Stuff The Internet Says On Scalability For November 18, 2011

Every kiss begins with HighScalability:

  • Amazon and the secret to life: 42; 10,240 cores
  • Many quatloos worth of quotable quotes:
    • @alesroubicek :  State kills scalability
    • @cincura_net : Wrong. *Shared* state kills scalability.
    • @kpshea : When I think "cloud" computing, I imagine the gaseous Star Trek blob that ate red blood cells (your sensitive data).
    • @kotobuki : I'm interested in scalability of personal fabrication. How to 'scale' in batch production stages will be a key, but still there are barriers.
    • @simonraikallen : The two rules of scalability testing: (1) The bottleneck is always the database (2) You can never predict what the bottleneck will be.
    • @mastachand : Spot on Post by @storagebod about PetaByte-Scale storage. it's not only about scaling. It's about "Simple Scalability" (http://www.storagebod.com/wordpress/?p=930)
    • @mschopman : don't forget about scalability, better response times due to shorter distances, ability to add nodes pretty fast
    • @marksbirch : Photo: newyorker: The way we are producing data, we may need a place even bigger than heaven to hold it all…
    • @htng : 3 scalability issues: supporting large # of channels vs. small; scaling to # of transactions, avoiding lengthy response times
    • @kellabyte : Architecture how NetFlix writes data low latency w/ 10s of in-mem queues but durable by 3 geo-located writes http://bit.ly/tCRdRC #BIGData
  • Why Stack Exchange Isn’t in the Cloud. It's about love, the love of computers, and what you love you don't let other people own. Also How StackOverflow Scales with SQL Server
  • NoSQL No More: Let’s double down with MoreSQL. Alex Tatiyants with an impassioned plea for programmers to throw down these new fangled databases and return to a comfortable and much loved past. To bring this world about Alex wants SQL Everywhere, to spread FUD about NoSQL, and to recognize with enough effort SQL can work for every problem. This brave old vision should bring comfort everywhere to people wearing very small shoes.
  • SpiderDuck: Twitter's Real-time URL Fetcher. A service at Twitter that fetches all URLs shared in Tweets in real-time, parses the downloaded content to extract metadata of interest and makes that metadata available for other Twitter services to consume within seconds.
  • Quora: In what ways is AWS better than most of its competitors? Some good answers: advanced feature set, vast ecosystem, lack of traditional competitors, simple services, not trying to recreate legacy datacenter environments, no lock-in, performance and flexibility.
  • Release -  is a project that aims to scale the radical concurrency-oriented programming paradigm to build reliable general-purpose software, such as server-based systems, on massively parallel machines (100 000 cores).
  • Riddle me this economists: why do people bid so much on the Amazon spot market? Is it a lack of inventory? Market manipulation? Some bid fiddling? Bad software architecture? Or is just normal market mechanics?
  • When the Internet shuts down maybe we can move to unlicensed spectrum: Open Wireless vs. Licensed Spectrum: Evidence from Market Adoption
  • Benchmark: JavaScript  vs single-threaded C++ vs multithreaded C++ vs JavaScript using Fabric Engine. JavaScript is slow, multithreaded C++ and JavaScript + Fabric are about the same speed. Modulo all the usual benchmark BS.
  • Someone actually trying to build the future instead of playing grownup FUD games: Big data requires fast networks -  research network consisting of more than 12,000 miles of fiberoptic cable and capable of 100 Gigabits-per-second speed.
  • The Story of Networks as told by Amira Skomorowska. A great look at the immense reach of graph theory: the famous seven bridges, random networks, 6 degrees of separation, the strength of weak ties, controlling systems, the hidden influence of social networks, and the power of networks.
  • More interesting Quick links brought to you by Greg Linden