Correlating Multiple TB of Performance Data to User Jobs

Similar documents
How To Build A Supermicro Computer With A 32 Core Power Core (Powerpc) And A 32-Core (Powerpc) (Powerpowerpter) (I386) (Amd) (Microcore) (Supermicro) (

Altix Usage and Application Programming. Welcome and Introduction

HPC and Big Data. EPCC The University of Edinburgh. Adrian Jackson Technical Architect

OpenMP Programming on ScaleMP

Parallel Programming Survey

HETEROGENEOUS HPC, ARCHITECTURE OPTIMIZATION, AND NVLINK

Hadoop on the Gordon Data Intensive Cluster

Building a Top500-class Supercomputing Cluster at LNS-BUAP

Computational infrastructure for NGS data analysis. José Carbonell Caballero Pablo Escobar

Purchase of High Performance Computing (HPC) Central Compute Resources by Northwestern Researchers

An Alternative Storage Solution for MapReduce. Eric Lomascolo Director, Solutions Marketing

CORRIGENDUM TO TENDER FOR HIGH PERFORMANCE SERVER

High Performance. CAEA elearning Series. Jonathan G. Dudley, Ph.D. 06/09/ CAE Associates

New Storage System Solutions

Appro Supercomputer Solutions Best Practices Appro 2012 Deployment Successes. Anthony Kenisky, VP of North America Sales

Scientific Computing Data Management Visions

Performance, Reliability, and Operational Issues for High Performance NAS Storage on Cray Platforms. Cray User Group Meeting June 2007

High Performance Computing OpenStack Options. September 22, 2015

Evoluzione dell Infrastruttura di Calcolo e Data Analytics per la ricerca

Hadoop: Embracing future hardware

Performance Characteristics of Large SMP Machines

HPC Cluster Decisions and ANSYS Configuration Best Practices. Diana Collier Lead Systems Support Specialist Houston UGM May 2014

Cluster Implementation and Management; Scheduling

PARALLELS CLOUD STORAGE

Architecting a High Performance Storage System

Lustre & Cluster. - monitoring the whole thing Erich Focht

Deploying and managing a Visualization Onera

Stovepipes to Clouds. Rick Reid Principal Engineer SGI Federal by SGI Federal. Published by The Aerospace Corporation with permission.

Cloud Storage. Parallels. Performance Benchmark Results. White Paper.

NetApp High-Performance Computing Solution for Lustre: Solution Guide

Unstructured Data Accelerator (UDA) Author: Motti Beck, Mellanox Technologies Date: March 27, 2012

Current Status of FEFS for the K computer

SR-IOV In High Performance Computing

IT of SPIM Data Storage and Compression. EMBO Course - August 27th! Jeff Oegema, Peter Steinbach, Oscar Gonzalez

Lustre failover experience

Maurice Askinazi Ofer Rind Tony Wong. Cornell Nov. 2, 2010 Storage at BNL

Commoditisation of the High-End Research Storage Market with the Dell MD3460 & Intel Enterprise Edition Lustre

præsentation oktober 2011

DIABLO TECHNOLOGIES MEMORY CHANNEL STORAGE AND VMWARE VIRTUAL SAN : VDI ACCELERATION

高 通 量 科 学 计 算 集 群 及 Lustre 文 件 系 统. High Throughput Scientific Computing Clusters And Lustre Filesystem In Tsinghua University

Performance Comparison of Intel Enterprise Edition for Lustre* software and HDFS for MapReduce Applications

The Hardware Dilemma. Stephanie Best, SGI Director Big Data Marketing Ray Morcos, SGI Big Data Engineering

Kriterien für ein PetaFlop System

IBM Spectrum Scale vs EMC Isilon for IBM Spectrum Protect Workloads

Lecture 2 Parallel Programming Platforms

The PHI solution. Fujitsu Industry Ready Intel XEON-PHI based solution. SC Denver

HITACHI VIRTUAL STORAGE PLATFORM FAMILY MATRIX

Low-cost

1 Bull, 2011 Bull Extreme Computing

Mississippi State University High Performance Computing Collaboratory Brief Overview. Trey Breckenridge Director, HPC

Transforming the UL into a Big Data University. Current status and planned evolutions

Lustre SMB Gateway. Integrating Lustre with Windows

Agenda. Enterprise Application Performance Factors. Current form of Enterprise Applications. Factors to Application Performance.

JUROPA Linux Cluster An Overview. 19 May 2014 Ulrich Detert

Performance Characteristics of VMFS and RDM VMware ESX Server 3.0.1

Lecture 1: the anatomy of a supercomputer

Lustre Monitoring with OpenTSDB

Using the Intel Xeon Phi (with the Stampede Supercomputer) ISC 13 Tutorial

Application and Micro-benchmark Performance using MVAPICH2-X on SDSC Gordon Cluster

SafePeak Case Study: Large Microsoft SharePoint with SafePeak

Scalable Cloud Computing Solutions for Next Generation Sequencing Data


LLNL Production Plans and Best Practices

Amadeus SAS Specialists Prove Fusion iomemory a Superior Analysis Accelerator

EOFS Workshop Paris Sept, Lustre at exascale. Eric Barton. CTO Whamcloud, Inc Whamcloud, Inc.

IBM System x SAP HANA

Large Scale Storage. Orlando Richards, Information Services LCFG Users Day, University of Edinburgh 18 th January 2013

Traditional v/s CONVRGD

Comparing SMB Direct 3.0 performance over RoCE, InfiniBand and Ethernet. September 2014

ALPS Supercomputing System A Scalable Supercomputer with Flexible Services

Lustre* is designed to achieve the maximum performance and scalability for POSIX applications that need outstanding streamed I/O.

Panasas: High Performance Storage for the Engineering Workflow

NVIDIA Tools For Profiling And Monitoring. David Goodwin

Removing Performance Bottlenecks in Databases with Red Hat Enterprise Linux and Violin Memory Flash Storage Arrays. Red Hat Performance Engineering

HP ProLiant BL660c Gen9 and Microsoft SQL Server 2014 technical brief

New and Improved Lustre Performance Monitoring Tool. Torben Kling Petersen, PhD Principal Engineer. Chris Bloxham Principal Architect

OBSERVEIT DEPLOYMENT SIZING GUIDE

Benchmarking Hadoop & HBase on Violin

HPC Update: Engagement Model

Parallel Processing using the LOTUS cluster

The Data Placement Challenge

1 DCSC/AU: HUGE. DeIC Sekretariat /RB. Bilag 1. DeIC (DCSC) Scientific Computing Installations

Sun Constellation System: The Open Petascale Computing Architecture

PCIe Over Cable Provides Greater Performance for Less Cost for High Performance Computing (HPC) Clusters. from One Stop Systems (OSS)

Future technologies for storage networks. Shravan Pargal Director, Compellent Consulting

Michael Kagan.

Lab Evaluation of NetApp Hybrid Array with Flash Pool Technology

Transcription:

Michael Kluge, ZIH Correlating Multiple TB of Performance Data to User Jobs Lustre User Group 2015, Denver, Colorado Zellescher Weg 12 Willers-Bau A 208 Tel. +49 351-463 34217 Michael Kluge (michael.kluge@tu-dresden.de)

About us ZIH: about us HPC and service provider Research Institute Big Data Competence Center Main research areas: Performance Analysis Tools Energy Efficiency Computer Architecture Standardization Efforts (OpenACC, OpenMP) Michael Kluge, ZIH 2

ZIH HPC and BigData Infrastructure Erlangen, Potsdam TU Dresden ZIH- Backbone 100 Gbit/s Cluster Uplink Home Archiv HTW, Chemnitz, Leipzig, Freiberg Parallel Machine BULL HPC Haswell processors 2 x 612 Nodes ~ 30.000 Cores High Throughput BULL Islands Westmere, Sandy Bridge, und Haswell processors, GPU-Nodes 700 Nodes ~15.000 Cores other clusters Megware Cluster IBM idataplex Shared Memory SGI Ultraviolet 2000 512 cores SandyBridge 8 TB RAM NUMA Storage Michael Kluge, ZIH 3

HRSK-II, Phase 2, currently installed GPU Isle 64 nodes Haswell 2 x 12 cores 64 GB RAM + 2 x Nvidia K 80 Troughput Isle 232 nodes Haswell 2 x 12 cores 64/128/256 GB RAM 128 GB SSD HPC Isle 612 nodes Haswell 2 x 12 cores 64 GB RAM 128 GB SSD HPC Isle 612 nodes Haswell 2 x 12 cores 64 GB RAM 128 GB SSD 64 GB RAM 24 Lustre Server 2 x Export 4 x Login 2 x SMP nodes 2 TB RAM, 48 cores FDR InfiniBand 5 PB NL- SATA 80 GB/s 400.000 IOPS 40 TB SSD 10 GB/s 1 Mio. IOPS to Campus 4 x 10GE Uplink, Home, Backup ~ 1,2 PF peak in CPUs, ~35.000 cores total ~ 300 TF peak in GPUs, 128 cards Michael Kluge, ZIH 4

Architecture of our storage concept (FASS) SCRATCH Switch 2 Switch 1 ZIH-Infrastructure Flexible Storage System (FASS) Login Batch-System Export HPC-Component User A User Z User A User Z Server/File systems Server 1 Transaction Checkpoint. Server 2 Scratch Server N Net Storage SSD SAS SATA Access statistics Analysis Throughput-Component Optimization and control Michael Kluge, ZIH 5

Performance Analysis for I/O Activities (1) What is of interest Application requests, Interface types Access sizes/patterns Network/Server/Storage utilization Level of detail: Record everything Record data every 5 to 10 seconds Challenge: How to analyze this? How to deal with anonymous data? Michael Kluge, ZIH 6

I/O and Job Correlation: Architecture/Software HPC File System SLURM Batch System node node node node I/O network application events node local events utilization errors MDS OSS OSS OSS SAN backend, cache RAID RAID RAID throughput, IOPS Global I/O Data Correlation System Batch System Log Files Cluster Formation Job Clusters Agent Interface Agent Interface Agent Interface Agent Interface Job Clusters with high I/O demands Job Clusters with low I/O demands Database Storage Data Processing Database Storage Database Storage Data Request s Dataheap Data Collection System https://fusionforge.zih.tu-dresden.de/projects/dataheap/ Michael Kluge, ZIH 7

Visualization of Live Data for Phase 1 Michael Kluge, ZIH 8

Visualization of Live Data for Phase 2 Michael Kluge, ZIH 9

Performance Analysis for I/O Activities (2) Amount of collected data: 240 OSTs, 20 OSS servers, Looking at stats and brw_stats (how to store a histogram in database?) ~75.000 tables about 1 GB of data per hour Analysis is supposed to cover 6 month 4 TB data No way to analyze this with any serial approach Michael Kluge, ZIH 10

Analysis Process map operations to a set of tables modify all tables (align data, scale, aggregate over time) create a new table from an input set of tables (aggregation) plot (time series, histograms) currently: single thread Haskell program Swift/T work in progress 4 TB means 50 seconds on an 80 GB/s file system for the initial read Swift/T has nice ways to describe data dependencies and to parallelize this kind of workflows (open for suggestions!) Michael Kluge, ZIH 11

Approach to look at the Raw Data analysis of the raw performance data take original data and create derived metrics look at plots and histograms correlate metrics with user jobs take job log and split jobs into classes check for correlations between job classes and performance metrics Michael Kluge, ZIH 12

Metric Example: NetApp vs. OST Bandwidth (20s) Michael Kluge, ZIH 13

Metric Example: NetApp vs. OST Bandwidth (60s) Michael Kluge, ZIH 14

Metric Example: NetApp vs. OST Bandwidth (10 minutes) Michael Kluge, ZIH 15

Metric Example: 3 Month bandwidth vs. IOPS Michael Kluge, ZIH 16

I/O and Job Correlation: Starting Point (Bandwidth) Michael Kluge, ZIH 17

Building Job Classes a class is defined by: user id, core count similar runtime similar start time (time isolation) sort all jobs by runtime, split list at gaps split these groups again if they are not overlapping (long gap between start) challenges: runtime fluctuation, large run time variations Michael Kluge, ZIH 18

I/O and Job Correlation: Preliminary Result (1) Michael Kluge, ZIH 19

I/O and Job Correlation: Preliminary Result (1) Michael Kluge, ZIH 20

I/O and Job Correlation: Correlation Factors Michael Kluge, ZIH 21

Open Topics How to store a histogram in database? Other approaches than Swift/T Deal with the zeros Cache intermediate results Michael Kluge, ZIH 22

Conclusion / Questions it is possible to map anonymous performance data back to user jobs advance towards storage controller analysis (caches, backend bandwidth, ) Michael Kluge, ZIH 23