Reading: Filesystems II

I've collected several papers that I think you might find interesting. I would encourage you to read them all, but my suspicion is that time might prevent you from doing so this weekend.

That said, you should read at least two. These are drawn from the USENIX organization (host to many awesome, open access papers on current research in systems), the ACM (ASPLOS, PLOS) and others.

I have ranked these in terms of my perception of how interesting or cool they are. (They all made the list, so don't take the last one to be awful, or anything like that.) I particularly like the first two because they tackle interesting problems in interesting ways.

• Making Problem Diagnosis Work for Large-Scale, Production Storage Systems PDF
  Michael P. Kasick and Priya Narasimhan, Kevin Harms

• TierStore: A Distributed Filesystem for Challenged Networks in Developing Regions PDF
  Michael Demmer, Bowei Du, and Eric Brewer

• Failure Trends in a Large Disk Drive Population PDF
  Eduardo Pinheiro, Wolf-Dietrich Weber and Luiz Andre Barroso

• Dynamo: Amazon’s Highly Available Key-value Store PDF
  Giuseppe DeCandia, Deniz Hastorun, Madan Jampani, Gunavardhan Kakulapati, Avinash Lakshman, Alex Pilchin, Swaminathan Sivasubramanian, Peter Vosshall and Werner Vogels

• Cumulus: Filesystem Backup to the Cloud PDF
  Michael Vrable, Stefan Savage, and Geoffrey M. Voelker

• The Google File System PDF
  Sanjay Ghemawat, Howard Gobioff, and Shun-Tak Leung

• Tolerating File-System Mistakes with EnvyFS PDF
  Lakshmi N. Bairavasundaram†, Swaminathan Sundararaman, Andrea C. Arpaci-Dusseau, Remzi H. Arpaci-Dusseau

• Finding a needle in Haystack: Facebook’s photo storage PDF
  Doug Beaver, Sanjeev Kumar, Harry C. Li, Jason Sobel, Peter Vajgel

• dsync: Efficient Block-wise Synchronization of Multi-Gigabyte Binary Data PDF
  Thomas Knauth and Christof Fetzer

Submission

For each of the two papers that you chose, please write up a summary PDF. That guide, produced by the writing center at the University of Frasier Valley should serve you well in structuring your summary.

As always, written work should be submitted as a Markdown document (.md) or a compressed file containing the LaTeX source of your report (.tar.gz) and a compiled PDF of the output. LaTeX submissions should be sources only, without temporary files (.aux, .log, etc.).

On Linux, you might use UberWriter or ReText for Markdown, and on Windows, MarkdownPad. Any editor works for LaTeX, although some prefer tools that are designed for the purpose.

Assignment: Summarizing FS Article

Naming Convention: username-filesystems

Moodle Link: http://moodle2.berea.edu/course/view.php?id=2243

---

---