Cliff Click explains 0xdata's H20, a clustering and in-memory math and statistics solution (available for Hadoop and standalone), writing H20's memory representation and compression in Java, low latency Java vs GCs, and much more.
Todd Montgomery talks about improving serialization times and throughput can by understanding how your computer processes and stores data. With this new understanding, architects and developers can build their own protocols to efficiently transmit data. Todd's advice sheds new light on why software developers choose their current serialization and marshaling techniques and how they can improve.
Graham Lee talks to InfoQ at QCon London 2013 about the creation of the Discworld app, and how the media-rich application benefited from automated testing and performance optimisations to be performant on retina class iPads.
Alex Papadimoulis shares his thoughts on distribution vs delivery, decoupling infrastructure (pull) from application (push) deployments and keeping delivery systems simple, especially for web scale applications. In particular Alex describes three different types of roll-outs: Live, Rolling and Parallel and their applicability (cloud-based delivery vs in-house servers).
Martijn Verburg discusses his new start-up jClarity, which offers performance tooling for the Cloud. He also provides an update on the Adopt a JSR and Adopt OpenJDK programs.
Jim Hirschauer describes the application monitoring tool landscape, KPIs and metrics to consider when monitoring, and compares monitoring traditional vs. cloud-based applications. He talks about performance considerations when instrumenting code, how organizations can be 'Smarter' about their Big Data, and looks at what's new in AppDynamics 3.7.
In this InfoQ interview, Michael Nygard explores some of the available loopholes in the CAP theorem helping architects to engineer distributed systems that meet their needs. He also discusses new patterns he’s observed since his book, Realease IT and shares his thoughts on continuous delivery, DevOps and ALM.
Ken Little talks about scaling Tumblr to keep up with their blogging users: scaling the data model, sharding, their PHP frontend and the Scala backend, and much more.
Serkan Piantino explains how Facebook has managed to scale up, what types of errors occur in an architecture that size and how to handle them, RAM vs disk, and much more.
Rich Hickey and Justin Sheehy talk about scalability and transactionability of datastores. They explain tradeoffs for achieving read and/or write scalability on top of Datomic and Riak.