InfoQ Homepage Performance & Scalability Content on InfoQ
-
Is Sun offering the new G1 Garbage Collector Only with a Paid Support Contract?
Sun's Garbage First garbage collector (nicknamed G1) has been released with Java Update 1.6.0_14 (6u14). Although this low pause, server style collector has been long awaited by the developer's community, at this point Sun allows production usage only to paying customers and this has raised concerns.
-
Rich Hickey on Clojure's Features and Implementation
In this interview from QCon London 2009, Rich Hickey talks about Clojure. The discussion includes the ideas behind Clojure's STM support, what other concurrency primitives Clojure supports and which ones might get added in the future. Other topics covered are Clojure's AOT support, the role and implementation of multimethods, Clojure ports to other systems and much more.
-
Performance Roundup: Heap Stacks Boost Threads in 1.8.x, MacRuby AOT, ZenProfile and EventHooks
New patches by Joe Damato improve the efficiency of Ruby 1.8.x's green threads with heap stacks: instead of copying the entire stack at every context switch, the patches actually switch between different stacks. Ryan Davis released zenprofile and event_hook for efficient profiling. Also: work on a MacRuby Ahead of Time compiler using LLVM has started.
-
Flex: Engine Yard's New Cloud Offering
Engine Yard announced Flex at this year's Rails Conf. Flex runs on Amazon's EC2, but unlike its smaller brother Solo, Flex scales over more than one instance. We talked to Michael Mullany, VP of Marketing at Engine Yard to get more information.
-
Erlang and Ruby Roundup: 37Signals, Erlectricity
37Signals is the latest company to use Erlang in combination with Ruby. The recent Erlang Factory conference also had other examples of Erlang use at EngineYard as well as a talk about Erlectricity, the library that connects Erlang and Ruby.
-
8 Best Practices to Improve Scalability
Wille Faler proposes 8 scalability and performance best practices like offloading the database, using caching, minimizing network traffic and others.
-
Ruby Performance Roundup: Ruby 1.9.1 Real World Performance, GC vs EventMachine, Ruby Compiler
When it comes to performance, Ruby 1.9.1 seems to clearly beat it's predecessor, as a new real world benchmark shows. There are other ways, though, to avoid inefficiencies in 1.8.x, by knowing how the GC works or when to disable pthreads. Also: Viktor Hokstad's been busy writing a Ruby compiler from scratch.
-
Presentation: The Market Risk System
In this presentation recorded at QCon London 2008, David Harper and Dominique Delarue present the Market Risk System used by BNP to retrieve, store and analyze risk data for all trading activities of their bank.
-
Presentation: Financial Transaction Exchange at BetFair.com
Betfair is the world's largest betting exchange with a transaction volume the equivalent of over half the combined equity trading volume of every major stock exchange in the world. In response to an increase in transaction volume coupled with a decrease in value per transaction, Betfair launched a number of initiatives to dramatically increase transaction processing capacity and reduce cost.
-
Declarative Concurrency For Ruby With Dataflow
Part of the Ruby language was influenced by functional programming techniques. Larry Diehl brings a declarative concurrent model to Ruby by importing the concept of unification from Oz Language with the Dataflow project.
-
CrossTwine Linker Aims to Boost MRI and 1.9.1 Performance
CrossTwine aims to improve the performance of MRI, 1.9.1, and Python. Unlike new Ruby VMs, the CrossTwine Linker allows to improve efficiency of the existing interpreters and VMs, while keeping their complete feature set. The company plans to offer services to use the CrossTwine Linker technology to speed up specific applications.
-
Adobe ConnectNow-Terracotta Integration Uses Distributed Shared Memory to Manage Scalability
Adobe ConnectNow is a free web conferencing and collaboration software. It's built on the Adobe Flash Collaboration Service (AFCS). For the Java server-side component of AFCS, Adobe integrated ConnectNow application with Terracotta software to take advantage of Terracotta's shared application memory store to address scalability of the web application.
-
What Is .NET 4.0 Beta 1 Going to Bring to PLINQ?
Ed Essey, a Program Manager on the Microsoft Parallel Computing team, wrote on the latest enhancements to PLINQ that are to appear in .NET 4.0 Beta 1 that is to be released soon. Some of them are: “With” Operators Pattern, Execution Mode, Cancellation, Refactoring, Performance Improvements.
-
Article: Performance Analysis and Monitoring with Perf4J
In this article Alex Devine explains how Java developers can take advantage of Perf4J, an open source toolset for adding code timing statements and for logging, analyzing and monitoring the results.
-
Is Parallel Programming Hard? Prof. Guy Blelloch Argues That It Isn’t
In an essay on Cilk Arts, Professor Guy Blelloch argues that parallel programming is not intrinsically hard, but rather a question of abstraction. The three problems identified by Blelloch are a lack of training in parallel thinking, separating parallel implementation from algorithms, and determinism. After detailing each, he explains why he thinks they can be overcome.