標籤：

OSDI 2016有哪些讓人眼前一亮的文章？

02-01

#OSDI 16 扒一扒 不完整

學校 篇數

UWashington 3
Google 1
Columbia,brown 1
Cambridge 1
IC,TUD 1
Toronto 2
USC 1
MIT &>1
UC Berkeley &>1
Michigan 2
上海交大 1
Wisconsin-Madison 3
utexas 3
Edinburgh 1
UCI 1

  1. UWashington:

  • Push-Button Verification of File Systems via Crash Refinement

    (Helgi Sigurbjarnarson, James Bornholt, Emina Torlak, Xi Wang)

  • Fast Replication with NOPaxos: Replacing Consensus with Network Ordering

    (Jialin Li, Ellis Michael, Naveen Kr. Sharma, Adriana Szekeres, Dan R. K. Ports)

  • Diamond: Automating Data Management and Storage for Wide-area, Reactive Applications (Irene Zhang, Niel Lebeck, Pedro Fonseca, Brandon Holt, Raymond Cheng, Ariadna Norberg, Arvind Krishnamurthy, Henry M. Levy)

2. Google: TensorFlow

TensorFlow: A system for large-scale machine learning

Martín Abadi, Paul Barham, Jianmin Chen, Zhifeng Chen, Andy Davis, Jeffrey Dean, Matthieu Devin, Sanjay Ghemawat, Geoffrey Irving, Michael Isard, Manjunath Kudlur, Josh Levenberg, Rajat Monga, Sherry Moore, Derek G. Murray, Benoit Steiner, Paul Tucker,Vijay Vasudevan, Pete Warden, Martin Wicke, Yuan Yu, Xiaoqiang Zheng

http://arxiv.org/abs/1605.08695

3. Brown and Columbia University

Shuffler

Shuffler: Fast and Deployable Continuous Code Re-randomization

David Williams-King, Graham Gobieski, Kent Williams-King, James P. Blake, Xinhao Yuan, Patrick Colp, Vasileios P. Kemerlis, Junfeng Yang, Bill Aiello

To appear in OSDI 16

4. Cambridge

scaling Firmaments scheduling logic to very large clusters

CamSaS – Firmament

5. IC 帝國理工， TUD, TUB 一篇 講docker安全應該很有意思

SCONE: Secure Linux Containers with Intel SGX

Sergei Arnautov, Bohdan Trach, Franz Gregor, Thomas Knauth, Andre Martin, Christian Priebe, Joshua Lind, Divya Muthukumaran, Daniel OKeeffe, Mark L. Stillwell, David Goltzsche, Dave Eyers, Rüdiger Kapitza, Peter Pietzuch, and Christof Fetzer

6. Toronto: 2篇 ding yuan組 zhou yuanyuan 學生

  • Non-intrusive Performance Profiling of Entire Software Stacks based on the Flow Reconstruction Principle

    Xu Zhao, Kirk Rodrigues, Yu Luo, Ding Yuan, and Michael Stumm. Proceedings of the 12th USENIX Symposium on Operating Systems Design and Implementation (OSDI』16), November 2016, Savannah, GA.
  • Dont Get Caught In the Cold, Warm-up Your JVM: Understand and Eliminate JVM Warm-up Overhead In Data-parallel Systems

    David Lion, Adrian Chiu, Hailong Sun, Xin Zhuang, Nikola Grcevski, and Ding Yuan. Proceedings of the 12th USENIX Symposium on Operating Systems Design and Implementation (OSDI』16), November 2016, Savannah, GA.

7. USC 2篇

Networked Systems Laboratory Main/Home

內容不詳

8. MIT 1篇至少

Alpenhorn: Bootstrapping Secure Communication Without Leaking Metadata.

David Lazar and Nickolai Zeldovich.

In Proceedings of the 12th Symposium on Operating Systems Design and Implementation (OSDI), Savannah, GA, November 2016.

9. UC berkeley 1篇至少

  • NetBricks: Taking the V out of NFV. Aurojit Panda, Sangjin Han, Keon Jang, Melvin Walls, Sylvia Ratnasamy, Scott Shenker. To appear OSDI 2016.

10 Michigan, Ann Arbor 2篇

  • Altruistic Scheduling in Multi-Resource Clusters

    R. Grandl, M. Chowdhury, A. Akella, G. Ananthanarayanan

    USENIX OSDI, 2016

    Acceptance Rate: 18.08%
  • EC-Cache: Load-Balanced, Low-Latency Cluster Caching with Online Erasure Coding

    K. V. Rashmi, M. Chowdhury, J. Kosaian, I. Stoica, K. Ramchandran

    USENIX OSDI, 2016

    Acceptance Rate: 18.08%

11. 上海交大 haibo chen 1篇

[OSDI] Fast and Concurrent RDF Queries with RDMA-based Distributed Graph Exploration. Jiaxin Shi, Youyang Yao, Rong Chen, Haibo Chen and Feifei Li. 2016 Usenix Symposium on Operating System Design and Implementation.

12. Uchicago 2篇 Shan Lu 組 zhou yuanyuan 學生

  1. Early Detection of Configuration Errors to Reduce Failure Damage [OSDI16]
  2. Yak: A High-Performance Big-Data-Friendly Garbage Collector [OSDI16]

13. University of Wisconsin-Madison 3篇

  • Clarinet: WAN-Aware Optimization for Analytics Queries [pdf]

    Raajay Viswanathan, Aditya Akella and Ganesh Ananthanarayanan.

    OSDI 2016, Savannah, GA.

  • Altruistic Scheduling in Multi-Resource Clusters [pdf]

    Robert Grandl, Mosharaf Chowdhury, Aditya Akella and Ganesh Ananthanarayanan.

    OSDI 2016, Savannah, GA.

  • Packing and Dependency-aware Scheduling for Data-Parallel Clusters [pdf]

    Robert Grandl, Srikanth Kandula, Sriram Rao, Aditya Akella and Janardhan Kulkarni.

    OSDI 2016, Savannah, GA.

    • 14. utexas： 3篇 Laboratory for Advanced Systems Research

    Coordinated and Efficient Huge Page Management with IngensYoungjin Kwon, Hangchen Yu, Simon Peter, Chris Rossbach, and Emmett Witchel. In OSDI 2016.

    Unobservable communication over fully untrusted infrastructureSebastian Angel and Srinath Setty. In OSDI 2016.

    Ryoan: A Distributed Sandbox for Untrusted Computation on Secret DataTyler Hunt, Zhiting Zhu, Yuanzhong Xu, Simon Peter, and Emmett Witchel. In OSDI 2016.

    15. 愛丁堡1篇 ：

    1. Praveen Tammana, Rachit Agarwal, and Myungjin Lee, "Simplifying Datacenter Network Debugging with PathDump", USENIX OSDI, Nov 2016 [pdf]

    16. UCI 1篇

    Khanh Nguyen, Lu Fang, Guoqing Xu, Brian Demsky, Shan Lu, Sanazsadat Alamian, and Onur Mutlu,

    "Yak: A High-Performance Big-Data-Friendly Garbage Collector",

    12th USENIX Symposium on Operating Systems Design and Implementation, (Acceptance rate: 47/267, 17.6%)

    還差很多篇的樣子：.....

    IPADS實驗室的參會學生們正在ChinaSys的微信公眾號CNSys上連載OSDI2016的見聞，包括了每篇論文從學生角度的簡要介紹

    —USENIX OSDI 2016 Technical Session

    Technical Sessions

    47篇 （program有一點小錯），國內有清華2篇，上交1篇，以及國防科大1篇

    Graph Processing and Machine Learning Session

    第1篇是 TensorFlow，後面3篇正好都是國內的 ：）

    就我個人來說最有趣的是這一篇：「Just Say NO to Paxos Overhead: Replacing Consensus Overhead with Network Ordering」。

    簡單的說，這篇文章把 Paxos 要實現的功能拆成了保證消息的順序（Order）以及原子性的廣播（Atomic Broadcast）兩部分。

    然後，通過把 Order 部分轉交給網路層（比如說用 SDN）去實現達到了比普通的 Paxos 快 4.7x 的結果（號稱只比完全不保證可靠性的直接 broadcast 慢 2%）。

    雖然其實我也還沒完全看明白，不過核心的技巧應該是用一個 centralized sequencer （文中描述了包括 SDN 和普通機器的兩種實現）來實現 order 的功能（基本上就是每個 package 打上一個序號）。

    我覺得這是一個非常好的「實際環境下有用的系統」的例子。

    理論界的人在研究 Consensus 的時候一般都假設完全對稱，沒有任何中心化節點。

    但是，在很多的實際的使用情況下經常會存在「根路由器」這麼一個中心節點，那麼就沒有道理不好好利用一下。

    正式的Paper List還沒出來呢。

    我們系今年中的：

    • Push-Button Verification of File Systems via Crash Refinement

      (Helgi Sigurbjarnarson, James Bornholt, Emina Torlak, Xi Wang)

    • Fast Replication with NOPaxos: Replacing Consensus with Network Ordering

      (Jialin Li, Ellis Michael, Naveen Kr. Sharma, Adriana Szekeres, Dan R. K. Ports)

    • Diamond: Automating Data Management and Storage for Wide-area, Reactive Applications

      (Irene Zhang, Niel Lebeck, Pedro Fonseca, Brandon Holt, Raymond Cheng, Ariadna Norberg, Arvind Krishnamurthy, Henry M. Levy)

    ————————————————————————————————

    等出了正式列表更新。

    等八月份Program出來吧，Google關於TensorFlow的paper會在（預印本：http://arxiv.org/abs/1605.08695）。

    我們組今年中的：

    Non-intrusive Performance Profiling of Entire Software Stacks based on the Flow Reconstruction Principle

    以及

    Dont Get Caught In the Cold, Warm-up Your JVM: Understand and Eliminate JVM Warm-up Overhead In Data-parallel Systems

    由於現在還在準備camera-ready版本，所以draft暫時不會對外公開。

    ==

    Update: final version submitted.

    我們投的兩篇都掛了

    UMich + MSR:

  • History-Based Harvesting of Spare Cycles and Storage in Large-Scale Datacenters

    Yunqi Zhang, George Prekas, Giovanni M. Fumarola, Marcus Fontoura, Inigo Goiri, Ricardo Bianchini


    The 12th USENIX Symposium on Operating Systems Design and Implementation (OSDI 2016)


    (to appear)

    • Altruistic Scheduling in Multi-Resource Clusters
    • Graphene: Packing and Dependency-Aware Scheduling for Data-Parallel Clusters
    • Firmament: Fast, Centralized Cluster Scheduling at Scale
    • Morpheus: Towards Automated SLOs for Enterprise Clusters?

    從一個外行角度來看沒啥有意思的文章。我是沖著INFLOW去的。當然那篇certified OS是很厲害。

    OSDI的camera ready版本一般什麼時候能出爐？

    現在還是camera ready......要等公布了才知道吧

    （2我們組有一篇好像是中了，隔壁組可能也是有一篇

    有沒有人可以告訴我tensorflow作為一個深度學習框架，跟Operating Systems Design and Implementation 有什麼關係？tensorflow算是Operating System嗎？

    推薦閱讀：

    TAG:操作系統 | 分散式存儲 | 分散式系統 | 計算機體系架構 |