SEARCH
ISTC-CC NEWSLETTER
RESEARCH HIGHLIGHTS
Ling Liu's SC13 paper "Large Graph Processing Without the Overhead" featured by HPCwire.
ISTC-CC provides a listing of useful benchmarks for cloud computing.
Another list highlighting Open Source Software Releases.
Second GraphLab workshop should be even bigger than the first! GraphLab is a new programming framework for graph-style data analytics.
ISTC-CC Abstract
AppFlux: Taming App Delivery via Streaming
Conference on Timely Results in Operating Systems (TRIOS'15), in conjunction with SOSP, Oct. 2015. The 25th ACM Symposium on Operating Systems Principles, October 4-7, 2015, Monterey, California, USA.
Ketan Bhardwaj, Pragya Agarwal, Ada Gavrilovska, Karsten Schwan
Georgia Institute of Technology
The number of apps downloaded for smart devices has surpassed 80 billion, with trends suggesting continued substantial increases. The resulting volume of app installs and updates puts pressure on the existing appdelivery infrastructure due to interactions of end user devices with app stores via the Internet that involve app stores’ data center, content delivery network’s (CDNs) points of presence and the Internet Service Provider (ISP) pipes. This paper presents ‘AppFlux’ – a novel app streaming approach to app delivery which reduces the load app delivery poses on the app infrastructure, and relieves users from having to deal with unnecessary updates potentially saving bytes on their costly data plan. By leveraging the emerging ‘edge cloud’ tier of the Internet, the AppFlux approach provides ‘just-in-time’ delivery of apps and their updates while (i) reducing the traffic due to app installs and updates seen in the last mile of Internet by up to 70%, (ii) facilitating twice as fast app delivery compared to CDNs and (iii) App streaming can potentially lead to 20% improvements in the app load times on devices. With its implementation for the Android app ecosystem, AppFlux achieves this in a completely invisible manner, i.e., without requiring any changes by app developers or any explicit actions by end users.
FULL PAPER: pdf
