Sharing Social Content from Home: A Measurement-driven Feasibility Study Massimiliano Marcon Bimal Viswanath Meeyoung Cha Krishna Gummadi NOSSDAV 2011.

Slides:



Advertisements
Similar presentations
Capacity Planning for LAMP Architectures John Allspaw Manager, Operations Flickr.com Web Builder 2.0 Las Vegas.
Advertisements

Multi-Access Services in Heterogeneous Wireless Networks Kameswari Chebrolu, Ramesh R. Rao Abstract Today's wireless world is characterized by heterogeneity.
1 Sizing the Streaming Media Cluster Solution for a Given Workload Lucy Cherkasova and Wenting Tang HPLabs.
Facebook: 747 class page John Lachmeyer Richard Tejada.
A Transport Protocol for Content-Centric Networking with Explicit Congestion Control Feixiong Zhang, Yanyong Zhang (Rutgers Univ.), Alex Reznik (InterDigital),
OPNET Technologies, Inc. Performance versus Cost in a Cloud Computing Environment Yiping Ding OPNET Technologies, Inc. © 2009 OPNET Technologies, Inc.
The Next I.T. Tsunami Paul A. Strassmann. Copyright © 2005, Paul A. Strassmann - IP4IT - 11/15/05 2 Perspective Months  Weeks.
On the Effectiveness of Measurement Reuse for Performance-Based Detouring David Choffnes Fabian Bustamante Fabian Bustamante Northwestern University INFOCOM.
Enabling the Social Web Krishna P. Gummadi Networked Systems Group Max Planck Institute for Software Systems.
Cloud Download : Using Cloud Utilities to Achieve High-quality Content Distribution for Unpopular Videos Yan Huang, Tencent Research, Shanghai, China Zhenhua.
Performance analysis and Capacity planning of Home LAN Mobile Networks Lab 4
Fresh Analysis of Streaming Media Stored on the Web Rabin Karki M.S. Thesis Presentation Advisor: Mark Claypool Reader: Emmanuel Agu 10 Jan, 2011.
Performance Analysis of Orb Rabin Karki and Thangam V. Seenivasan 1.
Named Data Networking for Social Network Content delivery P. Truong, B. Mathieu (Orange Labs), K. Satzke (Alu) E. Stephan (Orange Labs) draft-truong-icnrg-ndn-osn-00.txt.
An Analysis of Internet Content Delivery Systems Stefan Saroiu, Krishna P. Gommadi, Richard J. Dunn, Steven D. Gribble, and Henry M. Levy Proceedings of.
1 A Framework for Lazy Replication in P2P VoD Bin Cheng 1, Lex Stein 2, Hai Jin 1, Zheng Zhang 2 1 Huazhong University of Science & Technology (HUST) 2.
1 Simultaneous Distribution Control and Privacy Protection for Proxy based Media Distribution George Mason University Songqing Chen (George Mason University)
Network Traffic Measurement and Modeling CSCI 780, Fall 2005.
Peer-to-peer Multimedia Streaming and Caching Service by Won J. Jeon and Klara Nahrstedt University of Illinois at Urbana-Champaign, Urbana, USA.
Application Layer  We will learn about protocols by examining popular application-level protocols  HTTP  FTP  SMTP / POP3 / IMAP  Focus on client-server.
Social Media: YouTube as a Case. 2 New generation of video sharing service Feb.15th, 2005 Some statistics: 60 hours video uploaded very minute 4 billion.
Can Internet Video-on-Demand Be Profitable? SIGCOMM 2007 Cheng Huang (Microsoft Research), Jin Li (Microsoft Research), Keith W. Ross (Polytechnic University)
1 Proxy-Assisted Techniques for Delivering Continuous Multimedia Streams Lixin Gao, Zhi-Li Zhang, and Don Towsley.
Online Presence for SAIPs What’s Online Presence?
EE616 Technical Project Video Hosting Architecture By Phillip Sutton.
Can Internet VoD be Profitable? Cheng Huang (MSR), Jin Li (MSR), Keith W. Ross (NY Polytechnique)
Providing Controlled Quality Assurance in Video Streaming across the Internet Yingfei Dong, Zhi-Li Zhang and Rohit Rakesh Computer Networking and Multimedia.
GIS and Cloud Computing. Flickr  Upload and manage your photos online  Share your photos with your family and friends  Post your photos everywhere.
Chapter 4. After completion of this chapter, you should be able to: Explain “what is the Internet? And how we connect to the Internet using an ISP. Explain.
OSN Research As If Sociology Mattered Krishna P. Gummadi Networked Systems Research Group MPI-SWS.
Understanding the External Links of Video Sharing Sites: Measurement and Analysis.
Geographic Information Systems Cloud GIS. ► The use of computing resources (hardware and software) that are delivered as a service over the Internet ►
DELAYED CHAINING: A PRACTICAL P2P SOLUTION FOR VIDEO-ON-DEMAND Speaker : 童耀民 MA1G Authors: Paris, J.-F.Paris, J.-F. ; Amer, A. Computer.
Keeping on Top of Technological Trends and Uses of Existing Technology Daniel L. Appelman Heller Ehrman LLP.
School of EECS, Peking University Microsoft Research Asia UStore: A Low Cost Cold and Archival Data Storage System for Data Centers Quanlu Zhang †, Yafei.
Distributing Layered Encoded Video through Caches Authors: Jussi Kangasharju Felix HartantoMartin Reisslein Keith W. Ross Proceedings of IEEE Infocom 2001,
2: Application Layer1 Chapter 2 outline r 2.1 Principles of app layer protocols r 2.2 Web and HTTP r 2.3 FTP r 2.4 Electronic Mail r 2.5 DNS r 2.6 Socket.
ITU Workshop on "Future Trust and Knowledge Infrastructure", Phase 1 Geneva, Switzerland, 24 April 2015 Tussles for Edge Network Caching Patrick Poullie.
A Measurement Based Memory Performance Evaluation of High Throughput Servers Garba Isa Yau Department of Computer Engineering King Fahd University of Petroleum.
DIGITAL WORLDWIDE Ashish. s  107 trillion – The number of s sent on the Internet in  294 billion – Average number of messages.
Chris Reed Professor of Electronic Commerce Law. The perceived problem “If I put my information in the Cloud then I lose all my rights to it” But is this.
Multicast instant channel change in IPTV systems 1.
ALMA Archive Operations Impact on the ARC Facilities.
CONTENT DELIVERY NETWORKS
ACN Product Overview Broadband Australia Information correct as at October 1, 2015.
A P2P-Based Architecture for Secure Software Delivery Using Volunteer Assistance Purvi Shah, Jehan-François Pâris, Jeffrey Morgan and John Schettino IEEE.
Can Internet Video-on-Demand Be Profitable? Cheng Huang, Jin Li (Microsoft Research), Keith W. Ross (Polytechnic University) ACM SIGCOMM 2007.
Presenting By CH . MADHURI(12QU1D5806) Under the supervision of
1 Push-to-Peer Video-on-Demand System. 2 Abstract Content is proactively push to peers, and persistently stored before the actual peer-to-peer transfers.
Concepts of Video and File/Sharing System Reporters: Ma. Raizza M. Cantara Mary Jane Eule Richard Ravalo Maika Laguartilla.
Streaming and Content Delivery SECTIONS 7.4 AND 7.5.
A P2P On-Demand Video Streaming System with Multiple Description Coding Yanming Shen, Xiaofeng Xu, Shivendra Panwar, Keith Ross, Yao Wang Polytechnic University.
August 23, 2001ITCom2001 Proxy Caching Mechanisms with Video Quality Adjustment Masahiro Sasabe Graduate School of Engineering Science Osaka University.
ACN Product Overview nbn™ Australia Information correct as at October 1, 2015.
A Measurement Based Memory Performance Evaluation of Streaming Media Servers Garba Isa Yau and Abdul Waheed Department of Computer Engineering King Fahd.
MiddleMan: A Video Caching Proxy Server NOSSDAV 2000 Brian Smith Department of Computer Science Cornell University Ithaca, NY Soam Acharya Inktomi Corporation.
One Library’s Successful Venture in Providing Comprehensive Streaming Media Services Charleston Conference 2015 Saturday, November 7 10:45am - 11:15am.
A Latent Social Approach to YouTube Popularity Prediction Amandianeze Nwana Prof. Salman Avestimehr Prof. Tsuhan Chen.
Scalable Data Scale #2 site on the Internet (time on site) >200 billion monthly page views Over 1 million developers in 180 countries.
1 Internet Traffic Measurement and Modeling Carey Williamson Department of Computer Science University of Calgary.
ACN Product Overview Broadband Australia Information correct as at 17 February 2016.
Mobile Data Offloading: How Much Can WiFi Deliver? Kyunghan Lee, Injong Rhee, Joohyun Lee, Song Chong, Yung Yi CoNEXT Presentor: Seokshin.
1 Netflow Collection and Aggregation in the AT&T Common Backbone Carsten Lund.
Online Data Storage Companies MY Docs Online. Comparison Name Personal Edition Enterprise Edition Transcription Edition Price $9.95 monthly rate $4.99.
Does Internet media traffic really follow the Zipf-like distribution? Lei Guo 1, Enhua Tan 1, Songqing Chen 2, Zhen Xiao 3, and Xiaodong Zhang 1 1 Ohio.
Content Distribution Networks
ECE 671 – Lecture 16 Content Distribution Networks
Steve Ko Computer Sciences and Engineering University at Buffalo
Tussles for Edge Network Caching
AWS Cloud Computing Masaki.
Presentation transcript:

Sharing Social Content from Home: A Measurement-driven Feasibility Study Massimiliano Marcon Bimal Viswanath Meeyoung Cha Krishna Gummadi NOSSDAV /4/20151 MPI-SWS KAIST

Growth of social content Social content includes personal photos, videos, status updates – FB: 15 billion photos, 220 million weekly uploads Huge growth due to rising popularity of social networks Social content is often personal, users want control over: – What they share: – Whom they share with (access control, privacy) – How the content is being used (e.g. advertising) 11/4/20152

How is social content shared today? User generates content Content is uploaded to datacenter (e.g. Facebook) Datacenter delivers content via traditional Web – Web servers, CDNs 311/4/2015

Good side of current content sharing Performance and availability – OSNs use well-provisioned servers – Content accessible 24/7 from everywhere OSNs are storing content for users 411/4/2015

Bad side of current content sharing Restrictions on what / how much can be shared – Content type and quality Loss of ownership / copyrights – Terms of service of OSNs are complex – They may include broad rights on content, typically: Loss of privacy – OSNs privacy policies are complex, may change at any time – If malicious, OSNs can infringe on users’ privacy 511/4/2015 “worldwide, non-exclusive, royalty-free, sublicenseable and transferable license to use, reproduce, distribute, prepare derivative works of, display, and perform the Content” (YouTube, Facebook, Flickr) Users lose control over their data

Idea: What if we shared content from home? Assume you serve your content from a server in your home Bad side of current content sharing goes away: – No restrictions on what can be shared – No loss of ownership No terms of use with third parties – Access control managed at user’s homes Independent of third parties’ privacy policies What about the good side of current content sharing? 611/4/2015 Personal home server

Can we preserve the good side? Users must now store their content – Idea: use cheap commodity storage Availability may decrease – Idea: use a cheap always-on residential gateway Performance may decrease – Because bandwidth of residential links is limited – Observation: personal content has often limited audience Cost – Example: Residential gateway + 1TB disk ≅ $140 one time cost (Amazon prices) – For comparison: 1TB/month on Amazon S3 = $95/month 11/4/20157

This talk Is it feasible to share social content from home? – Characterize OSN workloads (Flickr/YouTube) – Characterize home network environment – Evaluate feasibility of content delivery from home 11/4/20158

Understanding OSN workloads Datasets Gathered Flickr: – 11,715 randomly chosen users – 1.3M public photos – Observed views received by photos for 19 days YouTube: – 77,575 randomly chosen users – 1.2M public videos – Collected views received by videos for 166 days Goal: understand how users upload and view content 911/4/2015

How much content do users upload? Avg. total uploaded data: 13.3MB Flickr 103MB YouTube Very few users have >10GB Shared photos/videos can easily fit in commodity storage 1011/4/2015 Users by uploaded content

How popular is the content? Social workloads are not too demanding Never requested in 1 week: 97% of Flickr pics 50% of YouTube videos 94% of videos <100 views 90% YouTube users serve <7.6GB/week = 100Kbps Demands for Flickr much smaller 1111/4/2015

This talk Is it feasible to share social content from home? – Characterize OSN workloads (Flickr/YouTube) – Characterize home network environment – Evaluate feasibility of content delivery from home 11/4/201512

Characterization of home networks Configured home routers to gather measurements Deployed them in 10 households (EU & Korea) Collected 79 days worth of data about: – Availability of gateways and local devices – Spare capacity of access links – Performance of content delivery Goal: estimate availability and performance 11/4/ USB Storage WLAN port

Availability of gateways/devices Gateways periodically send heartbeat messages – If no messages arrive in 5 min, gateway is disconnected Content delivery from gateways provide high availability Availability results Gateway average98% Local device average27% 1411/4/2015

How much spare capacity in access links? 11/4/ Gateways monitored utilization of access links – Recorded total upstream and downstream traffic values 80% of the time, upstream link is not used 95% of the time, upstream traffic less than 15Kbps

This talk Is it feasible to share social content from home? – Characterize OSN workloads (Flickr/YouTube) – Characterize home network environment – Evaluate feasibility of content delivery from home 11/4/201516

Performance of content delivery (photos) Every 10 minutes, gateways fetch 20 60KB photos – From Facebook – From a randomly chosen gateway among the home gateways Photo delivery times PercentileFacebookTestbed 50 th 0.36 sec1.91 sec 80 th 0.81 sec2.91 sec 95 th 1.38 sec5.32 sec 1711/4/2015 Performance generally acceptable for browsing Photo prefetching could improve it considerably

Performance of content delivery (videos) Every hour, gateways fetch an 18MB video – From Facebook – From a randomly chosen gateway in the testbed – Every second, throughput is recorded to simulate streaming Can we support delivery of streaming content? 1811/4/2015

Is there sufficient bandwidth for streaming content? 95% of the time, avg. bandwidth higher than 200Kbps 66% of the time, avg. bandwidth higher than 400Kbps Pre-buf. helps for high quality audio and YouTube-like bit-rates. 11/4/201519

Conclusions Social content sharing is very popular on OSNs – Current architectures results in loss of control over data Can we share social content from home? – Characterized social workload on Flickr/YouTube Volume of uploaded content easily fits on commodity storage A lot of content is never requested A lot of content is unpopular Estimated potential for home-based content delivery – Home gateways provide high availability – Promising for personal photos, and low bit-rate media 11/4/201520

Interesting future directions How to deal with high-quality media? – Idea: friends could help you deliver your content – Prefetching content could improve performance How to deal with very popular content? – Idea: They can be served from centralized infrastructure 11/4/201521