, TACC Performance guru
3 Motivation Thousands of different users Hundreds of codes Users don t (in general) profile their codes NUMA understanding is low Parallel I/O is hard
4 Goals Monitor all jobs Lightweight No user intervention required Send up red flags Rank jobs by severity of the flag Increase the overall code performance
5 Approach Systematically collect data Tag with current job id Data mine for better site understanding Send report cards to users Contact users with problematic jobs
6 Approach Tried to modify sar (sysstat) Fixed interval, not invokable Serializes to custom format Not job oriented Unaware of Lustre, IB, newer Linux counters
7 tacc_stats is a single exe. Runs: at job startup (batch sched. prolog) every 10 minutes (cron) at job shutdown (batch sched. epilog) Collects data on each node Data stored locally Collected nightly to a central location
8 Data Collected current time, job id, host uptime core and socket perf. counters IB perf. counters Lustre (llite, mdc, osc) statistics OS memory, process statistics network, block device counters
9 Performance counters AMD64 Opteron, per core FLOPS, DCSF, UC, DRAM, HT Intel Nehalem, per core Intel uncore Nehalem, per socket Program /dev/cpu/*/msr in prolog Don't interfere with user
10 InfiniBand ib First attempt, HCA counters 32-bit saturating. WTF? ib_sw Query extended counters on switch port connected to HCA, swap rx, tx. Ranger. ib_ext 64-bit. Lonestar. All use libibmad.
11 Lustre llite (VFS) File operation counters open, close, read, write, seek, flock,... read(), write() tallied by bytes requested not returned L-333, landed Directory operations missing create, lookup, (un)link, rename, {mk,rm,read}dir,... LU-673
12 Lustre mdc RPC opcode oriented close, getattr, setattr, Missing operations MDS_REINT REINT_{CREATE,UNLINK,...} LDLM_ENQUEUE LDLM_{IBITS,FLOCK,...}
13 Lustre osc RPC opcode based + I/O counters Overcounting on error. Must read 300 (#OST) files to get bytes on the wire for /scratch 1800 system calls (stdio), s. LU-334: Add OSC I/O counters to llite.
14 Linux memory Per NUMA node (socket) Total, Free, Used, Cached, Mapped, Anon, Page Tables, Slab, Not totally clear what's what.
15 Linux process/scheduler # Processes, context switches, forks 1, 5, 15 minute load averages Ticks (per core) in user, nice, system, sleep, iowait, irq, softirq, idle.
16 Linux counters, continued Block device (boring). Network interfaces. SysV shared memory. General VFS. VM/paging stats. Weird NUMA allocation stuff.
17 TACC Stats design KISS. Strip-down. Optimize for small /sys, /proc files. Collect raw counter values. Intermediate stats files. Flat text, awk-able. Self-describing, schema with names, types, units. Highly modular, disable types at build or run time.
18 230 main.c 283 dict.c 184 collect.c 119 schema.c 166 stats.c 253 stats_file.c 130 obd_to_mnt.c 147 llite.c 129 mdc.c 132 osc.c 53 lnet.c 177 ib_ext.c 75 block.c 93 cpu.c 134 mem.c 120 net.c 69 numa.c 113 ps.c 84 sysv_shm.c 62 tmpfs.c 78 vfs.c 46 vm.c 238 amd64_pmc.c 322 intel_pmc3.c 331 intel_uncore.c
19 #define KEYS \ X(rd_ios, "E", "read requests processed"), \ X(rd_merges, "E", "read requests merged with in queue requests"), \ X(rd_sectors, "E,U=512B", "sectors read"), \ X(rd_ticks, "E,U=ms", "wait time for read requests"), \ X(wr_ios, "E", "write requests processed"), \ X(wr_merges, "E", "write requests merged with in queue requests"), \ X(wr_sectors, "E,U=512B", "sectors written"), \ X(wr_ticks, "E,U=ms", "wait time for write requests"), \ X(in_flight, "", "requests in flight"), \ X(io_ticks, "E,U=ms", "time active"), \ X(time_in_queue, "E,U=ms", "wait time for all requests") static void collect_block_dev(struct stats_type *type, const char *dev) { struct stats *stats = get_current_stats(type, dev); if (stats == NULL) return; char path[80]; snprintf(path, sizeof(path), "/sys/block/%s/stat", dev); #define X(k,r...) #k collect_key_list(stats, path, KEYS, NULL); #undef X }
20 static void collect_block(struct stats_type *type) { const char *path = "/sys/block"; DIR *dir = opendir(path); if (dir == NULL) { ERROR("cannot open `%s': %m\n", path); return; } struct dirent *ent; while ((ent = readdir(dir))!= NULL) { if (ent >d_name[0] == '.') continue; collect_block_dev(type, ent >d_name); } } if (dir!= NULL) closedir(dir); struct stats_type block_stats_type = {.st_name = "block",.st_collect = &collect_block, #define X SCHEMA_DEF.st_schema_def = JOIN(KEYS), #undef X };
21 $tacc_stats $hostname i ranger.tacc.utexas.edu $uname Linux x86_ el5_TACC \ #18 SMP Mon Mar 14 22:24:19 CDT 2011 $uptime !block rd_ios,e rd_merges,e rd_sectors,e,u=512b \ rd_ticks,e,u=ms wr_ios,e wr_merges,e \ wr_sectors,e,u=512b wr_ticks,e,u=ms in_flight \ io_ticks,e,u=ms time_in_queue,e,u=ms block hda \
22 Data Sizes 1 record: 3.7 KB, ~545 values 1 file (1 node day): ~160 records = 0.5+ MB 10K lines 180 KB compressed 1 node month: 16+ MB 1 Ranger day: 2 GB 1 Ranger month: 60 GB
23
24
25
26
27
28
29 Thanks!
System Administration
Performance Monitoring For a server, it is crucial to monitor the health of the machine You need not only real time data collection and presentation but offline statistical analysis as well Characteristics
OpenMP Programming on ScaleMP
OpenMP Programming on ScaleMP Dirk Schmidl [email protected] Rechen- und Kommunikationszentrum (RZ) MPI vs. OpenMP MPI distributed address space explicit message passing typically code redesign
Performance Counter. Non-Uniform Memory Access Seminar Karsten Tausche 2014-12-10
Performance Counter Non-Uniform Memory Access Seminar Karsten Tausche 2014-12-10 Performance Counter Hardware Unit for event measurements Performance Monitoring Unit (PMU) Originally for CPU-Debugging
Porting Lustre to Operating Systems other than Linux. Ken Hornstein US Naval Research Laboratory April 16, 2010
Porting Lustre to Operating Systems other than Linux Ken Hornstein US Naval Research Laboratory April 16, 2010 Motivation We do a lot of data visualization on Lustre data, and would like to do that on
Multi-Threading Performance on Commodity Multi-Core Processors
Multi-Threading Performance on Commodity Multi-Core Processors Jie Chen and William Watson III Scientific Computing Group Jefferson Lab 12000 Jefferson Ave. Newport News, VA 23606 Organization Introduction
Lecture 1: the anatomy of a supercomputer
Where a calculator on the ENIAC is equipped with 18,000 vacuum tubes and weighs 30 tons, computers of the future may have only 1,000 vacuum tubes and perhaps weigh 1½ tons. Popular Mechanics, March 1949
Lecture 17: Virtual Memory II. Goals of virtual memory
Lecture 17: Virtual Memory II Last Lecture: Introduction to virtual memory Today Review and continue virtual memory discussion Lecture 17 1 Goals of virtual memory Make it appear as if each process has:
Tyche: An efficient Ethernet-based protocol for converged networked storage
Tyche: An efficient Ethernet-based protocol for converged networked storage Pilar González-Férez and Angelos Bilas 30 th International Conference on Massive Storage Systems and Technology MSST 2014 June
Performance of Software Switching
Performance of Software Switching Based on papers in IEEE HPSR 2011 and IFIP/ACM Performance 2011 Nuutti Varis, Jukka Manner Department of Communications and Networking (COMNET) Agenda Motivation Performance
An Implementation Of Multiprocessor Linux
An Implementation Of Multiprocessor Linux This document describes the implementation of a simple SMP Linux kernel extension and how to use this to develop SMP Linux kernels for architectures other than
Configuration Maximums VMware vsphere 4.0
Topic Configuration s VMware vsphere 4.0 When you select and configure your virtual and physical equipment, you must stay at or below the maximums supported by vsphere 4.0. The limits presented in the
VIRTUALIZATION AND CPU WAIT TIMES IN A LINUX GUEST ENVIRONMENT
VIRTUALIZATION AND CPU WAIT TIMES IN A LINUX GUEST ENVIRONMENT James F Brady Capacity Planner for the State Of Nevada [email protected] The virtualization environment presents the opportunity to better
Locating Cache Performance Bottlenecks Using Data Profiling
Locating Cache Performance Bottlenecks Using Data Profiling Aleksey Pesterev Nickolai Zeldovich Robert T. Morris Massachusetts Institute of Technology Computer Science and Artificial Intelligence Lab {alekseyp,
Object Classes and Permissions
Object Classes and Permissions Security Policy Development Primer for Security Enhanced Linux (Module 5) 2 SE Linux Policy Structure Top-level sections of policy.conf: Flask definitions object classes,
How To Understand How A Process Works In Unix (Shell) (Shell Shell) (Program) (Unix) (For A Non-Program) And (Shell).Orgode) (Powerpoint) (Permanent) (Processes
Content Introduction and History File I/O The File System Shell Programming Standard Unix Files and Configuration Processes Programs are instruction sets stored on a permanent medium (e.g. harddisc). Processes
RDMA over Ethernet - A Preliminary Study
RDMA over Ethernet - A Preliminary Study Hari Subramoni, Miao Luo, Ping Lai and Dhabaleswar. K. Panda Computer Science & Engineering Department The Ohio State University Outline Introduction Problem Statement
A Study on the Scalability of Hybrid LS-DYNA on Multicore Architectures
11 th International LS-DYNA Users Conference Computing Technology A Study on the Scalability of Hybrid LS-DYNA on Multicore Architectures Yih-Yih Lin Hewlett-Packard Company Abstract In this paper, the
Chapter 4 System Unit Components. Discovering Computers 2012. Your Interactive Guide to the Digital World
Chapter 4 System Unit Components Discovering Computers 2012 Your Interactive Guide to the Digital World Objectives Overview Differentiate among various styles of system units on desktop computers, notebook
Configuration Maximums
Configuration s vsphere 6.0 This document supports the version of each product listed and supports all subsequent versions until the document is replaced by a new edition. To check for more recent editions
Linux Driver Devices. Why, When, Which, How?
Bertrand Mermet Sylvain Ract Linux Driver Devices. Why, When, Which, How? Since its creation in the early 1990 s Linux has been installed on millions of computers or embedded systems. These systems may
How To Monitor Performance On A Microsoft Powerbook (Powerbook) On A Network (Powerbus) On An Uniden (Powergen) With A Microsatellite) On The Microsonde (Powerstation) On Your Computer (Power
A Topology-Aware Performance Monitoring Tool for Shared Resource Management in Multicore Systems TADaaM Team - Nicolas Denoyelle - Brice Goglin - Emmanuel Jeannot August 24, 2015 1. Context/Motivations
Discovering Computers 2011. Living in a Digital World
Discovering Computers 2011 Living in a Digital World Objectives Overview Differentiate among various styles of system units on desktop computers, notebook computers, and mobile devices Identify chips,
Configuration Maximums
Topic Configuration s VMware vsphere 5.1 When you select and configure your virtual and physical equipment, you must stay at or below the maximums supported by vsphere 5.1. The limits presented in the
Parallel Processing using the LOTUS cluster
Parallel Processing using the LOTUS cluster Alison Pamment / Cristina del Cano Novales JASMIN/CEMS Workshop February 2015 Overview Parallelising data analysis LOTUS HPC Cluster Job submission on LOTUS
A Survey of Parallel Processing in Linux
A Survey of Parallel Processing in Linux Kojiro Akasaka Computer Science Department San Jose State University San Jose, CA 95192 408 924 1000 [email protected] ABSTRACT Any kernel with parallel processing
Lustre & Cluster. - monitoring the whole thing Erich Focht
Lustre & Cluster - monitoring the whole thing Erich Focht NEC HPC Europe LAD 2014, Reims, September 22-23, 2014 1 Overview Introduction LXFS Lustre in a Data Center IBviz: Infiniband Fabric visualization
Binary search tree with SIMD bandwidth optimization using SSE
Binary search tree with SIMD bandwidth optimization using SSE Bowen Zhang, Xinwei Li 1.ABSTRACT In-memory tree structured index search is a fundamental database operation. Modern processors provide tremendous
KVM PERFORMANCE IMPROVEMENTS AND OPTIMIZATIONS. Mark Wagner Principal SW Engineer, Red Hat August 14, 2011
KVM PERFORMANCE IMPROVEMENTS AND OPTIMIZATIONS Mark Wagner Principal SW Engineer, Red Hat August 14, 2011 1 Overview Discuss a range of topics about KVM performance How to improve out of the box experience
Delivering Quality in Software Performance and Scalability Testing
Delivering Quality in Software Performance and Scalability Testing Abstract Khun Ban, Robert Scott, Kingsum Chow, and Huijun Yan Software and Services Group, Intel Corporation {khun.ban, robert.l.scott,
HPC @ CRIBI. Calcolo Scientifico e Bioinformatica oggi Università di Padova 13 gennaio 2012
HPC @ CRIBI Calcolo Scientifico e Bioinformatica oggi Università di Padova 13 gennaio 2012 what is exact? experience on advanced computational technologies a company lead by IT experts with a strong background
Lecture 24 Systems Programming in C
Lecture 24 Systems Programming in C A process is a currently executing instance of a program. All programs by default execute in the user mode. A C program can invoke UNIX system calls directly. A system
Unix::Statgrab - System Monitoring
Unix::Statgrab - System Monitoring Jens Rehsack 2013 Jens Rehsack () Unix::Statgrab - System Monitoring 2013 1 / 28 Overview Part I Introduction 1 Introduction Jens Rehsack () Unix::Statgrab - System Monitoring
How To Monitor Infiniband Network Data From A Network On A Leaf Switch (Wired) On A Microsoft Powerbook (Wired Or Microsoft) On An Ipa (Wired/Wired) Or Ipa V2 (Wired V2)
INFINIBAND NETWORK ANALYSIS AND MONITORING USING OPENSM N. Dandapanthula 1, H. Subramoni 1, J. Vienne 1, K. Kandalla 1, S. Sur 1, D. K. Panda 1, and R. Brightwell 2 Presented By Xavier Besseron 1 Date:
GNS Managed Services Report For abcsun02
GNS Managed Services Report For abcsun02 Report Date: 04-30-2008 Client Details: Name: ABC Funds Address: 242 West 30th Street, Suite 1005, New York, NY 10001 Contact: John Doe Host Details: Hostname:
Performance Characteristics of Large SMP Machines
Performance Characteristics of Large SMP Machines Dirk Schmidl, Dieter an Mey, Matthias S. Müller [email protected] Rechen- und Kommunikationszentrum (RZ) Agenda Investigated Hardware Kernel Benchmark
ELEC 377. Operating Systems. Week 1 Class 3
Operating Systems Week 1 Class 3 Last Class! Computer System Structure, Controllers! Interrupts & Traps! I/O structure and device queues.! Storage Structure & Caching! Hardware Protection! Dual Mode Operation
Architecting a High Performance Storage System
WHITE PAPER Intel Enterprise Edition for Lustre* Software High Performance Data Division Architecting a High Performance Storage System January 2014 Contents Introduction... 1 A Systematic Approach to
Performance Testing. Configuration Parameters for Performance Testing
Optimizing an ecommerce site for performance on a global scale requires additional oversight, budget, dedicated technical resources, local expertise, and specialized vendor solutions to ensure that international
Removing Performance Bottlenecks in Databases with Red Hat Enterprise Linux and Violin Memory Flash Storage Arrays. Red Hat Performance Engineering
Removing Performance Bottlenecks in Databases with Red Hat Enterprise Linux and Violin Memory Flash Storage Arrays Red Hat Performance Engineering Version 1.0 August 2013 1801 Varsity Drive Raleigh NC
Commoditisation of the High-End Research Storage Market with the Dell MD3460 & Intel Enterprise Edition Lustre
Commoditisation of the High-End Research Storage Market with the Dell MD3460 & Intel Enterprise Edition Lustre University of Cambridge, UIS, HPC Service Authors: Wojciech Turek, Paul Calleja, John Taylor
Sun Constellation System: The Open Petascale Computing Architecture
CAS2K7 13 September, 2007 Sun Constellation System: The Open Petascale Computing Architecture John Fragalla Senior HPC Technical Specialist Global Systems Practice Sun Microsystems, Inc. 25 Years of Technical
HPC Hardware Overview
HPC Hardware Overview John Lockman III February 7, 2012 Texas Advanced Computing Center The University of Texas at Austin Outline Some general comments Lonestar System Dell blade-based system InfiniBand
Cray Lustre File System Monitoring
Cray Lustre File System Monitoring esfsmon Jeff Keopp OSIO/ES Systems Cray Inc. St. Paul, MN USA [email protected] Harold Longley OSIO Cray Inc. St. Paul, MN USA [email protected] Abstract The Cray Data Management
CS 377: Operating Systems. Outline. A review of what you ve learned, and how it applies to a real operating system. Lecture 25 - Linux Case Study
CS 377: Operating Systems Lecture 25 - Linux Case Study Guest Lecturer: Tim Wood Outline Linux History Design Principles System Overview Process Scheduling Memory Management File Systems A review of what
Performance monitoring. in the GNU/Linux environment. Linux is like a wigwam - no Windows, no Gates, Apache inside!
1 Performance monitoring in the GNU/Linux environment Linux is like a wigwam - no Windows, no Gates, Apache inside! 2 1 Post-conditions To be familiar with some performance-tuning options To be able to
Comprehensive Resource Use Monitoring for HPC Systems with TACC Stats
214 First International Workshop on HPC User Support Too Comprehensive Resource Use Monitoring for HPC Systems with TACC Stats Todd Evans, William L. Barth, James C. Browne, Robert L. DeLeon, Thomas R.
Configuration Maximums
Configuration s vsphere 6.0 This document supports the version of each product listed and supports all subsequent versions until the document is replaced by a new edition. To check for more recent editions
Process Scheduling in Linux
Process Scheduling in Linux Scheduling Mechanism: how to switch. Scheduling Policy: when to switch and what process to choose. Some scheduling objectives: fast process response time avoidance of process
Distributed File Systems
Distributed File Systems Paul Krzyzanowski Rutgers University October 28, 2012 1 Introduction The classic network file systems we examined, NFS, CIFS, AFS, Coda, were designed as client-server applications.
The Lagopus SDN Software Switch. 3.1 SDN and OpenFlow. 3. Cloud Computing Technology
3. The Lagopus SDN Software Switch Here we explain the capabilities of the new Lagopus software switch in detail, starting with the basics of SDN and OpenFlow. 3.1 SDN and OpenFlow Those engaged in network-related
Configuration Maximums VMware vsphere 4.1
Topic Configuration s VMware vsphere 4.1 When you select and configure your virtual and physical equipment, you must stay at or below the maximums supported by vsphere 4.1. The limits presented in the
Sockets vs. RDMA Interface over 10-Gigabit Networks: An In-depth Analysis of the Memory Traffic Bottleneck
Sockets vs. RDMA Interface over 1-Gigabit Networks: An In-depth Analysis of the Memory Traffic Bottleneck Pavan Balaji Hemal V. Shah D. K. Panda Network Based Computing Lab Computer Science and Engineering
Putting it all together: Intel Nehalem. http://www.realworldtech.com/page.cfm?articleid=rwt040208182719
Putting it all together: Intel Nehalem http://www.realworldtech.com/page.cfm?articleid=rwt040208182719 Intel Nehalem Review entire term by looking at most recent microprocessor from Intel Nehalem is code
High-Performance Processing of Large Data Sets via Memory Mapping A Case Study in R and C++
High-Performance Processing of Large Data Sets via Memory Mapping A Case Study in R and C++ Daniel Adler, Jens Oelschlägel, Oleg Nenadic, Walter Zucchini Georg-August University Göttingen, Germany - Research
Performance Tuning best pracitces and performance monitoring with Zabbix
Performance Tuning best pracitces and performance monitoring with Zabbix Andrew Nelson Senior Linux Consultant May 28, 2015 NLUUG Conf, Utrecht, Netherlands Overview Introduction Performance tuning is
Packet Capture in 10-Gigabit Ethernet Environments Using Contemporary Commodity Hardware
Packet Capture in 1-Gigabit Ethernet Environments Using Contemporary Commodity Hardware Fabian Schneider Jörg Wallerich Anja Feldmann {fabian,joerg,anja}@net.t-labs.tu-berlin.de Technische Universtität
Multicore Processor and GPU. Jia Rao Assistant Professor in CS http://cs.uccs.edu/~jrao/
Multicore Processor and GPU Jia Rao Assistant Professor in CS http://cs.uccs.edu/~jrao/ Moore s Law The number of transistors on integrated circuits doubles approximately every two years! CPU performance
HPC Update: Engagement Model
HPC Update: Engagement Model MIKE VILDIBILL Director, Strategic Engagements Sun Microsystems [email protected] Our Strategy Building a Comprehensive HPC Portfolio that Delivers Differentiated Customer Value
Parallel Large-Scale Visualization
Parallel Large-Scale Visualization Aaron Birkland Cornell Center for Advanced Computing Data Analysis on Ranger January 2012 Parallel Visualization Why? Performance Processing may be too slow on one CPU
Current Status of FEFS for the K computer
Current Status of FEFS for the K computer Shinji Sumimoto Fujitsu Limited Apr.24 2012 LUG2012@Austin Outline RIKEN and Fujitsu are jointly developing the K computer * Development continues with system
Green Telnet. Making the Client/Server Model Green
Green Telnet Reducing energy consumption is of growing importance. Jeremy and Ken create a "green telnet" that lets clients transition to a low-power, sleep state. By Jeremy Blackburn and Ken Christensen,
PERFORMANCE TUNING ORACLE RAC ON LINUX
PERFORMANCE TUNING ORACLE RAC ON LINUX By: Edward Whalen Performance Tuning Corporation INTRODUCTION Performance tuning is an integral part of the maintenance and administration of the Oracle database
How System Settings Impact PCIe SSD Performance
How System Settings Impact PCIe SSD Performance Suzanne Ferreira R&D Engineer Micron Technology, Inc. July, 2012 As solid state drives (SSDs) continue to gain ground in the enterprise server and storage
Operating Systems Design 16. Networking: Sockets
Operating Systems Design 16. Networking: Sockets Paul Krzyzanowski [email protected] 1 Sockets IP lets us send data between machines TCP & UDP are transport layer protocols Contain port number to identify
LoadLeveler Overview. January 30-31, 2012. IBM Storage & Technology Group. IBM HPC Developer Education @ TIFR, Mumbai
IBM HPC Developer Education @ TIFR, Mumbai IBM Storage & Technology Group LoadLeveler Overview January 30-31, 2012 Pidad D'Souza ([email protected]) IBM, System & Technology Group 2009 IBM Corporation
Configuration Maximums VMware Infrastructure 3
Technical Note Configuration s VMware Infrastructure 3 When you are selecting and configuring your virtual and physical equipment, you must stay at or below the maximums supported by VMware Infrastructure
Red Hat Linux Internals
Red Hat Linux Internals Learn how the Linux kernel functions and start developing modules. Red Hat Linux internals teaches you all the fundamental requirements necessary to understand and start developing
Lustre Monitoring with OpenTSDB
Lustre Monitoring with OpenTSDB 2015/9/22 DataDirect Networks Japan, Inc. Shuichi Ihara 2 Lustre Monitoring Background Lustre is a black box Users and Administrators want to know what s going on? Find
Oracle Exadata Database Machine for SAP Systems - Innovation Provided by SAP and Oracle for Joint Customers
Oracle Exadata Database Machine for SAP Systems - Innovation Provided by SAP and Oracle for Joint Customers Masood Ahmed EMEA Infrastructure Solutions Oracle/SAP Relationship Overview First SAP R/3 release
These sub-systems are all highly dependent on each other. Any one of them with high utilization can easily cause problems in the other.
Abstract: The purpose of this document is to describe how to monitor Linux operating systems for performance. This paper examines how to interpret common Linux performance tool output. After collecting
UNIX File Management (continued)
UNIX File Management (continued) OS storage stack (recap) Application FD table OF table VFS FS Buffer cache Disk scheduler Device driver 2 Virtual File System (VFS) Application FD table OF table VFS FS
Configuration Maximums
Topic Configuration s VMware vsphere 5.0 When you select and configure your virtual and physical equipment, you must stay at or below the maximums supported by vsphere 5.0. The limits presented in the
Stovepipes to Clouds. Rick Reid Principal Engineer SGI Federal. 2013 by SGI Federal. Published by The Aerospace Corporation with permission.
Stovepipes to Clouds Rick Reid Principal Engineer SGI Federal 2013 by SGI Federal. Published by The Aerospace Corporation with permission. Agenda Stovepipe Characteristics Why we Built Stovepipes Cluster
Lecture 2 Parallel Programming Platforms
Lecture 2 Parallel Programming Platforms Flynn s Taxonomy In 1966, Michael Flynn classified systems according to numbers of instruction streams and the number of data stream. Data stream Single Multiple
Near-Dedicated Scheduling
Near-Dedicated Scheduling Chris Brady, CRI, Boulder, Colorado, USA, Mary Ann Ciuffini, NCAR, Boulder, Colorado, USA, Bryan Hardy, CRI, Boulder, Colorado, USA ABSTRACT: With the advent of high-performance
Recommended hardware system configurations for ANSYS users
Recommended hardware system configurations for ANSYS users The purpose of this document is to recommend system configurations that will deliver high performance for ANSYS users across the entire range
Parallel Programming Survey
Christian Terboven 02.09.2014 / Aachen, Germany Stand: 26.08.2014 Version 2.3 IT Center der RWTH Aachen University Agenda Overview: Processor Microarchitecture Shared-Memory
Installing, upgrading and troubleshooting your CLIO system under the Windows environment.
Installing, upgrading and troubleshooting your CLIO system under the Windows environment. Audiomatica Srl Rev. 1.1. June 2001. Contents 1 What does this document cover?... 1 2 Windows compatibility...
SAP HANA - Main Memory Technology: A Challenge for Development of Business Applications. Jürgen Primsch, SAP AG July 2011
SAP HANA - Main Memory Technology: A Challenge for Development of Business Applications Jürgen Primsch, SAP AG July 2011 Why In-Memory? Information at the Speed of Thought Imagine access to business data,
How To Build A Supermicro Computer With A 32 Core Power Core (Powerpc) And A 32-Core (Powerpc) (Powerpowerpter) (I386) (Amd) (Microcore) (Supermicro) (
TECHNICAL GUIDELINES FOR APPLICANTS TO PRACE 7 th CALL (Tier-0) Contributing sites and the corresponding computer systems for this call are: GCS@Jülich, Germany IBM Blue Gene/Q GENCI@CEA, France Bull Bullx
Oracle Database Scalability in VMware ESX VMware ESX 3.5
Performance Study Oracle Database Scalability in VMware ESX VMware ESX 3.5 Database applications running on individual physical servers represent a large consolidation opportunity. However enterprises
Running a Workflow on a PowerCenter Grid
Running a Workflow on a PowerCenter Grid 2010-2014 Informatica Corporation. No part of this document may be reproduced or transmitted in any form, by any means (electronic, photocopying, recording or otherwise)
Storage benchmarking cookbook
Storage benchmarking cookbook How to perform solid storage performance measurements Stijn Eeckhaut Stijn De Smet, Brecht Vermeulen, Piet Demeester The situation today: storage systems can be very complex
Cray DVS: Data Virtualization Service
Cray : Data Virtualization Service Stephen Sugiyama and David Wallace, Cray Inc. ABSTRACT: Cray, the Cray Data Virtualization Service, is a new capability being added to the XT software environment with
Building a computer. Electronic Numerical Integrator and Computer (ENIAC)
Building a computer Electronic Numerical Integrator and Computer (ENIAC) CSCI 255: Introduc/on to Embedded Systems Keith Vertanen Copyright 2011 Layers of abstrac
The Linux Kernel: Process Management. CS591 (Spring 2001)
The Linux Kernel: Process Management Process Descriptors The kernel maintains info about each process in a process descriptor, of type task_struct. See include/linux/sched.h Each process descriptor contains
LOAD BALANCING DISTRIBUTED OPERATING SYSTEMS, SCALABILITY, SS 2015. Hermann Härtig
LOAD BALANCING DISTRIBUTED OPERATING SYSTEMS, SCALABILITY, SS 2015 Hermann Härtig ISSUES starting points independent Unix processes and block synchronous execution who does it load migration mechanism
HPC and Big Data. EPCC The University of Edinburgh. Adrian Jackson Technical Architect [email protected]
HPC and Big Data EPCC The University of Edinburgh Adrian Jackson Technical Architect [email protected] EPCC Facilities Technology Transfer European Projects HPC Research Visitor Programmes Training
Alexander Reinefeld, Thorsten Schütt, ZIB 1
Alexander Reinefeld, Thorsten Schütt, ZIB 1 Solving Puzzles with MapReduce Alexander Reinefeld and Thorsten Schütt Zuse Institute Berlin Hadoop Get Together, 29.09.2009, Berlin How to solve very large
Performance Comparison of Intel Enterprise Edition for Lustre* software and HDFS for MapReduce Applications
Performance Comparison of Intel Enterprise Edition for Lustre software and HDFS for MapReduce Applications Rekha Singhal, Gabriele Pacciucci and Mukesh Gangadhar 2 Hadoop Introduc-on Open source MapReduce
CFS-v: I/O Demand-driven VM Scheduler in KVM
CFS-v: Demand-driven VM Scheduler in KVM Hyotaek Shim and Sung-Min Lee (hyotaek.shim, [email protected] com) Software R&D Center, Samsung Electronics 2014. 10. 16 Problem in Server Consolidation 2/16
CIT 470: Advanced Network and System Administration. Topics. Performance Monitoring. Performance Monitoring
CIT 470: Advanced Network and System Administration Performance Monitoring CIT 470: Advanced Network and System Administration Slide #1 Topics 1. Performance monitoring. 2. Performance tuning. 3. CPU 4.
