Datacenter Wide-areaEnterprise LOAD-BALANCER Client Servers.

Slides:



Advertisements
Similar presentations
Performance Analysis of Peer-to-Peer File Transfer Network Sayantan Mitra Vibhor Goyal 1.
Advertisements

Distributed NFV Optimization
Building Cloud-ready Video Transcoding System for Content Delivery Networks(CDNs) Zhenyun Zhuang and Chun Guo Speaker: 饒展榕.
Nanxi Kang Princeton University
Designing a New Routing Simulator for DiffServ MPLS Networks Peng Zhang Zhansong Ma Raimo Kantola {pgzhang, zhansong,
Yu Stephanie Sun 1, Lei Xie 1, Qi Alfred Chen 2, Sanglu Lu 1, Daoxu Chen 1 1 State Key Laboratory for Novel Software Technology, Nanjing University, China.
OpenFlow-Based Server Load Balancing GoneWild
On Selfish Routing In Internet-like Environments Lili Qiu (Microsoft Research) Yang Richard Yang (Yale University) Yin Zhang (AT&T Labs – Research) Scott.
What’s the Problem Web Server 1 Web Server N Web system played an essential role in Proving and Retrieve information. Cause Overloaded Status and Longer.
Towards Virtual Routers as a Service 6th GI/ITG KuVS Workshop on “Future Internet” November 22, 2010 Hannover Zdravko Bozakov.
Copyright 2007, Information Builders. Slide 1 Workload Distribution for the Enterprise Mark Nesson, Vashti Ragoonath June, 2008.
Mohamed Hefeeda 1 School of Computing Science Simon Fraser University, Canada ISP-Friendly Peer Matching without ISP Collaboration Mohamed Hefeeda (Joint.
1 A Vehicle Route Management Solution Enabled by Wireless Vehicular Networks Kevin Collins and Gabriel-Miro Muntean IEEE INFOCOM 2008.
Volcano Routing Scheme Routing in a Highly Dynamic Environment Yashar Ganjali Stanford University Joint work with: Nick McKeown SECON 2005, Santa Clara,
©NEC Laboratories America 1 Hui Zhang Samrat Ganguly Sudeept Bhatnagar Rauf Izmailov NEC Labs America Abhishek Sharma University of Southern California.
Beneficial Caching in Mobile Ad Hoc Networks Bin Tang, Samir Das, Himanshu Gupta Computer Science Department Stony Brook University.
Peer-to-Peer Based Multimedia Distribution Service Zhe Xiang, Qian Zhang, Wenwu Zhu, Zhensheng Zhang IEEE Transactions on Multimedia, Vol. 6, No. 2, April.
Detecting Network Intrusions via Sampling : A Game Theoretic Approach Presented By: Matt Vidal Murali Kodialam T.V. Lakshman July 22, 2003 Bell Labs, Lucent.
Yashar Ganjali, and Abtin Keshavarzian Presented by: Isaac Keslassy Computer Systems Laboratory Department of Electrical Engineering Stanford University.
December 20, 2004MPLS: TE and Restoration1 MPLS: Traffic Engineering and Restoration Routing Zartash Afzal Uzmi Computer Science and Engineering Lahore.
ICNP'061 Benefit-based Data Caching in Ad Hoc Networks Bin Tang, Himanshu Gupta and Samir Das Computer Science Department Stony Brook University.
Decentralized resource management for a distributed continuous media server Cyrus Shahabi and Farnoush Banaei-Kashani IEEE Transactions on Parallel and.
Hardware-based Load Generation for Testing Servers Lorenzo Orecchia Madhur Tulsiani CS 252 Spring 2006 Final Project Presentation May 1, 2006.
ICNP'061 Benefit-based Data Caching in Ad Hoc Networks Bin Tang, Himanshu Gupta and Samir Das Department of Computer Science Stony Brook University.
Yashar Ganjali Joint work with: Abtin Keshavarzian June 4, 2003 Single-Path vs. Multi-Path Routing in Ad Hoc Networks.
Hardness of Approximation and Greedy Algorithms for the Adaptation Problem in Virtual Environments Ananth I. Sundararaj, Manan Sanghi, John R. Lange and.
Rethinking Internet Traffic Management: From Multiple Decompositions to a Practical Protocol Jiayue He Princeton University Joint work with Martin Suchara,
Adaptive Content Delivery for Scalable Web Servers Authors: Rahul Pradhan and Mark Claypool Presented by: David Finkel Computer Science Department Worcester.
Multipath Protocol for Delay-Sensitive Traffic Jennifer Rexford Princeton University Joint work with Umar Javed, Martin Suchara, and Jiayue He
Tradeoffs in CDN Designs for Throughput Oriented Traffic Minlan Yu University of Southern California 1 Joint work with Wenjie Jiang, Haoyuan Li, and Ion.
Datacenter Wide-areaEnterprise LOAD-BALANCER Client Servers.
1 Latency Equalization: A Programmable Routing Service Primitive Minlan Yu Joint work with Marina Thottan, Li Li at Bell Labs.
Distributing Content Simplifies ISP Traffic Engineering Abhigyan Sharma* Arun Venkataramani* Ramesh Sitaraman*~ *University of Massachusetts Amherst ~Akamai.
LABERIO: Dynamic load- balanced routing in OpenFlow- enabled networks Author:Hui Long, Yao Shen*, Minyi Guo, Feilong Tang IEEE 27th International.
Cloud MapReduce : a MapReduce Implementation on top of a Cloud Operating System Speaker : 童耀民 MA1G Authors: Huan Liu, Dan Orban Accenture.
Monitoring Latency Sensitive Enterprise Applications on the Cloud Shankar Narayanan Ashiwan Sivakumar.
Mobile Agents For Personalized Information Retrieval: When are they a good idea? Telcordia Technologies Proprietary – Internal Use Only This document contains.
An Efficient Approach for Content Delivery in Overlay Networks Mohammad Malli Chadi Barakat, Walid Dabbous Planete Project To appear in proceedings of.
Performance Evaluation of ATM Shortcuts in Overlaid IP/ATM Networks Jim Kurose Don Towsley Department of Computer Science Univ. of Massachusetts, Amherst.
Scalable Web Server on Heterogeneous Cluster CHEN Ge.
Sensitivity of Cluster File System Access to I/O Server Selection A. Apon, P. Wolinski, and G. Amerson University of Arkansas.
CSCI 465 D ata Communications and Networks Lecture 15 Martin van Bommel CSCI 465 Data Communications & Networks 1.
University of Padova Department of Information Engineering On the Optimal Topology of Bluetooth Piconets: Roles Swapping Algorithms Med-Hoc-Net 2002, Chia.
The Performance Evaluation of Intra-domain Bandwidth Allocation and Inter-domain Routing Algorithms for a QoS-guaranteed Routing Path Discovery Bo Li,
Intradomain Traffic Engineering By Behzad Akbari These slides are based in part upon slides of J. Rexford (Princeton university)
6 December On Selfish Routing in Internet-like Environments paper by Lili Qiu, Yang Richard Yang, Yin Zhang, Scott Shenker presentation by Ed Spitznagel.
1 Part VII Component-level Performance Models for the Web © 1998 Menascé & Almeida. All Rights Reserved.
정하경 MMLAB Fundamentals of Internet Measurement: a Tutorial Nevil Brownlee, Chris Lossley, “Fundamentals of Internet Measurement: a Tutorial,” CMG journal.
1 An Arc-Path Model for OSPF Weight Setting Problem Dr.Jeffery Kennington Anusha Madhavan.
1 Simple provisioning, complex consolidation – An approach to improve the efficiency of provisioning oriented optical networks Tamás Kárász Budapest University.
CS 6401 Overlay Networks Outline Overlay networks overview Routing overlays Resilient Overlay Networks Content Distribution Networks.
Content Delivery Networks: Status and Trends Speaker: Shao-Fen Chou Advisor: Dr. Ho-Ting Wu 5/8/
DECOR: A Distributed Coordinated Resource Monitoring System Shan-Hsiang Shen Aditya Akella.
On the Placement of Web Server Replicas Yu Cai. Paper On the Placement of Web Server Replicas Lili Qiu, Venkata N. Padmanabhan, Geoffrey M. Voelker Infocom.
R2C2: A Network Stack for Rack-scale Computers Paolo Costa, Hitesh Ballani, Kaveh Razavi, Ian Kash Microsoft Research Cambridge EECS 582 – W161.
Datacenter Wide-areaEnterprise LOAD-BALANCER Client Servers.
Chen Qian, Xin Li University of Kentucky
Link-Level Internet Structures
University of Maryland College Park
Content Distribution Networks
ECE 544: Traffic engineering (supplement)
Principles of Network Applications
Mohammad Malli Chadi Barakat, Walid Dabbous Alcatel meeting
Distributed Systems CS
Capacity Analysis, cont. Realistic Server Performance
Distributed Channel Assignment in Multi-Radio Mesh Networks
Distributed Systems CS
EE 122: Lecture 22 (Overlay Networks)
pathChirp Efficient Available Bandwidth Estimation
Authors: Jinliang Fan and Mostafa H. Ammar
Presentation transcript:

Datacenter Wide-areaEnterprise

LOAD-BALANCER Client Servers

Outline and goals  A new architecture for distributed load-balancing  joint (server, path) selection  Demonstrate a nation-wide prototype  Interesting preliminary results

I’m here to ask for your help!

Data Path (Hardware) Control Path OpenFlow OpenFlow Controller OpenFlow Protocol (SSL) Control Path

Custom Hardware OS Network OS Feature Software Defined Networking Feature 7

Load Balancing is just Smart Routing

Custom Hardware Network OS Load-balancing logic Load-balancing as a network primitive Load-balancing decision 9

So far…  A new architecture for distributed load-balancing  joint (server, path) selection  Aster*x – a nation-wide prototype  Promising results that joint (server, path) selection might have great benefits

What next?

How big is the pie? Characterizing and quantifying the performance of joint (server, path) selection

MININET-RT

Clients CDN ISP Model

Parameters Topology  Intra-AS topologies  BRITE (2000 topologies)  CAIDA (1000 topologies)  Rocketfuel (~100 topos.)  nodes  Uniform link capacity

Parameters Servers  5-10 servers  Random placement Service  Simple HTTP service  Serving 1 MB file  Additional server-side computation

Parameters Clients  3-5 client locations  Random placement Request pattern  Poisson process  Mean rate: 5-10 req/sec

Load-balancing strategies?

Simple but suboptimal Complex but optimal Design space Disjoint-Shortest-Path Joint Disjoint-Traffic-Engineering

Anatomy of a request- response ClientLoad-BalancerServer Response Time Deliver Retrieve Choose Request Response 1 st byte Response last byte Last byte ack

Disjoint-Shortest-Path  CDN selects the least loaded server  Load = retrieve + deliver  ISP independently selects the shortest path

Disjoint-Traffic-Engineering  CDN selects the least loaded server  Load = retrieve + deliver  ISP independently selects path to minimize max load  Max bandwidth headroom

Joint  Single controller jointly selects the best (server, path) pair Total latency = retrieve + estimated deliver

Disjoint-Shortest-Path vs Joint Disjoint-Shortest-Path performs ~2x worse than Joint

Disjoint-Traffic-Engg. vs Joint Disjoint-Traffic-Engineering performs almost as well as Joint

Is Disjoint truly disjoint? ClientLoad-BalancerServer Response Time Deliver Retrieve Choose Request Response 1 st byte Response last byte Last byte ack Server response time contains network information

The bottleneck effect A single bottleneck resource along the path determines the performance.

Clients CDN ISP The CDN-ISP game

 System load monotonically decreases  Both push system in the same direction

Summary of observations  Disjoint-SP is ~2x worse than Joint  Disjoint-TE performs almost as well as Joint (despite decoupling of server selection and traffic engineering)  Game theoretic analysis supports the empirical observation

Questions for you!  How should I change the model to mimic a real CDN?

Questions for you!  How can I get real data?  What network topologies should I use?  How should I decide the no. of servers and their location?  How should I decide the client request pattern?

Questions for you!  How can I try it out in your network?  Elastic Load Balancing in EC2  Amazon CloudFront

Conclusion  A new architecture for distributed load-balancing  joint (server, path) selection  Aster*x - a nation-wide prototype  Interesting preliminary results  Future – Evaluation with real data

Let’s chat more!

Extra slides…

Sample topologies BRITECAIDA

Disjoint-SP vs Ideal

Disjoint-TE vs Ideal