Progress Report Chia-Lun Wu 2015.6.9.

Slides:



Advertisements
Similar presentations
Circling Back To Littles Law Now that we have tools to gather information.
Advertisements

Tag line, tag line Perforce Benchmark with PAM over NFS, FCP & iSCSI Bikash R. Choudhury.
Fast and Safe Performance Recovery on OS Reboot Kenichi Kourai Kyushu Institute of Technology.
Recovery CPSC 356 Database Ellen Walker Hiram College (Includes figures from Database Systems by Connolly & Begg, © Addison Wesley 2002)
Milestone 1 Workshop in Information Security – Distributed Databases Project Access Control Security vs. Performance By: Yosi Barad, Ainat Chervin and.
Evaluation of NoSQL databases for DIRAC monitoring and beyond
G Robert Grimm New York University Recoverable Virtual Memory.
User Level Interprocess Communication for Shared Memory Multiprocessor by Bershad, B.N. Anderson, A.E., Lazowska, E.D., and Levy, H.M.
Behaviour and Performance of Interactive Multi-player Game Servers Ahmed Abdelkhalek, Angelos Bilas, and Andreas Moshovos.
OS Fall ’ 02 Performance Evaluation Operating Systems Fall 2002.
Scuola Superiore Sant’Anna Project Assignments Operating Systems.
Performance Evaluation
Adaptive Content Delivery for Scalable Web Servers Authors: Rahul Pradhan and Mark Claypool Presented by: David Finkel Computer Science Department Worcester.
CacheMind: Fast Performance Recovery Using a Virtual Machine Monitor Kenichi Kourai Kyushu Institute of Technology, Japan.
OS Fall ’ 02 Performance Evaluation Operating Systems Fall 2002.
Everest: scaling down peak loads through I/O off-loading D. Narayanan, A. Donnelly, E. Thereska, S. Elnikety, A. Rowstron Microsoft Research Cambridge,
1 The Problem of Power Consumption in Servers L. Minas and B. Ellison Intel-Lab In Dr. Dobb’s Journal, May 2009 Prepared and presented by Yan Cai Fall.
计算机系 信息处理实验室 Lecture 14 Cache Manager
CSCI 6962: Server-side Design and Programming Introduction to AJAX.
Location-aware MapReduce in Virtual Cloud 2011 IEEE computer society International Conference on Parallel Processing Yifeng Geng1,2, Shimin Chen3, YongWei.
Distributed File Systems
1 Wenguang WangRichard B. Bunt Department of Computer Science University of Saskatchewan November 14, 2000 Simulating DB2 Buffer Pool Management.
A User-Lever Concurrency Manager Hongsheng Lu & Kai Xiao.
1 Scheduling The part of the OS that makes the choice of which process to run next is called the scheduler and the algorithm it uses is called the scheduling.
ICOM 6115: Computer Systems Performance Measurement and Evaluation August 11, 2006.
Difference of Degradation Schemes among Operating Systems -Experimental analysis for web application servers- Hideaki Hibino*(Tokyo Tech) Kenichi Kourai.
1 Evaluation of Cooperative Web Caching with Web Polygraph Ping Du and Jaspal Subhlok Department of Computer Science University of Houston presented at.
A remote control robot with webcam. Responsibilities User Interface Communicate with server Webcam Display Server Web Server Collaborators Work: Harkins.
Foundation API The Fast Version. Follow my examples Example: building App for GMAP on Stampede Materials are in the Data Store: – Community Data/iplantcollaborative/example_data/gmap.
1 Chapter 17 Shared Memory Contention. 2 Overview Specifically talking about SGA – Buffer Cache – Redo Log Buffer Contention in these areas of SGA – Can.
LFC Replication Tests LCG 3D Workshop Barbara Martelli.
5/8/09 Titanium Performance Update 2 Dot Hill Confidential.
Data Acquisition in a PACS Weina Ma Sep 24 th, 2013.
Improving Disk Throughput in Data-Intensive Servers Enrique V. Carrera and Ricardo Bianchini Department of Computer Science Rutgers University.
Green Computing Metrics: Power, Temperature, CO2, … Computing system: Many-cores, Clusters, Grids and Clouds Algorithm and model: task scheduling, CFD.
GLOBAL EDGE SOFTWERE LTD1 R EMOTE F ILE S HARING - Ardhanareesh Aradhyamath.
Distributed applications monitoring at system and network level A.Brunengo (INFN- Ge), A.Ghiselli (INFN-Cnaf), L.Luminari (INFN-Roma1), L.Perini (INFN-Mi),
Implement Prefetching Technique Using Supply Chain Theory Lixiong Chen.
EMI INFSO-RI ARC tools for revision and nightly functional tests Jozef Cernak, Marek Kocan, Eva Cernakova (P. J. Safarik University in Kosice, Kosice,
Performance Testing Test Complete. Performance testing and its sub categories Performance testing is performed, to determine how fast some aspect of a.
Online Monitoring System at KLOE Alessandra Doria INFN - Napoli for the KLOE collaboration CHEP 2000 Padova, 7-11 February 2000 NAPOLI.
CSE 451: Operating Systems Section 9: Storage; networks.
Energy Efficient Prefetching and Caching Athanasios E. Papathanasiou and Michael L. Scott. University of Rochester Proceedings of 2004 USENIX Annual Technical.
Display Page (HTML/CSS)
Time Management.  Time management is concerned with OS facilities and services which measure real time.  These services include:  Keeping track of.
Em Spatiotemporal Database Laboratory Pusan National University File Processing : Database Management System Architecture 2004, Spring Pusan National University.
Eris Performance & Load Testing. Performance & Load Testing Basics Part 1.
Disk Drive Architecture Exploration VisualSim Mirabilis Design.
Multithreaded Programming ( ). Components of a Process Figure 3.1.
1 Presented by: Val Pennell, Test Tool Manager Date: March 9, 2004 Software Testing Tools – Load Testing.
Java Autonomous Artificial Intelligence By : Abhishek Singh
This document is provided for informational purposes only and Microsoft makes no warranties, either express or implied, in this document. Information.
Cache Manager Orlando Robertson Brown Bag Seminar
Diskpool and cloud storage benchmarks used in IT-DSS
Implementing Localization
PES Lessons learned from large scale LSF scalability tests
Comparison of LAN, MAN, WAN
Zhen Xiao, Qi Chen, and Haipeng Luo May 2013
Preventing Performance Degradation on Operating System Reboots
Distributed File Systems
Distributed File Systems
Distributed File Systems
February 5, 2004 Adrienne Noble
Software System Performance
Replica Placement Heuristics of Application-level Multicast
Dan Granahan Saagar Patel Shen Li Rob Williams
Ch 9 – Distributed Filesystem
Performance And Scalability In Oracle9i And SQL Server 2000
Distributed File Systems
Distributed File Systems
Presentation transcript:

Progress Report Chia-Lun Wu 2015.6.9

Simulate client/server environment 2 VM (server/client) Python script to simulate client n users -> n threads Call bosc_test binary N users, each client issue M requests with K records

Experiment Settings Parameters BOSC_SCAN_THREESHOLD (10000, 100000) IO_thread max sleep time (60s, 30s) 50 users issue 1000 requests with 10 records Monitor throughput overtime Random + update intensive workload 50% insert, 50% modify

Experiment Results When flushing, throughput degrades Flushing to disk can’t keep up with memory updates After flushing, dirty blocks get more (ex: 1981 -> flush -> 86794) Each user sleep 5s between requests -> dirty blocks get more later End up flushing all the time Need throttling mechanism to limit the rate of memory updates ? Flushing to disk is slow Real-time ingestion of high-frequency small-sized sensor data streams Streaming random disk writes: > 100000 records per second per server

Experiment Results SCAN_THRESHOLD = 10000 SLEEP = 60s No flush (Throughput)

Experiment Results SCAN_THRESHOLD = 10000 SLEEP = 60s Flush

Experiment Results SCAN_THRESHOLD = 100000 SLEEP = 60s No flush

Experiment Results SCAN_THRESHOLD = 100000 SLEEP = 30s No flush

Experiment Results SCAN_THRESHOLD = 10000 SLEEP = 60s 100 Client sleep 5 s every request Experiment Results

Experiment Results SCAN_THRESHOLD = 10000 SLEEP = 60s 100 Client sleep 5 s every request Experiment Results Dirty blocks still get more after flushing Flush Flush Flush Flush Flush

Experiment Results (50 users issue 1000 requests with 10 records) (Average Throughput)