Stuff The Internet Says On Scalability For June 10, 2011
Submitted for your scaling pleasure:
- Achievements:
- Every day, Amazon Web Services adds enough new capacity to support all of Amazon.com’s global infrastructure through the company’s first 5 years, when it was $2.7 billion annual revenue. From Cloud Computing Is Driving Infrastructure Innovation by James Hamilton. Where's is all that money be spent? Facilities, servers, power, and popcorn.
- Evernote hits 10 million users. StackExchange hits 1 million users. No lawsuits expected in either case.
- Neural waves of brain. The brain's waves drive computation, sort of, in a 5 million core, 9 Hz computer.
- Scaling, Scaling, Scaled: textPlus Turns Two, Hits 10 Billion Messages Sent Milestone
- Quotes of a quotable essence:
- robinduckett: FACT: You are not a web developer if you need third party services which handle scalability so you can "focus on the programming".
- Twitter’s Bain: Facebook May Have More Scale, We Have More Engagement
- shervin: Fallibility without malleability sheds scalability.
- uisdans: Fat client/server is over. We're moving from #apps #social web #iaas to a #nui #richapp #bigdata #paas spanning the private/public cloud
- Ex-Google Engineer Says the Company's Software Infrastructure is Obsolete. Arguments don't follow IMHO. Creating a global infrastructure in the large is a very different goal that following the latest trends for personal projects. Though it is no doubt limiting to have to use this infrastructure for everything.
- Datacenters are becoming their own technological niche. Why use Internet tech in datacenter? TCP becomes Data Center TCP (DCTCP) - an enhancement to the TCP congestion control algorithm for data center networks.
- The Grasshopper Outage: Co-Founders Response. Action items: replace NetApp with Pillar, fix core networking issue at the disaster recovery site, perform full disaster recovery audit.
- Computer Systems Colloquium (Fall 2009). Guest lecturers at Stanford lay down the knowledge on the Self, Biological Circuits, DRAM errors, Wave Glider, Parallel Computation, Rethinking Time in Distributed Systems, Scaling Ports to 100,000+.
- The Story Behind Ruby 1.9.3 Getting 36% Faster Loading Times.
- ABS: The guts of Android. Jake Edge has an interesting look as Android as an embedded OS.
- Zero-downtime Deployment (and Rollback) in Tomcat; a walkthrough and a checklist. Parallel deployment is the ability to deploy more than one version of your web application in parallel, making all versions available under the exact same URL. Saves a lot of work.
- Reinout van Rees with a Scalability panel talking about Django, blocking operations, monitoring, Puppet vs Chef, ORM, and a lot more.
- PyCon 2011: Handling ridiculous amounts of data with probabilistic data structures. Titus Brown hates C++, but still managed to stumbled across some neat CS techniques that scale well, and are easy to understand and trivial to implement.
- Martin Odersky with a Recap of Scala Days. Find the papers here at Scala Days 2011. Allow me to sum up Scala Days 2011: multi-core; distributed; applicative; and we're hiring
- MIT on Which technologies get better faster? Researchers found that the greater a technology’s complexity, the more slowly it changes and improves over time.
- App Engine: Google's deepest secrets as a service. Cade Metz with a really go look at the App Engine.
- Measuring the scalability of SQL and NoSQL systems. Roberto Zicari with an interesting interview of Adam Silberstein and Raghu Ramakrishnan, from the PNUTS team, and benchmarking database systems. It could be that future hardware devices will mask the cost of random writes so well that higher-level log structured techniques will become redundant. On the other hand, higher-level log structured approaches have more computational resources at their disposal, and also have more information about the application.
- A LOT of bandwidth that is never evenly utilised, Greg Ferro describes in L2 MultiPath Basic Design Differences, fundamental changes in how high density networks are designed. Plus Show 47 – Load Balancers – Good Thing We Didn’t Step in It.
- StackExchange open sourced .NET web performance mini-profiler responsible for 2x-10x performance improvements.
- Content Delivery Summit Videos Now Available For Viewing
