Stuff The Internet Says On Scalability For April 1, 2011

Submitted for your reading pleasure, no foolin'...

• Quotable Quotes:
  • @zateriosystems: thinking about scalability?, are you OK to double your capacity in one week?, a startup should be ready...ready to jump.
  • @sklacy: Maybe what I should have said is "Design for scalability, deploy without it."
  • @MikeHale: Scalability is customer 2000 having the same experience as customer 1 #sqlsat67
  • @LusciousPear: The meaning of #NoSQL is shut up
  • @deobrat: The biggest bottleneck to scalability are ignorant developers. Most don't even try saving extra CPU cycles or memory bytes :(
  • @w_westendorp: .@ijansch: Cloud computing is like outsourcing your scalability problems
  • @edyavno: Billy Newport essentially just affirmed the theme I've been propagating: "Distributed Caching is the enterprise NoSQL" #strangeloop #nosql
  • @monkchips: HP CEO Leo Apotheker says "relational databases are becoming less and less relevant to the future stack"
• 12,000 Requests per second with Nginx, PHP and Memcached and Implementing Full-Page caching with NGinx and PHP. Martin Fjordvald with a great pair of articles on full page caching with an implementation in PHP.
• Prioritizing Principlas in "On Designing and Deploying Internet-Scale Services by Brad Porter. Brad highlights some of the most important ideas in implementing internet-scale services: 1) Keep it simple, 2) Expect failures, 3) Support version roll-back, 4) Maintain forward-and-backward compatibility, 5) Give enough information to diagnose. 3&4 are tough tough tough. Some comments by readers as well.
• Waypoints From Big Data to Smart Data.  Krishna Sankar thinks  a transition from Big Data to Smart Data is happening...Smart Data = Big Data (data at scale, connected and probably with hidden variables) + context + embedded interactive (inference/reasoning) models.
• Greg Linden with some good hits from his latest reading list. Topics include Google on social, Yahoo raising chickens, Google on VLDBs, the doom of both netbooks and tablets, YouTube becoming the TV of the future.
• Quora: What should Larry Page focus on now that he is CEO of Google? Fire all the people they worked so hard to hire, refocus products, forget social, create new markets, quit dying your organization and return to your startup roots, have a good soak in the hot tub. Repeat.
• Christina Warren with How Groupon Uses the Cloud to Scale Its Business. I did not know that they used  Salesforce and its Force.com to power their deals. Good description of the deal process too.
• Dan Rayburn with Three New White Papers Released On The Telco CDN Space: breaking the constraints of scale; six trends in CDNs; 4 keys to telco CDN success.
• Microsoft with an interesting new graph database called Trinity. Trinity is a graph database and computation platform over distributed memory cloud. As a database, it provides features such as highly concurrent query processing, transaction, consistency control. As a computation platform, it provides synchronous and asynchronous batch-mode computations on large scale graphs. Trinity can be deployed on one machine or hundreds of machines.
• Now for something completely different, TalkTechToMe will read HighScalability.com and other tech feeds to you over a phone.
• A co-Relational Model of Data for Large Shared Data Banks by Erik Meijer, Gavin Bierman. Contrary to popular belief, SQL and NoSQL are just two sides of the same coin.
• What's this, James Golick says Most NoSQL DBs Are Not Scalable and follows it up with What does "scalable database" mean? To be a scalable a product must provide automatic partitioning. 
• George Beech with a StackExchange experience report on Performance tuning Intel NICs. By tuning their NICs peak traffic went much smoother. Have you tuned your NICs today?
• Google releases snappy,  a compression/decompression library. A point Google has made in the past is that cheap compression can double your bandwidth with very little cost.
• Data Center Knowledge with A Look Inside A Microserver-Powered Data Center. What does a microserver-driven data center look like? Different.
• Venture Beat reports LinkedIn reaches 100 million users.
• SnowFlock: Rapid Virtual Machine Cloning for Cloud Computing. A new cloud computing abstraction that instantaneously clones a VM into multiple replicas on different hosts.
• replicas running on different hosts
• Retouched Bloom Filters: Allowing Networked Applications to Trade Off Selected False Positives Against False Negatives. An extension that makes the Bloom filter more flexible by permitting the
  removal of selected false positives at the expense of generating random false negatives.