Updates on Technologies, Trends, and Tools
Linux Pro Online * www.linuxpromagazine.com
Off the Beat * Bruce Byfield
The birth of SpinachCon. User testing is often limited in free software, but Deb Nicholson is developing SpinachCon to get developers and users together.
Mir vs. Wayland Show Why Upstream Projects Matter. This controversy wouldn't be happening if upstream development was kept upstream.
Productivity Sauce * Dmitri Popov
Dig up Photos Buried in Email with Lost Photos Found. When you finally pull all your photos from the cloud to your machine, you'll need the Lost Photos Found script.
Keep Track of Your Current Work with a Bash Function. With a simple custom now() function, you can keep track of all the things you are working on.
git-sh: Bash Environment for Git Work. Git-sh transforms the Bash shell into an environment for working with Git.
Paw Prints * Jon "maddog" Hall
University Students! Earn money while coding cool Free and Open Source projects! Check out this way to make some extra money while getting experience!
A Library for Many Jobs * Carsten Schnober
The Joblib Python Library handles problems like parallelization, memorization, and saving and loading objects in no time.
Monitoring HPC Systems: Ganglia * Jeff Layton
Ganglia is probably the most popular monitoring framework/tool. We show you how to install and configure Ganglia and get it up and running on a simple two-node system.
Cloud Orchestration with Cloudify * Martin Loschwitz
Cloud computing is forcing admins to rethink automation, because tools like Puppet do not provide a sufficient range of configuration options. Cloudify offers a new direction for orchestration in the cloud.
Supercharge Your Website with Amazon CloudFront * Joseph Guarino
Users who lose interest in websites that don't respond in the expected time take their clicks elsewhere. We look at ways to improve your WordPress website performance.
Smarter Threat Management with OSSIM * Joseph Guarino
The Open Source Security Information and Event Management system helps IT security professionals cut through noise and gain wisdom and foresight in defending and managing their networks.
Facebook Rolls Out New PHP-Like Programming Language
Facebook has released an open source version of Hack, a programming language for the HHVM runtime platform. HHVM is a virtual machine for PHP created by Facebook.
According to Facebook's Julien Verlaguet, Hack developed as a way for Facebook to address the problem of delivering accurate code in a timely manner at the vast scale of a company like Facebook. "Could early error detection co-exist with rapid iteration, all the while preserving our investment in PHP? Could improved code analysis and introspection help make developers more productive with tools like auto-complete?"
Verlaguet's blog post explains that "Hack has deep roots in PHP …. In fact, most PHP files are already valid Hack files." The main innovation added with Hack is static typing. Hack's type system adds features such as generics, nullable types, type aliasing, and contraints on type parameters. Hack also adds Collections (a type-safe alternative to PHP arrays), lambda expressions, and runtime enforcement of return types and parameter types.
Facebook says it has already migrated nearly all its PHP codebase to Hack. The Hack source code is available at the project website at http://hacklang.org/
Newly Discovered Android Vulnerability Causes Privilege Escalation
Security researchers at Indiana University and Microsoft Corporation have uncovered a new category of vulnerabilities affecting Android operating systems. The new vulnerabilities – known as Pileup problems (short for Privilege Escalation through Update) – are thought to affect every Android device: up to a billion devices around the world.
A Pileup vulnerability allows a malicious application to pre-configure a carefully selected set of privileges that will be enabled when the system updates to a new version. When the system updates, the privileges escalate in a manner that is invisible to the user.
Researchers found these Pileup flaws after an extensive analysis of Android Package Management Service. The issues affect all official Android versions and 3,000 customized versions. The team then developed SecUP, which "deploys a scanner on the user's device to capture the malicious apps designed to exploit Pileup vulnerabilities."
New Encryption System Prevents Server Snooping
MIT computer scientists have developed a new privacy system designed to protect data even from attackers who have full access to the server. The Mylar platform encrypts file data in the browser and then stores the data on the server in encrypted form. The data is thus protected from snooping by NSA or anyone else who might have server access. The user's browser then decrypts the data the next time the user accesses the file.
Client-side encryption is nothing new, but Mylar adds some innovations that make it especially practical for production environments. For example, Mylar supports keyword searches over encrypted documents, even if the data is encrypted using different keys. Mylar also offers a secure means for users to share keys and encrypted data, and it provides a way of ensuring that client code is authentic – even if the server is malicious.
Prototype versions of Mylar are built on the Meteor web framework. The Mylar developers say the presence of the Mylar encryption layer adds an overhead of only 17% with 50 ms latency increase for sending a message in a chat application.
Cisco Announces $1 Billion Cloud Investment
Cisco has announced a new initiative to invest US$ 1 billion in a new cloud service. The plan is to build a "global Intercloud – a network of clouds – together with a set of partners."
According to Cisco, this Intercloud is designed for "high-value application workloads, real-time analytics, scalability, and full compliance with local data sovereignty laws."
Cisco will work with global partners to build this Intercloud network, including communications providers in Europe, Australia, and North America. The company will leverage its existing line of cloud services to provide a platform for rapid development of custom applications.
Robert Lloyd, Cisco's president of development and sales, said, "We have the capability to enable a seamless world of many clouds in which our customers have the choice to enable the right, highly secure cloud for the right workload, while creating strategic advantages for rapid innovation, and ultimately, business growth."
An Even Smaller Raspberry Pi
The Raspberry Pi Foundation announced the impending debut of an even smaller and less expensive version of the credit-card-sized Raspberry Pi computer. The new product, which they call the Raspberry Pi Compute Module, doesn't come with the standard ports but is, instead, designed to fit into a DDR2 SO-DIMM connector.
The developers say their goal with creating the Raspberry Pi Compute Module is to make it easy for users to create their own custom-printed circuit boards and then add the Rasp Pi as a ready-made CPU, with its accompanying OS, APIs, and application software.
Perhaps the most interesting aspect of the announcement is the possibility for integrating the Raspberry Pi Compute Module into industrial uses. Although the Compute Module fits in a DIMM2 slot, the foundation points out that you can't just plug it into the memory slot of your laptop, because the connectors do different things. According to Raspberry Pi Foundation Hardware Director James Adams, the Raspberry Pi developers are launching another product called the Compute Module IO Board to act as a breakout board for prototyping designs based on the Compute Module.
Will Russia and India Join Forces for HPC?
The Russian Academy of Sciences, along with the Russian computer company RSA Group, are reaching out to India in hopes of establishing a collaboration for building a supercomputer to compete with China's dazzling Tianhe-2 system. Tianhe-2 has held onto its lead position in the latest edition of the TOP500 supercomputer list. Russia and India both have supercomputer programs now, but they have lost some momentum over the past five years as China, Japan, the USA, and several European countries have forged ahead.
According to a report in the Economic Times of India, Russian Academy's Boris Shabanov has "…invited a team from the Indian Institute of Science and the Karnataka government to explore the possibility of a supercomputing center in Bangalore, India." According to the report, Indian and Russian scientists believe "the Chinese are way ahead" and are looking for a way to close the gap.
Most supercomputing experts are well aware that building a big computer cluster is only part of the solution – the system is useful only if you have sufficient numbers of professional resources to operate and manage the cluster, and a large enough scientific community to create experiments and write programs to keep the system busy and productive. Besides combining funding sources, a partnership between India and Russia could lead to greater scientific collaboration, which could benefit both scientific communities.
AMD Announces New Graphics Cards
AMD announced a pair of new graphics card for next-generation environments. The FirePro W9100 professional graphics is intended for "4K workstations accelerated by OpenCL." The card comes with "2.62 TFLOPS double-precision of industry leading GPU compute power and ultra-high resolution (4K) multi-display capabilities…"
The company also announced the water-cooled Radeon R9 295X2, which they call "the world's fastest, period." AMD says the R9 295X2 can deliver 11.5 TFLOPS.
The implications for the HPC industry are unclear at this time. At this point, AMD is targeting the FirePro W9100 for professional workstations and aiming the Radeon 295X2 at gamers; however, the affinity of the FirePro W9100 for OpenCL could one day lead to applications with high-performance clusters. The 11.5 TFLOPS computing power of the Radeon is bound to attract attention with the HPC community, but the US$ 1499 per card price point might not lend itself to the commodity pricing of massive HPC installations.
NVidia Unveils Next-Generation GPU
At the 2014 GPU Technology conference in San Jose, CA, NVidia announced details of their Pascal GPU architecture, which they will officially release in 2016. The new Pascal GPUs will support stacked memory, a technology that allows multiple layers of DRAM memory components on a single unit. According to NVidia, stacked memory provides "several times greater bandwidth, more than twice the capacity, and quadrupled energy efficiency" compared with present alternatives.
The increased capacity of GPUs requires a higher-bandwidth connection between GPUs or GPU and CPU. NVidia also unveiled NVLink technology that it says will replace PCI Express for communication between CPU and GPU components. NVidia says NVLink will provide "a more energy-efficient, high-bandwidth path between the GPU and the CPU at data rates 5 to 12 times that of the current PCIe Gen3. NVLink will provide between 80 and 200 GBps of bandwidth, allowing the GPU full-bandwidth access to the CPU's memory system." The first Pascal GPU products will appear in 2016.
Buy this article as PDF
The company is collaborating with Google and Intel to use Kubernetes as an engine for Fuel
Customers can take a free test drive of SLES for HPC on the Azure Cloud
San Francisco-based chip company announces their first fully open source chip platform.
The whole distro gets rebuilt on glibc 2.3
Ubuntu Vendor tries to solve app packaging and distribution problem across distributions.
Founder of ownCloud launches the Nextcloud project.
Will The Machine change the way future programmers think about memory?
The new Torus distributed storage system is available under an open source license on GitHub
Juries decides Google’s use of Java APIs Was Fair Use