efficient memory disaggregation with infiniswap
play

Efficient Memory Disaggregation with Infiniswap Juncheng Gu , - PowerPoint PPT Presentation

Sep 27, 2022 •205 likes •857 views

Efficient Memory Disaggregation with Infiniswap Juncheng Gu , Youngmoon Lee, Yiwen Zhang, Mosharaf Chowdhury, Kang G. Shin Agenda Motivation and related work Design and system overview Implementation and evaluation Future work and

  1. Efficient Memory Disaggregation with Infiniswap Juncheng Gu , Youngmoon Lee, Yiwen Zhang, Mosharaf Chowdhury, Kang G. Shin

  2. Agenda • Motivation and related work • Design and system overview • Implementation and evaluation • Future work and conclusion 3/30/17 1

  3. Memory-intensive applications 2 3/30/17

  4. Memory-intensive applications 3 3/30/17

  5. Performance degradation 1 Normalized Performance 0.8 0.6 0.4 0.2 0 VoltDB Memcached PowerGraph GraphX (TPC-C) (Facebook/FB SYS) (TunkRank) (PageRank) 100% working sets in memory 75% working sets in memory 50% working sets in memory 3/30/17 4

  6. Performance degradation 1 1 Normalized Performance Normalized Performance 0.8 0.8 0.6 0.6 0.4 0.4 0.18 0.2 0.2 0 0 VoltDB VoltDB Memcached Memcached PowerGraph PowerGraph GraphX GraphX (TPC-C) (TPC-C) (Facebook/FB SYS) (Facebook/FB SYS) (TunkRank) (TunkRank) (PageRank) (PageRank) 100% working sets in memory 100% working sets in memory 75% working sets in memory 75% working sets in memory 50% working sets in memory 50% working sets in memory 3/30/17 5

  7. Performance degradation 1 1 1 Normalized Performance Normalized Performance Normalized Performance 0.8 0.8 0.8 0.6 0.6 0.6 0.47 0.4 0.4 0.4 0.18 0.18 0.2 0.2 0.2 0 0 0 VoltDB VoltDB VoltDB Memcached Memcached Memcached PowerGraph PowerGraph PowerGraph GraphX GraphX GraphX (TPC-C) (TPC-C) (TPC-C) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (TunkRank) (TunkRank) (TunkRank) (PageRank) (PageRank) (PageRank) 100% working sets in memory 100% working sets in memory 100% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 3/30/17 6

  8. Performance degradation 1 1 1 1 Normalized Performance Normalized Performance Normalized Performance Normalized Performance 0.8 0.8 0.8 0.8 0.6 0.6 0.6 0.6 0.47 0.47 0.4 0.4 0.4 0.4 0.18 0.18 0.18 0.2 0.2 0.2 0.2 0.04 0 0 0 0 VoltDB VoltDB VoltDB VoltDB Memcached Memcached Memcached Memcached PowerGraph PowerGraph PowerGraph PowerGraph GraphX GraphX GraphX GraphX (TPC-C) (TPC-C) (TPC-C) (TPC-C) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (TunkRank) (TunkRank) (TunkRank) (TunkRank) (PageRank) (PageRank) (PageRank) (PageRank) 100% working sets in memory 100% working sets in memory 100% working sets in memory 100% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 3/30/17 7

  9. Performance degradation 1 1 1 1 1 Normalized Performance Normalized Performance Normalized Performance Normalized Performance Normalized Performance 0.8 0.8 0.8 0.8 0.8 0.6 0.6 0.6 0.6 0.6 0.47 0.47 0.47 0.4 0.4 0.4 0.4 0.4 0.18 0.18 0.18 0.18 0.2 0.2 0.2 0.2 0.2 0.06 0.04 0.04 0 0 0 0 0 VoltDB VoltDB VoltDB VoltDB VoltDB Memcached Memcached Memcached Memcached Memcached PowerGraph PowerGraph PowerGraph PowerGraph PowerGraph GraphX GraphX GraphX GraphX GraphX (TPC-C) (TPC-C) (TPC-C) (TPC-C) (TPC-C) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (TunkRank) (TunkRank) (TunkRank) (TunkRank) (TunkRank) (PageRank) (PageRank) (PageRank) (PageRank) (PageRank) 100% working sets in memory 100% working sets in memory 100% working sets in memory 100% working sets in memory 100% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 3/30/17 8

  10. Performance degradation 1 1 1 1 1 1 0.94 Normalized Performance Normalized Performance Normalized Performance Normalized Performance Normalized Performance Normalized Performance 0.8 0.8 0.8 0.8 0.8 0.8 0.6 0.6 0.6 0.6 0.6 0.6 0.47 0.47 0.47 0.47 0.4 0.4 0.4 0.4 0.4 0.4 0.18 0.18 0.18 0.18 0.18 0.2 0.2 0.2 0.2 0.2 0.2 0.06 0.06 0.04 0.04 0.04 0 0 0 0 0 0 VoltDB VoltDB VoltDB VoltDB VoltDB VoltDB Memcached Memcached Memcached Memcached Memcached Memcached PowerGraph PowerGraph PowerGraph PowerGraph PowerGraph PowerGraph GraphX GraphX GraphX GraphX GraphX GraphX (TPC-C) (TPC-C) (TPC-C) (TPC-C) (TPC-C) (TPC-C) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (TunkRank) (TunkRank) (TunkRank) (TunkRank) (TunkRank) (TunkRank) (PageRank) (PageRank) (PageRank) (PageRank) (PageRank) (PageRank) 100% working sets in memory 100% working sets in memory 100% working sets in memory 100% working sets in memory 100% working sets in memory 100% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 3/30/17 9

  11. Performance degradation 0.97 1 1 1 1 1 1 1 0.94 0.94 Normalized Performance Normalized Performance Normalized Performance Normalized Performance Normalized Performance Normalized Performance Normalized Performance 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.47 0.47 0.47 0.47 0.47 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.18 0.18 0.18 0.18 0.18 0.18 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.06 0.06 0.06 0.04 0.04 0.04 0.04 0 0 0 0 0 0 0 VoltDB VoltDB VoltDB VoltDB VoltDB VoltDB VoltDB Memcached Memcached Memcached Memcached Memcached Memcached Memcached PowerGraph PowerGraph PowerGraph PowerGraph PowerGraph PowerGraph PowerGraph GraphX GraphX GraphX GraphX GraphX GraphX GraphX (TPC-C) (TPC-C) (TPC-C) (TPC-C) (TPC-C) (TPC-C) (TPC-C) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (TunkRank) (TunkRank) (TunkRank) (TunkRank) (TunkRank) (TunkRank) (TunkRank) (PageRank) (PageRank) (PageRank) (PageRank) (PageRank) (PageRank) (PageRank) 100% working sets in memory 100% working sets in memory 100% working sets in memory 100% working sets in memory 100% working sets in memory 100% working sets in memory 100% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 3/30/17 10

  12. Performance degradation 0.97 0.97 1 1 1 1 1 1 1 1 0.94 0.94 0.94 Normalized Performance Normalized Performance Normalized Performance Normalized Performance Normalized Performance Normalized Performance Normalized Performance Normalized Performance 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.47 0.47 0.47 0.47 0.47 0.47 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.12 0.06 0.06 0.06 0.06 0.04 0.04 0.04 0.04 0.04 0 0 0 0 0 0 0 0 VoltDB VoltDB VoltDB VoltDB VoltDB VoltDB VoltDB VoltDB Memcached Memcached Memcached Memcached Memcached Memcached Memcached Memcached PowerGraph PowerGraph PowerGraph PowerGraph PowerGraph PowerGraph PowerGraph PowerGraph GraphX GraphX GraphX GraphX GraphX GraphX GraphX GraphX (TPC-C) (TPC-C) (TPC-C) (TPC-C) (TPC-C) (TPC-C) (TPC-C) (TPC-C) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (Facebook/FB SYS) (TunkRank) (TunkRank) (TunkRank) (TunkRank) (TunkRank) (TunkRank) (TunkRank) (TunkRank) (PageRank) (PageRank) (PageRank) (PageRank) (PageRank) (PageRank) (PageRank) (PageRank) 100% working sets in memory 100% working sets in memory 100% working sets in memory 100% working sets in memory 100% working sets in memory 100% working sets in memory 100% working sets in memory 100% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 75% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 50% working sets in memory 3/30/17 11

Recommend

Disaggregation and the Application Sebastian Angel Mihir Nanavati Siddhartha Sen Traditional

Disaggregation and the Application Sebastian Angel Mihir Nanavati Siddhartha Sen Traditional

Disaggregation and the Application Sebastian Angel Mihir Nanavati Siddhartha Sen Traditional data center racks RDMA NIC GPUs CPUs Memory Storage Prior and current disaggregation efforts Towards DDCs OS kernel + Cache Why? Many

1.06k views • 21 slides
NVMe-over-Fabrics Performance Characterization and the Path to Low-Overhead Flash Disaggregation

NVMe-over-Fabrics Performance Characterization and the Path to Low-Overhead Flash Disaggregation

NVMe-over-Fabrics Performance Characterization and the Path to Low-Overhead Flash Disaggregation Zvika Guz, Harry Li, Anahita Shayesteh, and Vijay Balakrishnan Memory Solution Lab Samsung Semiconductor Inc. 1 Synopsis Performance

224 views • 21 slides
Efforts in Monitoring SDG with Disaggregation in the Philippines International Workshop on Data

Efforts in Monitoring SDG with Disaggregation in the Philippines International Workshop on Data

Efforts in Monitoring SDG with Disaggregation in the Philippines International Workshop on Data Disaggregation for the SDGs Bangkok, Thailand 28 30 January 2019 Bernadette B. Balamban Chief Statistical Specialist Philippine Statistics

722 views • 27 slides
Last class: Virtual Memory Today: Virtual Memory Uses Efficient Use of Physical

Last class: Virtual Memory Today: Virtual Memory Uses Efficient Use of Physical

Last class: Virtual Memory Today: Virtual Memory Uses Efficient Use of Physical Memory Through virtual memory N 2 32 -sized address spaces All isolated by default Uses for memory Make a new process Address

516 views • 22 slides
Disaggregation March 4, 2020 D. Pallme, A. Kosanovic, TN-DOT K. Pujats, M. Golias, S. Mishra,

Disaggregation March 4, 2020 D. Pallme, A. Kosanovic, TN-DOT K. Pujats, M. Golias, S. Mishra,

Methods and T ools for Freight Flow Disaggregation March 4, 2020 D. Pallme, A. Kosanovic, TN-DOT K. Pujats, M. Golias, S. Mishra, Univ. of Memphis Brief project overview Production/Attraction Methods Contents OD Disaggregation GIS Tools

421 views • 18 slides
Energy-Efficient In-Memory Data Stores on Hybrid Memory Hierarchies Eleventh International

Energy-Efficient In-Memory Data Stores on Hybrid Memory Hierarchies Eleventh International

Energy-Efficient In-Memory Data Stores on Hybrid Memory Hierarchies Eleventh International Workshop on Data Management on New Hardware June 2015 Ahmad Hassan (SAP), Hans Vandierendonck (QUB) and Dimitrios S. Nikolopoulos (QUB) May 2015

513 views • 30 slides
SILT A Memory-Efficient, High-Performance Key- Value Store Based on paper of H. Lim, B. Fan,

SILT A Memory-Efficient, High-Performance Key- Value Store Based on paper of H. Lim, B. Fan,

SILT A Memory-Efficient, High-Performance Key- Value Store Based on paper of H. Lim, B. Fan, D.G. Andersen, and M. Kaminsky: Presenter: SILT: A Memory-Efficient, High-Performance Key-Value Store Andrzej Findeisen in Proceedings ACM SOSP

590 views • 19 slides
Efficient Hardware-assisted Logging with Asynchronous and Direct Update for Persistent Memory

Efficient Hardware-assisted Logging with Asynchronous and Direct Update for Persistent Memory

Efficient Hardware-assisted Logging with Asynchronous and Direct Update for Persistent Memory Jungi Jeong , Chang Hyun Park, Jaehyuk Huh, and Seungryoul Maeng International Symposium on Microarchitecture (MICRO) 2018 Storage-Class Memory

1.95k views • 155 slides
HOOP: Efficient Hardware-Assisted Out-of-Place Update for Non-Volatile Memory Miao Cai Chance

HOOP: Efficient Hardware-Assisted Out-of-Place Update for Non-Volatile Memory Miao Cai Chance

HOOP: Efficient Hardware-Assisted Out-of-Place Update for Non-Volatile Memory Miao Cai Chance Coats Jian Huang Systems Platform Research Group Non-Volatile Memory is a Revolutionary Technology Close-to-DRAM Performance Data

367 views • 21 slides
CCHL: Compression-Consolidation Hardware Logging for Efficient Failure-Atomic Persistent Memory

CCHL: Compression-Consolidation Hardware Logging for Efficient Failure-Atomic Persistent Memory

CCHL: Compression-Consolidation Hardware Logging for Efficient Failure-Atomic Persistent Memory Updates Xueliang Wei, Dan Feng, Wei Tong, Jingning Liu, Chengning Wang, Liuqing Ye Huazhong University of Science and Technology Persistent Memory

817 views • 38 slides
Search Lookaside Buffer: Efficient Caching for Index Data Structures Xingbo Wu, Fan Ni, Song

Search Lookaside Buffer: Efficient Caching for Index Data Structures Xingbo Wu, Fan Ni, Song

Search Lookaside Buffer: Efficient Caching for Index Data Structures Xingbo Wu, Fan Ni, Song Jiang Background Large-scale in-memory applications. In-memory databases In-memory NoSQL stores and caches Software routing tables

731 views • 36 slides
The Basics What is memory management? Programs contain Objects Data Occupy memory

The Basics What is memory management? Programs contain Objects Data Occupy memory

The Basics What is memory management? Programs contain Objects Data Occupy memory Runtime system must allocate and reclaim memory for program in an efficient manner Why is this important? Why is this hard? Why is

319 views • 20 slides
Tolerating Faults in Disaggregated Datacenters Amanda Carbonari , Ivan Beschastnikh University

Tolerating Faults in Disaggregated Datacenters Amanda Carbonari , Ivan Beschastnikh University

Tolerating Faults in Disaggregated Datacenters Amanda Carbonari , Ivan Beschastnikh University of British Columbia HotNets17 Todays Datacenters 2 The future: Disaggregation 3 The future: Disaggregation The future: Disaggregation is

497 views • 25 slides
Efficient Memory Integrity Verification and Encryption for Secure Processors G. Edward Suh,

Efficient Memory Integrity Verification and Encryption for Secure Processors G. Edward Suh,

Efficient Memory Integrity Verification and Encryption for Secure Processors G. Edward Suh, Dwaine Clarke, Blaise Gassend, Marten van Dijk, Srinivas Devadas Massachusetts Institute of Technology New Security Challenges Current computer

411 views • 26 slides
Efficient Parallel Functional Programming with Hierarchical Memory Management Sam Westrick

Efficient Parallel Functional Programming with Hierarchical Memory Management Sam Westrick

Efficient Parallel Functional Programming with Hierarchical Memory Management Sam Westrick Carnegie Mellon University Joint work with: Ram Raghunathan, Adrien Guatto, Stefan Muller, Rohan Yadav, Umut Acar, Guy Blelloch, Matthew Fluet Setting

892 views • 42 slides
CoMerge Toward Efficient Data Placement in Shared Heterogeneous Memory Systems Thaleia Dimitra

CoMerge Toward Efficient Data Placement in Shared Heterogeneous Memory Systems Thaleia Dimitra

CoMerge Toward Efficient Data Placement in Shared Heterogeneous Memory Systems Thaleia Dimitra Doudali Ada Gavrilovska Motivation Performance slowdown in heterogeneous memory systems. Application data objects How to reduce the

343 views • 12 slides
An Efficient Memory-Mapped Key-Value Store for Flash Storage Anastasios Papagiannis, Giorgos

An Efficient Memory-Mapped Key-Value Store for Flash Storage Anastasios Papagiannis, Giorgos

An Efficient Memory-Mapped Key-Value Store for Flash Storage Anastasios Papagiannis, Giorgos Saloustros, Pilar Gonzlez-Frez, and Angelos Bilas Institute of Computer Science (ICS) Foundation for Research and Technology Hellas (FORTH)

2.85k views • 31 slides
Configurable and Efficient Memory Access Tracing via Selective Expression-based x86 Binary

Configurable and Efficient Memory Access Tracing via Selective Expression-based x86 Binary

Simone Economo, Davide Cingolani, Alessandro Pellegrini and Francesco Quaglia DIAG - Sapienza University of Rome Configurable and Efficient Memory Access Tracing via Selective Expression-based x86 Binary Instrumentation

729 views • 46 slides
Communication and Memory Efficient Testing of Discrete Distributions Themis Gouleakis USC

Communication and Memory Efficient Testing of Discrete Distributions Themis Gouleakis USC

Communication and Memory Efficient Testing of Discrete Distributions Themis Gouleakis USC MPI July 21, 2019 Joint work with: Ilias Diakonikolas (USC), Daniel Kane (UCSD) and Sankeerth Rao (UCSD) M OTIVATION Datasets growing too

489 views • 15 slides
Efficient and Reliable Lock-Free Memory Introduction The Problem Reclamation

Efficient and Reliable Lock-Free Memory Introduction The Problem Reclamation

Outline Efficient and Reliable Lock-Free Memory Introduction The Problem Reclamation Lock-free synchronization Based on Reference Counting Our solution Anders Gidenstam, Marina Papatriantafilou, Idea Hkan Sundell and

81 views • 6 slides
Energy Efficient Content-Addressable Memory Advanced Seminar Computer Engineering Institute of

Energy Efficient Content-Addressable Memory Advanced Seminar Computer Engineering Institute of

Energy Efficient Content-Addressable Memory Advanced Seminar Computer Engineering Institute of Computer Engineering Heidelberg University Fabian Finkeldey 26.01.2016 Fabian Finkeldey, Energy Efficient Content-Addressable Memory 1 Table of

498 views • 33 slides
Data Generation and Disaggregation for Monitoring and Evaluation of SDGs Md. Alamgir Hossen

Data Generation and Disaggregation for Monitoring and Evaluation of SDGs Md. Alamgir Hossen

Action Plan and Methodological Guidelines for Data Generation and Disaggregation for Monitoring and Evaluation of SDGs Md. Alamgir Hossen Deputy Director and Member-Secretary, SDG Cell Bangladesh Bureau of Statistics (BBS) Background Under

284 views • 12 slides
Memory II. Memory improvement III. Problems with memory 3 systems/stages of Memory: memory

Memory II. Memory improvement III. Problems with memory 3 systems/stages of Memory: memory

Main Focus I. Memory as a process Memory II. Memory improvement III. Problems with memory 3 systems/stages of Memory: memory the process by which I. Sensory Memory information is - acquired, II. Short -Term Memory - stored,

169 views • 5 slides
Handling Imperfections in Energy Disaggregation Authors: Angshul Majumdar, Megha

Handling Imperfections in Energy Disaggregation Authors: Angshul Majumdar, Megha

Handling Imperfections in Energy Disaggregation Authors: Angshul Majumdar, Megha Gupta Presenter: Manoj Gula7 Spikes or Surges Short-term increases in the electrical supply

473 views • 20 slides

More recommend