Subscribe: All Things Distributed
http://www.allthingsdistributed.com/index.xml
Added By: Feedage Forager Feedage Grade A rated
Language: English
Tags:
amazon  aws  business  companies  customers  data  deep learning  deep  digital  learning  machine learning  new  services 
Rate this Feed
Rate this feedRate this feedRate this feedRate this feedRate this feed
Rate this feed 1 starRate this feed 2 starRate this feed 3 starRate this feed 4 starRate this feed 5 star

Comments (0)

Feed Details and Statistics Feed Statistics
Preview: All Things Distributed

All Things Distributed



Werner Vogels' weblog on building scalable and robust distributed systems.



Last Build Date: Mon, 22 May 2017 14:54:13 PDT

Copyright: Copyright 2011
 



Weekend Reading: Amazon Aurora: Design Considerations for High Throughput Cloud-Native Relational Databases.

Fri, 19 May 2017 15:00:00 PDT

(image)

In many high-throughput OLTP style applications, the database plays a crucial role in achieving scale, reliability, high-performance, and cost efficiency. For a long time, these requirements were almost exclusively served by commercial, proprietary databases. Soon after the launch of the AWS Relational Database Service (RDS) customers gave us feedback that they would love to migrate to RDS. Yet, what they desired more, was a reality that unshackled them from the high-cost, punitive licensing schemes, which came with proprietary databases.

They would love to migrate to an open-source style database like MySQL or PostgreSQL, if such a database could meet the enterprise-grade reliability and performance these high-scale applications required.

We decided to use our inventive powers to design and build a new database engine that would give database systems such as MySQL and PostgreSQL reliability and performance at scale. Meaning, at a level that could serve even the most demanding OLTP applications. It gave us the opportunity to invent a new database architecture that would address to needs of modern cloud-scale applications, departing from the traditional approaches that had their roots in databases of the nineties. That database engine is now known as "Amazon Aurora" and launched in 2014 for MySQL, and in 2016 for PostgreSQL.

Amazon Aurora has become the fastest-growing service in the history of AWS and frequently is the target of migration from on-premise proprietary databases.

In a paper published this week at SIGMOD'17, the Amazon Aurora team presents the design considerations for the new database engine and how they addressed them. From the abstract:

Amazon Aurora is a relational database service for OLTP workloads offered as part of Amazon Web Services (AWS). In this paper, we describe the architecture of Aurora and the design considerations leading to that architecture. We believe the central constraint in high throughput data processing has moved from compute and storage to the network. Aurora brings a novel architecture to the relational database to address this constraint, most notably by pushing redo processing to a multi-tenant scaleout storage service, purpose-built for Aurora. We describe how doing so not only reduces network traffic, but also allows for fast crash recovery, failovers to replicas without loss of data, and fault-tolerant, self-healing storage. We then describe how Aurora achieves consensus on durable state across numerous storage nodes using an efficient asynchronous scheme, avoiding expensive and chatty recovery protocols. Finally, having operated Aurora as a production service for over 18 months, we share lessons we have learned from our customers on what modern cloud applications expect from their database tier.

I hope you will enjoy this weekend's reading, as it contain many gems about modern database design.

"Amazon Aurora: Design Considerations for HighThroughput Cloud-Native Relational Databases", Alexandre Verbitski, Anurag Gupta, Debanjan Saha, Murali Brahmadesam, Kamal Gupta, Raman Mittal, Sailesh Krishnamurthy, Sandor Maurice, Tengiz Kharatishvili, Xiaofeng Bao, in SIGMOD '17 Proceedings of the 2017 ACM International Conference on Management of Data, Pages 1041-1052 May 14 – 19, 2017, Chicago, IL, USA.




Faster, higher, stronger: How the digitalization of industry is redefining value creation

Wed, 03 May 2017 10:00:00 PDT

This article titled "Wie die Digitalisierung Wertschöpfung neu definiert" appeared in German last week in the "Größer, höher, weiter (bigger, higher, further)" column of Wirtschaftwoche. Germany's "hidden champions" – family-owned companies, engineering companies, specialists – are unique in the world. They stand for quality, reliability and a high degree of know-how in manufacturing. Hidden champions play a significant role in the German economy; as a result, Germany has become one of the few countries in Western Europe where manufacturing accounts for more than 20% of GDP. By contrast, neighboring countries have seen a continuous decline in their manufacturing base. What's more, digital technologies and business models that are focused on Industry 4.0 (i.e., the term invented in Germany to refer to the digitalization of production) have the potential to reinforce Germany's lead even more. According to estimates by Bitkom, the German IT industry association, and the Fraunhofer Institute of Industrial Engineering IAO, Germany's hidden champions will contribute a substantial portion to the country's economic growth by 2025 and create new jobs. At the same time, many experts believe the fundamental potential of Industry 4.0 has not even been fully leveraged yet. The power of persistence versus the speed of adjustment Most of Germany's hidden champions have earned their reputation through hard work: they have been optimizing their processes over decades. They have invested the time to perfect their processes and develop high-quality products for their customers. This has paid off – and continues to do so. However, digital technologies are now ushering in a paradigm change in value creation. Manufacturing can be fully digitalized to become part of a connected "Internet of Things" (IoT), controlled via the cloud. And control is not the only change: IoT creates many new data streams that, through cloud analytics, provide companies with much deeper insight into their operations and customer engagement. This is forcing Mittelstand companies to break down silos between departments, think beyond their traditional activities, and develop new business models. In fact, almost every industrial company in Germany already has a digitalization project in place. Most of these companies are extracting additional efficiency gains in their production by using digital technologies. Other companies have established start-ups for certain activities, or pilot projects aimed at creating showcases. But many of these initiatives never get beyond that point. The core business, which is doing well, remains untouched by all this. And one of the main reasons why is because the people with the necessary IT expertise in Mittelstand companies sometimes are not sitting at the strategy table as often as they should. Will these initiatives be enough to secure the pole position for Germany's Mittelstand? Probably not. Companies in growth markets are catching up. China's industry, for example, is making huge progress – something that took years to achieve in other places. The role of Chinese manufacturers in the worldwide market is changing: from low-cost workbench to global provider of advanced technology. Market leaders from Germany therefore realize they cannot afford to rest on their laurels. Competitors from the software side are also reshuffling the balance of power, because their offerings will create a completely new market alongside the traditional business of Mittelstand toolmakers and mechanical and systems engineering companies. If new and innovative companies, such as providers of data analytics, specialized software providers or companies that can bundle complementary offerings, appear on the scene, traditional manufacturing would suddenly become just one module among many – namely manufacturing-as-a-service. Creating added value in an Industry-4.0 environment often happens when B2B companies integrate B2C approaches, in turn sparking change in[...]



Coming to STATION F: The first Mentor's Office powered by AWS!

Wed, 12 Apr 2017 01:00:00 PDT

I am excited to announce that AWS is opening its first Mentor's Office at STATION F in Paris! The Mentor's Office is a workplace exclusively dedicated to meetings between AWS experts and the startups. STATION F is the world's biggest startup campus. With this special offer starting at the end of June, at the campus opening, AWS increases the support already available to startup customers in France. All year long, AWS experts will deliver technical and business assistance to startups based on campus. AWS Solutions Architects will meet startup members for face-to-face sessions, to share guidance on how cloud services can be used for their specific use cases, workloads, or applications. Startup members will also have the possibility to meet with AWS business experts such as account managers, business developers, and consultants. They can explore the possibilities of the AWS Cloud and take advantage of our IT experience and business expertise. With these 1:1 meetings, AWS delivers mentoring to startups to help them bring their ideas to life and accelerate their business using the cloud. AWS will also provide startups with all of the benefits of the AWS Activate program, including AWS credits, training, technical support, and a special startup community forum to help them successfully build their business. For more details about the Mentor's Office at STATION F, feel free to contact the AWS STATION F team. With this opening, Amazon continues to build out global programs to support startup growth and to speed up innovation. Startups can also apply to other Amazon programs to boost their businesses, such as: Amazon Launchpad, which makes it easy for startups to launch, market, and distribute their products to hundreds of millions of Amazon customers across the globe. Alexa Fund, which provides up to $100 million in venture capital funding to fuel voice technology innovation. After the launch of AWS in 2006, we saw an acceleration of French startups adopting the cloud. Successful French startups already using AWS to grow their businesses, across Europe and around the world, include Captain Dash, Dashlane, Botify, Sketchfab,Predicsis, Yomoni, BidMotion, Teads, FrontApp, Iconosquare, and many others. They all get benefits from AWS's highly flexible, scalable, and secure global platform. AWS eliminates the undifferentiated heavy lifting of managing underlying infrastructure and provides elastic, pay-as-you-go IT resources. We have also seen start-ups in France using AWS to grow and become household names in their market segment, such as Aldebaran Robotics (SoftBank Robotics Europe). This startup uses AWS to develop new technologies. They are able to concentrate their engineering resources on innovation, rather than maintaining technology infrastructure, which is leading to the development of autonomous and programmable humanoid robots. Cloud is also an opportunity for startups to reach security standards that were not accessible before. For example, PayPlug is an online payment by credit card solution enabling e-merchants to enrich the customer experience by reinventing the payment experience. Such a service requires suppliers to get PCI DSS certification for the "Service Provider" level, a very demanding certification level. Using AWS's PCI DSS Level 1 compliant infrastructure, Payplug has been certified by L'ACPR (L'Autorité de contrôle prudentiel et de resolution, the French supervisory for prudential and resolution authority) as a financial institution, a major step in their development. I look forward to meeting the builders of tomorrow at STATION F in the near future. Go French Startups! [...]



Back-to-Basics Weekend Reading: Twenty years of functional MRI: The science and the stories.

Fri, 07 Apr 2017 09:00:00 PDT

(image)

I will be returning this weekend to the US from a very successful AWS Summit in Sydney, so I have ample time to read during travels. This weekend however I would like to take a break from reading historical computer science material, to catch up on another technology I find fascinating, that of functional Magnetic Resonace Imaging, aka fMRI.

fMRI is a functional imagine technology, meaning that it just records the state of the brain at one particular point in time, but the changing state over a period of time. The basic technology records brain activity by measuring changes in blood flow through the brain. The technology relies on the fact that cerebral blood flow and neuronal activation are coupled. When an area of the brain is in use, blood flow to that region also increases.

There have been significant advances in the use of fMRI technology, but mostly in research. It also comes with significant ethical questions: if you can "read" someone's brain, what are you allowed to do what that knowledge?

For my flight back to the US this weekend I will read two papers: one by Peter Bandettini pubslished in NeruImagine about the history of fMRI and one from Poldrack and Farah on the state of the art in fMRI and its applications, published in Nature.

"Twenty years of functional MRI: The science and the stories, Peter A. Bandettini, Neuroimage 62, 575–588 (2012)

"Progress and challenges in probing the human brain", Russell A. Poldrack and Martha J. Farah, Nature 526, 371–379 (15 October 2015)




Välkommen till Stockholm – An AWS Region is coming to the Nordics

Tue, 04 Apr 2017 00:00:00 PDT

Today, I am very excited to announce our plans to open a new AWS Region in the Nordics! The new region will give Nordic-based businesses, government organisations, non-profits, and global companies with customers in the Nordics, the ability to leverage the AWS technology infrastructure from data centers in Sweden. The new AWS EU (Stockholm) Region will have three Availability Zones and will be ready for customers to use in 2018. Over the past decade, we have seen tremendous growth at AWS. As a result, we have opened 42 Availability Zones across 16 AWS Regions worldwide. Last year, we opened new regions in Canada, India, Korea, the UK, and the US. Throughout the next year we will see another five zones, across two AWS Regions, come online in France and China. However, we do not plan to slow down and we are not stopping there. We are actively working to open new regions in the locations our customers need them most. In Europe, we have been constantly expanding our footprint. In 2007, we opened our first AWS Region in Ireland and since then have opened additional regions, in Germany and the UK, with France still to come. After the launch of the AWS EU (Stockholm) Region, there will be 13 Availability Zones in Europe for customers to build flexible, scalable, secure, and highly available applications. It will also give customers another region where they can store their data with the knowledge that it will not leave the EU unless they move it. As well as AWS Regions, we also have 24 AWS Edge Network Locations in Europe. This enables customers to serve content to their end users with low latency, giving them the best application experience. This continued investment in Europe has led to strong growth as many customers across the region move to AWS. Organizations across the Nordics—Denmark, Finland, Iceland, Norway, and Sweden—have been increasingly moving their mission-critical applications to AWS. This has led us to steadily increase our investment in the Nordics to serve our growing base of enterprise, public sector, and startup customers. In 2011, AWS opened a Point of Presence (PoP) in Stockholm to enable customers to serve content to their end users with low latency. In 2014 and 2015 respectively, AWS opened offices in Stockholm and Espoo, Finland. We have also added teams in the Nordics to help customers of all sizes as they move to AWS, including account managers, solutions architects, business developers, partner managers, professional services consultants, technology evangelists, start-up community developers, and more. Some of the most successful startups in the world, including Bambora, iZettle, King, Mojang, and Supercell are already using AWS to deliver highly reliable, scalable, and secure applications to customers. Supercell is responsible for several of the highest grossing mobile games in history, and they rely on AWS for their entire infrastructure. With titles like Boom Beach, Clash of Clans, Clash Royale, and Hay Day, Supercell has 100 million people playing their games every single day. iZettle, a mobile payments startup, is also ‘all-in’ on AWS. After finding it cost prohibitive to use colocation centers in local markets where their users are based, iZettle decided to give up hardware. They migrated their IT infrastructure, including mission-critical payments platforms, to AWS in just six weeks. After migrating, database queries that took six seconds now take three seconds in their AWS infrastructure. That’s 100% faster. Some of the largest, and most well respected, enterprises in the Nordics also depend on AWS to power their businesses, enabling them to be more agile and responsive to their customers. These companies include ASSA ABLOY, Finnair, Husqvarna Group, IKEA, Kauppalehti, Kesko, Sanoma, Scania, Schibsted, Telenor, and WOW Air. Scania, a world leading manufacturer of commercial vehicles, is using AWS to bring advanced technologies to their trucks, buses, coa[...]



Back-to-Basics Weekend Reading: An Implementation of a Log-Structured File System

Thu, 30 Mar 2017 09:00:00 PDT

(image)

This weekend I am travelling to Australia for the first AWS Summit of 2017. I find on such a long trip, to keep me from getting distracted, I need an exciting paper that is easy to read. Last week's 'Deep Learning' overview would have not met those requirements.

One topic that always gets me excited is how to take computer science research and implement it in production systems. There are often so many obstacles that we do not see much of this work happening. For example when building Dynamo, where we put a collection of different research technologies together in production, we struggled with all the assumptions the researchers had made. At times, it makes research unsuitable for production (e.g. real systems do not fail by stopping in a nice and clean way).

In the early nineties, Mendel Rosenblum and John Ousterhout had made a major breakthrough in the design of file systems with "The Design and Implementation of a Log-Structured File System." That alone is an interesting paper to read, but this weekend we will be looking at the actual implementation of an LFS by Margo Seltzer and other members of the BSD team.

It is one of the first papers to describe the implementation of a research system, and measure the result within a production system. I hope you will also enjoy it!

"An implementation of a log-structured file system for UNIX.", Margo Seltzer, Keith Bostic, Marshall Kirk Mckusick, and Carl Staelin. 1993, In Proceedings of the USENIX Winter 1993 Conference Proceedings on USENIX Winter 1993 Conference Proceedings (USENIX'93). USENIX Association, Berkeley, CA, USA, 3-3.




Back-to-Basics Weekend Reading: Deep learning in neural networks

Fri, 24 Mar 2017 21:00:00 PDT

(image)

In the past few years, we have seen an explosion in the use of 'Deep Learning' as its software platforms and the supporting hardware mature, especially as GPUs with larger memories become widely available. Even though this is a recent development, 'Deep Learning' has entrenched historical roots, tracing back all the way to the sixties or possibly earlier.

By reading-up on its history, we get a better understanding of the current state of the art of 'Deep Learning algorithms' and the 'Neural Networks' that you build with them.

There is a broad set of papers to read if we want to dive deep into the history. It would take us multiple weekends. Instead, we will be reading an excellent overview paper from 2014 by Jürgen Schmidhuber. Jürgen evaluates the current state of the art in 'Deep Learning' by tracing it back to its roots. Ergo, we get excellent historical context.

Enjoy!

"Deep Learning in Neural Networks: An Overview." Jürgen Schmidhuber, in Neural Networks, Volume 61, January 2015, Pages 85-117 (DOI: 10.1016/j.neunet.2014.09.003)




Amazon Makes it Free for Developers to Build and Host Most Alexa Skills Using AWS

Wed, 15 Mar 2017 10:00:00 PDT

Amazon today announced a new program that will make it free for tens of thousands of Alexa developers to build and host most Alexa skills using Amazon Web Services (AWS). Many Alexa skill developers currently take advantage of the AWS Free Tier, which offers one million AWS Lambda requests and up to 750 hours of Amazon Elastic Compute Cloud (Amazon EC2) compute time per month at no charge. However, if developers exceed the AWS Free Tier limits, they may incur AWS usage fees each month. Now, developers with a live Alexa skill can apply to receive a $100 AWS promotional credit and can earn an additional $100 per month in AWS promotional credits if they incur AWS usage charges for their skill – making it free for developers to build and host most Alexa skills. Our goal is to free up developers to create more robust and unique skills that can take advantage of AWS services. We can't wait to see what you create. How It Works If you have one or more live Alexa skills, you are eligible to receive a $100 AWS promotional credit to be used toward AWS fees incurred in connection with your skills. Additionally, if you continue to incur skill-related AWS charges that exceed the initial $100 promotional credit, you will also be eligible to receive monthly AWS promotional credits of $100. All you need to do is apply once. Apply Now > Build and Host Alexa Skills with AWS With the new program, if you exceed the AWS Free Tier due to growth of your skill, or are looking to scale your skill using AWS services, you will be eligible to receive AWS promotional credits to be applied to AWS services such as Amazon EC2, Amazon Simple Storage Service (Amazon S3), Amazon DynamoDB, and Amazon CloudFront. For example, you can use DynamoDB to create more engaging skills that have context and memory. In a game with memory, you could pause for a few hours and then keep going (like the Wayne Investigation, or Sub War). Or, to give your customers a more immersive experience, consider incorporating audio files via Amazon S3 to stream short audio bursts, games, podcasts, or news stories in your skill. Many of our most engaging skills, like Ambient Noise and RuneScape Quests – One Piercing Note, add audio sounds to soothe and voiceovers and sound effects to make the in-game experience more immersive. Build a Skill Today - Special Offers Our skill templates and step-by-step guides are a valuable way to quickly learn the end-to-end process for building and publishing an Alexa skill. You can get started quickly with the city guide template or fact skill template, or use the Alexa SDK for Node on GitHub to create a custom skill. Plus, if you publish a skill, you'll receive an Alexa dev t-shirt. Quantities are limited. See Terms and Conditions. Additional Resources For more information on getting started with developing for Alexa, check out the following resources: Voice Design Best Practices Alexa Skills Kit (ASK) Alexa Voice Service (AVS) The Alexa Fund ASK Developer Forums Weekly Developer Office Hours [...]



How companies can become magnets for digital talent

Mon, 13 Mar 2017 10:00:00 PDT

This article titled "Wie Unternehmen digitale Talente anziehen" appeared in German last week in the "Tipps für Arbeitgeber" section of Wirtschaftwoche. The rise in digital business models is a huge challenge for recruiting and talent selection. The sort of skills businesses need today are in short supply. How companies can prepare themselves to attract the best talents for shaping their digital business. Digitalization offers almost endless possibilities to communicate faster, work more efficiently, and be more creative – in real-time. But groundbreaking digital business models need pioneers: creators, forward-looking thinkers and inventors who don't hesitate to leave the beaten path, embody ownership, and who understand how to translate customers' wishes into superb new products, services and solutions that evolve with speed. It is a no-brainer, that getting the right talent on board can decisively accelerate a company's digital transformation. At the same time, if your daily corporate practice doesn't fulfill their expectations regarding a vibrant and flexible working culture and a social media-minded environment, digital natives will simply turn their back on you and go elsewhere. Finding those kind of people is not easy. There are probably only a few companies that can say, they already have a sufficient number of such employees among their staff. Job openings for machine learning scientists, data analytics experts, IT security experts or developers are already difficult to fill, and the demand for this knowledge will increase significantly in the next few years as customers show their demand for digital engagements. The market for digital skills is "hot", in the U.S. as well as in Germany. And these talents are by no means coveted only by companies that always had a digital business model to begin with; suppliers to the automotive industry, financial services companies, and retailers also, urgently need product managers, and technical staff who can quickly make their organizations digitally attractive to their customers. Recruiting and selection in the digital age therefore needs to be tackled in a more strategic way than in the past. So how do you position your company as an attractive employer for digital talent? Preparing the organization for a new beginning One way is to eliminate rigid structures, previously the enemy to digital thinking. Digitalization involves, among others, suddenly converging areas that used to be siloed. Take industrial companies. In the past, their sales departments defined specifications according to the customer's wishes, which were then transferred step by step into the manufacturing process. These days, it's expected that everything should happen almost simultaneously. Previously, the top priorities for IT departments were equipping data centers with hardware, purchasing software, and further developing proprietary software. Today, companies take their server capacity and software from the cloud. These changes have to be taken into account when scanning the market for talent. At Düsseldorf-based fashion retailer Peek&Cloppenburg, for example, the business, development and IT functions are increasingly cooperating with each other because they realize that isolated departments and rigid hierarchies can slow down the organization's innovative strength and speed. That is also why employees have more and more room to make decisions themselves. P&C's digital transformation is supported by an in-house consulting team that helps the specialized departments analyze and digitize those processes that strengthen the customer touchpoints. The freedom to create Another way to make your company attractive for digital talent is to give them as much creative freedom as possible AutoScout24, a Munich-based online marketplace fo[...]



Back-to-Basics Weekend Reading: The Foundations of Blockchain

Fri, 10 Mar 2017 08:00:00 PST

(image)

More and more we see stories appearing, like this one in HBR by MIT Media Lab's Joi Ito and crew. It praises the power of blockchain as a disruptive technology, on par with how "the internet" changed everything.

I am always surprised to see that these far-reaching predictions are made, without diving into the technology itself. This weekend I would like to read about some of the technologies that predate blockchain, as they are its fundamental building blocks.

Blockchain technology first came on the scene in 2008, as a core component of the bitcoin cryptocurrency. Blockchain provides transactional, distributed ledger functionality that can operate without a centralized, trusted authority. Updates recorded in the ledger are immutable, with cryptographic time-stamping to achieve serializability. Blockchain's robust, decentralized functionality is very attractive for global financial systems, but can easily be applied to contracts, or operations such as global supply chain tracking.

When we look at the foundation of blockchain, there are three papers from the nineties that describe different components whose principles found its way into blockchain. The 91 paper by Haber and Stornetta describes how to use crypto signatures to time-stamp documents. The 98 paper by Schneier and Kelsey describes how to use crypto to protect sensitive information in log files on untrusted machines. Finally, the 96 paper by Ross Anderson describes a decentralized storage system, from which recorded updates cannot be deleted.

I hope these will enlighten your fundamental understanding of blockchain technology.

"How to Time-Stamp a Digital Document", Stuart Haber, and W. Scott Stornetta, In Advances in Cryptology – Crypto ’90, pp. 437–455. Lecture Notes in Computer Science v. 537, Springer-Verlag, Berlin 1991.

"Cryptographic Support for Secure Logs on Untrusted Machines", Bruce Schneier, and John Kelsey, in The Seventh USENIX Security Symposium Proceedings, pp. 53–62. USENIX Press, Januar 1998.

"The Eternity Service", Ross J. Anderson. Pragocrypt 1996.




Back-to-Basics Weekend Reading: Why Do Computers Stop and What Can Be Done About It?

Sat, 04 Mar 2017 09:00:00 PST

(image)

"Everything fails, all the time." A humble computer scientist once said. With all the resources we have today, it is easier for us to achieve fault-tolerance than it was many decades ago when computers began playing a role in critical systems such as health care, air traffic control and financial market systems. In the early days, the thinking was to use a hardware approach to achieve fault-tolerance. It was not until the mid-nineties that software fault-tolerance became more acceptable.

Tandem Computer was one of the pioneers in building these fault-tolerant, mission-critical systems. They used a shared-nothing multi-cpu approach. This is where each CPU had its own memory- and io-bus, and all were connected through a replicated shared bus, over which the independent OS instances could communication and run in lock step. In the late seventies and early eighties, this was considered state of the art in fault-tolerance.

Jim Gray, the father of concepts like transactions, worked for Tandem on software fault-tolerance. To be able to build better systems, he went deep in deconstructing the kind of failures Tandem customers were experiencing. He wrote up his findings in his "Why do Computers Stop" report. For a very longtime, this would be the only study available on reliability in production computer systems.

As important as the study is, the paper additionally covers "What can be done about it." Jim, for the first time, introduces concepts like process-pairs and transactions as the basis for software fault-tolerance. This is one of the fundamental papers of fault-tolerance in distributed systems, and I am going to enjoy reading it this weekend. I hope you will also.

"Why Do Computers Stop and What Can Be Done About It?", Jim Gray, June 1985, Tandem Technical report 85.7




Back-to-Basics Weekend Reading: Byzantine Generals

Fri, 24 Feb 2017 20:00:00 PST

(image)

In Reliable Distribution Systems, we need to handle different failure scenarios. Many of those deal with message loss and process failure. However, there is a class of scenarios that deal with malfunctioning processes, which send out conflicting information. The challenge is developing algorithms that can reach an agreement in the presence of these failures.

Lamport described that he was frustrated with the attention that Dijkstra had gotten for describing a computer science problem as the story of dining philosophers. He decided the best way to attract attention to a particular distributed systems problem was to present it in terms of a story; hence, the Byzantine Generals.

Abstractly, the problem can be described in terms of a group of generals of the Byzantine army, who camped with their troops around an enemy city. Communicating only by messenger, the generals were required to agree upon a common battle plan. However, one or more of them may be traitors who would try to confuse the others. The problem is: to find an algorithm that ensures the loyal generals will reach an agreement.

It is shown, using only oral messages, this problem is solvable if and only if more than two-thirds of the generals are loyal. So, a single traitor can confound two loyal generals. With unforgeable written messages, the problem is solvable for any number of generals and possible traitors. This weekend, I will be going back in time and reading three fundamental papers that laid-out the problems, and its first solutions. In the SIFT paper, the problem is first described, the "reaching agreement" paper describes the fundamental 3n+1 processor solution, and the last paper reviews and generalizes the previous results.

Maybe you will enjoy them as well. "SIFT: Design and Analysis of a Fault-Tolerant Computer for Aircraft Control" John H. Wensley, Leslie Lamport, Jack Goldberg, Milton W. Green, Karl N. Levitt, P. M. Melliar-Smith, Robert E. Shostak, Charles B. Weinstock, in Proceedings of the IEEE 66, October 5, 1978

"Reaching Agreement in the Presence of Faults" M. Pease, R. Shostak, and L. Lamport, 1980, J. ACM 27, 2 (April 1980), 228-234.

"The Byzantine Generals Problem", Lamport, L.; Shostak, R.; Pease, M. (1982), ACM Transactions on Programming Languages and Systems. 4 (3): 382–401. doi:10.1145/357172.357176.




Back-to-Basic Weekend Reading: Monte-Carlo Methods

Fri, 10 Feb 2017 11:00:00 PST

(image)

I always enjoy looking for solutions to difficult challenges in non-obvious places. That is probably why I like using probabilistic techniques for problems that appear to be hard, or impossible to solve deterministically. The probabilistic approach may not result in the perfect result, but it may get you very close, and much faster than deterministic techniques (which may even be computationally impossible).

Some of the earliest approaches using probabilities in physics experiments resulted in the Monte Carlo methods. Their essential idea is using randomness to solve problems that might be deterministic in principle. These are a broad class of computational algorithms that rely on repeated random sampling to obtain numerical results.

The Monte Carlo methods can be traced back to Stanislaw Ulam, John van Neuman, and Nick Metropolis at the Los Alamos Scientific Laboratory in the late 40s. The Monte Carlo methods were crucial in the simulations of the Manhattan Project, given the limited computational power available in those days

The paper I will be reading this weekend is the original paper from 1949, by Metropolis and Ulam. For fun, I’ve also decided to add a second paper by Herbert Anderson, who was a member of the Manhattan project. Anderson’s paper describes the use of Monte Carlo methods, and the computers in the Manhattan project.

The Monte Carlo Method”, Nicholas Metropolis, S. Ulam, Journal of the American Statistical Association, Vol. 44, No. 247. (Sep., 1949), pp. 335-341.

"Metropolis, Monte Carlo and the MANIAC", Anderson, Herbert L., Los Alamos Science, (1986) 14: 96–108.




Back-to-Basics Weekend Reading - Bloom Filters

Fri, 03 Feb 2017 11:00:00 PST

(image)

Listening to the "Algorithms to Live By" audio on my commute this morning, once again I was struck by the beauty of Bloom Filters. So, I decided it is time to resurrect the 'Back-to-Basics Weekend Reading' series, as I will be re-reading some fundamental CS papers this weekend.

In the past, I have done some weekend reading about Counting Bloom Filters, but now I am going even more fundamental, and I invite you to join me.

Bloom Filters, conceived by Burton Bloom in 1970, are probabilistic data structures to test whether an item is in a set. False positives are possible, but false negatives are not. Meaning, if a bit in the filter is not set, you can be sure the item is not in the set. If it is in the set, the mapped item may be in the set.

This is a hugely important technique if you need to process and track massive amounts of unique data units, as it is very space-efficient. From Dynamo and Postgresql, to HBase and Bitcoin, Bloom Filters are used in almost all modern distributed systems. This weekend I will be reading the original paper by Bloom from 1970, and another more recent survey paper that describes several variants and applications that have been developed over the years.

"Space/Time Trade-offs in Hash Coding with Allowable Errors", Bloom, Burton H., in Communications of the ACM, 13 (7): 422–426

"Cache-, Hash- and Space-Efficient Bloom Filters", Putze, F.; Sanders, P.; Singler, J., in Demetrescu, Camil, Experimental Algorithms, 6th International Workshop, WEA 200




A survival strategy for the digital transformation

Sun, 29 Jan 2017 09:00:00 PST

This article titled "Überlebensstrategie für die digitale Transformation" appeared in German last week in the "Die Zukunft beginnt heute (the future starts today)" section of Wirtschaftwoche. Smaller companies have a lot to gain in the digital era – provided they adopt the right mindset. The winners will be those that view their business from the eyes of their customers and understand that fast-paced innovation is the key to long-term growth. With this mindset they can take on even the largest enterprises who are slow to adapt to the fast moving digital reality. The digital era is here. Companies that haven't realized that by now will fall behind. In many industry segments and markets, for example, platform services, we've already witnessed how start-ups and niche providers have unleashed a revolution. Companies that used to be dominant but stare at all the changes around them for too long in a state of paralysis can quickly end up in a struggle to survive – look no further than the entertainment and music industry where streaming services have eaten a significant piece of the cake of the hard-copy providers. The better you understand why and how small and medium-sized players can conquer global markets, you'll be better positioned to come out as a winner. Digitalization allows even the smallest companies to think big because it puts technology into their hands that was previously hard to access and too costly to acquire. But adopting modern technologies alone is not enough to win the market battle. However, when new technologies are combined with a passion for putting the interests of the customer at the heart of everything you do, they can give agile companies that decisive push to the front of the pack. And Mittelstand companies have fantastic opportunities, provided they digitize more of their existing business models. Especially in manufacturing-based industries, introducing more software that complements hardware can eliminate fixed costs and allow you to quickly scale up to a global level. Companies that embrace this can rise to become leading players, taking the place previously reserved for the 'big shots' in their industries. Digitalization starts with having the right mindset: namely one aimed at creating innovative digital experiences. Continuous customer-centric experimentation has been the leading principle at Amazon from the start, in both our e-commerce activities as well as in Amazon Web Services. We found out that by organizing our innovation efforts around customers' needs, we could innovate very fast. Since 2006, Amazon Web Services has introduced far more than 2,500 new services and features, and around 90% of them were the result of wishes articulated directly by customers. The first requirement for developing an innovation mindset is to adapt your offerings fast to changing customer behavior. There are great examples of German companies that already do this. One is Vorwerk and its all-in-one cooking machine, called Thermomix. This premium product has been on the market for more than 50 years. But the way people cook today is different than in the 1960s. Today, cooking must be convenient, fast, and healthy. People want to prepare meals without too much effort, yet some appreciate a bit of guidance during the entire cooking process, from picking a new recipe from Vorwerk's cloud-based database to putting the finished meal on the table. Companies that want to adopt a digital innovation mindset should start leaving their comfort zone – even if they don't (yet) feel any pressure to change. Or put another way: they have to develop an inner drive to not just deliver on their custo[...]



Expanding the AWS Cloud Introducing the AWS Europe (London) Region

Wed, 14 Dec 2016 00:00:00 PST

In November 2015, Amazon Web Services announced that it would launch a new AWS infrastructure region in the United Kingdom. Today, I'm happy to announce that the AWS Europe (London) Region, our 16th technology infrastructure region globally, is now generally available for use by customers worldwide. UK companies are using AWS to innovate across diverse industries, such as energy, manufacturing, medicaments, retail, media, and financial services and the UK is home to some of the world's most forward-thinking businesses. These include startups like Fanduel, JustEat, and Monzo to enterprises such as British Gas, Trainline, Travis Perkins, News UK, the Financial Times. The British Government is also helping to drive innovation and has embraced a cloud-first policy for technology adoption. Take Peterborough City Council as an example. The council has deployed IoT Weather Stations in Schools across the City and is using the sensor information collated in a Data Lake to gain insights on whether the weather or pollution plays a part in learning outcomes. London has also established itself as a critical center for the financial services sector and a significant hub for venture capital activity across all Europe. The City's thriving venture capital and start-up accelerator communities are fueling growth and innovation, making it one of the most important locations in the world to do business. AWS is working with incubators and accelerators such as SeedCamp and Techstars, in London; Ignite100 in Newcastle; and DotForge in Sheffield and Manchester to help startups make the most of the cloud. We believe in our customers and are investing for the long term. With the AWS Europe (London) Region, we look to better serve end users in the UK. With the launch of the AWS Europe (London) Region, AWS can enable many more UK enterprise, public sector and startup customers to reduce IT costs, address data locality needs, and embark on rapid transformations in critical new areas, such as big data analysis and Internet of Things. All around us we see that the AWS capabilities foster a culture of experimentation with businesses of all sizes. AWS is not only affordable but it is secure and scales reliably to drive efficiencies into business transformations. I have been humbled by just how much our UK customers have been able to achieve using AWS technology so far. In just this past month we've had HSBC, ARM, Missguided, and most recently at re:Invent 2016, Trainline, talking with us about how they are using AWS to transform and scale their businesses. Following are just a few of the reasons that customers have given us for building their business on the AWS Cloud: Blend seamlessly into the digital world: With the rising importance of technology-driven business transformation, an emphasis on certain enterprise and consumer-based opportunities emerges. To take advantage of the game-changing opportunities, businesses are looking to blend into the digital world. Take GoSquared, a UK startup that runs all its development and production processes on AWS, as an example. GoSquared provides various analytics services that web and mobile companies can use to understand their customers' behaviors. With AWS, GoSquared can process tens of billions of data points every day from four continents to provide customers with a single view. Use catalysts for real-time business models: The Internet of Things (IoT) is undoubtedly driving a philosophy of interconnecting people, process, and machines to create massive volumes of data that has potential for disruptive change. The BMW Group is using AWS for its new connected-car applic[...]



Expanding the AWS Cloud: Introducing the AWS Canada (Central) Region

Thu, 08 Dec 2016 10:30:00 PST

Earlier this year, Amazon Web Services (AWS) announced it would launch a new AWS infrastructure region in Montreal, Quebec. Today, I'm happy to share that the Canada (Central) Region is available for use by customers worldwide. The AWS Cloud now operates in 40 Availability Zones within 15 geographic regions around the world, with seven more Availability Zones and three more regions coming online in China, France, and the U.K. in the coming year. The Canadian opportunity Canada has set forth a bold innovation agenda grounded in entrepreneurship, scientific research, growing small and medium-sized businesses with a focus on environmentally friendly technologies, and the transition to a digital economy. This agenda leverages the transformative aspects of technology and encourages Canadian companies, universities, governments, not-for-profits, and entrepreneurs to contribute to building a durable innovation economy. Given this, enterprises, public sector bodies, startups, and small businesses are looking to adopt agile, scalable, and secure public cloud solutions. The new Canada (Central) Region offers a robust suite of infrastructure, management, and developer services that can enable innovators to deploy market-leading applications. Access to secure, scalable, low-cost AWS infrastructure in Canada allows customers to innovate and provide tools to meet privacy, sovereignty, and compliance requirements. The new AWS Canada (Central) Region also continues the company's focus on delivering cloud technologies to customers in an environmentally friendly way. AWS data centers in Canada will draw from a regional electricity grid that is 99 percent powered by hydropower. For more information about AWS efforts, see AWS & Sustainability. Some examples of how current customers use AWS are: Cost-effective solutions Kik Interactive is a Canadian chat platform with hundreds of millions of users around the globe. It adopted Amazon Redshift, Amazon EMR and AWS Lambda to power its data warehouse, big data, and data science applications, supporting the development of product features at a fraction of the cost of competing solutions. Rapid time to market The Globe and Mail (Globe) is one of Canada's most read newspapers, with a national weekly circulation of 4.7 million. To increase online readership, it worked with AWS Partner Network (APN) Partner ClearScale to develop a personal recommendation capability. The solution, which leverages Amazon Kinesis, Amazon DynamoDB, and Amazon EMR to collect, store, and process the data, as well as AWS CloudFormation and AWS OpsWorks to support the Globe's DevOps environment, was deployed in three months—less than half the time it would have taken had the newspaper built it on-premises. Enterprise-class services available from Canada Box is an enterprise content management and collaboration platform used by more than 41 million users and 59,000 businesses—including 59% of the Fortune 500. It relies on the scale and power of Amazon Simple Storage Service (Amazon S3) to deliver in-region storage options to businesses and organizations across the world in Canada, Japan, Singapore, Australia, Ireland, Germany, and the U.S., as part of its Box Zones ecosystem. Having the ability to provide these services locally enables Box to better serve Canadian enterprises looking for cloud solutions while ensuring their data is stored inside Canada. Increasing agility Lululemon Athletica is a Canadian athletic apparel company that is using AWS Lambda, AWS CodePipeline, and AWS CodeDeploy to rapidly build and deploy their digital marketing and e-commerce so[...]



Transforming Development with AWS

Thu, 01 Dec 2016 12:00:00 PST

In my keynote at AWS re:Invent today, I announced 13 new features and services (in addition to the 15 we announced yesterday). My favorite parts of James Bond movies is are where 007 gets to visit Q to pick up and learn about new tools of the trade: super-powered tools with special features which that he can use to complete his missions, and, in some cases, get out of some nasty scrapes. Bond always seems to have the perfect tool for every situation that he finds himself in. * At AWS, we want to be the Q for developers, giving them the super-powered tools and services with deep features in the Cloud. In the hands of builders, the impact of these services has been to completely transform the way applications are developed, debugged, and delivered to customers. I was joined by 32,000 James Bonds at the conference today from all around the world, and we introduced new services focused on accelerating this transformation across development, testing and operations, data and analytics, and computation itself. Transformation in Development, Testing, & Operations Although development and operations are often overlooked, they are the engines of agility for most organizations. Today, cCompanies cannot afford to wait two or three years between releases, and; customers have found that continually releasing incremental functionality to customer frequently reduces risk and improves quality. Today, we're making available broad new services which that let builders prepare and operate their applications more quickly and efficiently, and respond to changes in both their business and their operating environment, swiftly. We launched the following new services and features today to help. AWS OpsWorks for Chef : a fully managed Chef Automate environment, available through AWS OpsWorks to fuel even more automation and reduce the heavy lifting associated with continuous deployment. Amazon EC2 Systems Manager : A collection of tools for package installation, patching, resource configuration, and task automation on Amazon EC2. AWS Codebuild: A new, fully managed, extensible service for compiling source code and running unit tests, which is integrated with other application lifecycle management services— such as AWS CodeDeploy, AWS CodeCommit, and AWS CodePipeline— for dramatically decreasing the time between iterations of software. Amazon X-Ray: A new service to analyze, visualize, and debug distributed applications, allowing builders to identify performance bottlenecks and errors. Personal Health Dashboard: A new personalized view of AWS service health for all customers, allowing developers to gain visibility into service health issues which that may be affecting their application. AWS Shield : protective Protective armor against distributed denial of service (DDoS) attacks, available as Shield Standard and Shield Advanced. Shield Standard gives DDoS protection to all customers using API Gateway, Elastic Load Balancing, Route 53, CloudFront, and EC2. Shield Advanced protects against more sophisticated DDoS attacks, with access to help through a 24x7 AWS DDoS response team. Transformation in Data In the old world, access to infrastructure resources was a big differentiator for big, wealthy companies. No more. Today, any developer can have access to a wealth of infrastructure technology services which that bring advanced technology to their fingertips times in the Cloud. The days of differentiation through infrastructure are behind us; the technology is now evenly distributed. Instead, most companies today and in the future will differentiate themselves through [...]



Bringing the Magic of Amazon AI and Alexa to Apps on AWS.

Wed, 30 Nov 2016 10:00:00 PST

From the early days of Amazon, Machine learning (ML) has played a critical role in the value we bring to our customers. Around 20 years ago, we used machine learning in our recommendation engine to generate personalized recommendations for our customers. Today, there are thousands of machine learning scientists and developers applying machine learning in various places, from recommendations to fraud detection, from inventory levels to book classification to abusive review detection. There are many more application areas where we use ML extensively: search, autonomous drones, robotics in fulfillment centers, text processing and speech recognition (such as in Alexa) etc. Among machine learning algorithms, a class of algorithms called deep learning has come to represent those algorithms that can absorb huge volumes of data and learn elegant and useful patterns within that data: faces inside photos, the meaning of a text, or the intent of a spoken word.After over 20 years of developing these machine learning and deep learning algorithms and end user services listed above, we understand the needs of both the machine learning scientist community that builds these machine learning algorithms as well as app developers who use them. We also have a great deal of machine learning technology that can benefit machine scientists and developers working outside Amazon. Last week, I wrote a blog about helping the machine learning scientist community select the right deep learning framework from among many we support on AWS such as MxNet, TensorFlow, Caffe, etc. Today, I want to focus on helping app developers who have chosen to develop their apps on AWS and have in the past developed some of the seminal apps of our times on AWS, such as Netflix, AirBnB, or Pinterest or created internet connected devices powered by AWS such as Alexa and Dropcam. Many app developers have been intrigued by the magic of Alexa and other AI powered products they see being offered or used by Amazon and want our help in developing their own magical apps that can hear, see, speak, and understand the world around them. For example, they want us to help them develop chatbots that understand natural language, build Alexa-style conversational experiences for mobile apps, dynamically generate speech without using expensive voice actors, and recognize concepts and faces in images without requiring human annotators. However, until now, very few developers have been able to build, deploy, and broadly scale applications with AI capabilities because doing so required specialized expertise (with Ph.D.s in ML and neural networks) and access to vast amounts of data. Effectively applying AI involves extensive manual effort to develop and tune many different types of machine learning and deep learning algorithms (e.g. automatic speech recognition, natural language understanding, image classification), collect and clean the training data, and train and tune the machine learning models. And this process must be repeated for every object, face, voice, and language feature in an application. Today, I am excited to announce that we are launching three new Amazon AI services that eliminate all of this heavy lifting, making AI broadly accessible to all app developers by offering Amazon's powerful and proven deep learning algorithms and technologies as fully managed services that any developer can access through an API call or a few clicks in the AWS Management Console. These services are Amazon Lex, Amazon Polly, and Amazon Rekognition that will help AWS app developers build these next genera[...]



MXNet - Deep Learning Framework of Choice at AWS

Tue, 22 Nov 2016 09:00:00 PST

Machine learning is playing an increasingly important role in many areas of our businesses and our lives and is being employed in a range of computing tasks where programming explicit algorithms is infeasible. At Amazon, machine learning has been key to many of our business processes, from recommendations to fraud detection, from inventory levels to book classification to abusive review detection. And there are many more application areas where we use machine learning extensively: search, autonomous drones, robotics in fulfillment centers, text and speech recognitions, etc. Among machine learning algorithms, a class of algorithms called deep learning hascome to represent those algorithms that can absorb huge volumes of data and learn elegant and useful patterns within that data: faces inside photos, the meaning of a text, or the intent of a spoken word. A set of programming models has emerged to help developers define and train AI models with deep learning; along with open source frameworks that put deep learning in the hands of mere mortals. Some examples of popular deep learning frameworks that we support on AWS include Caffe, CNTK, MXNet, TensorFlow, Theano, and Torch. Among all these popular frameworks, we have concluded that MXNet is the most scalable framework. We believe that the AI community would benefit from putting more effort behind MXNet. Today, we are announcing that MXNet will be our deep learning framework of choice. AWS will contribute code and improved documentation as well as invest in the ecosystem around MXNet. We will partner with other organizations to further advance MXNet. AWS and Support for Deep Learning Frameworks At AWS, we believe in giving choice to our customers. Our goal is to support our customers with tools, systems, and software of their choice by providing the right set of instances, software (AMIs), and managed services. Just like in Amazon RDS―where we support multiple open source engines like MySQL, PostgreSQL, and MariaDB, in the area of deep learning frameworks, we will support all popular deep learning frameworks by providing the best set of EC2 instances and appropriate software tools for them. Amazon EC2, with its broad set of instance types and GPUs with large amounts of memory, has become the center of gravity for deep learning training. To that end, we recently made a set of tools available to make it as easy as possible to get started: a Deep Learning AMI, which comes pre-installed with the popular open source deep learning frameworks mentioned earlier; GPU-acceleration through CUDA drivers which are already installed, pre-configured, and ready to rock; and supporting tools such as Anaconda and Jupyter. Developers can also use the distributed Deep Learning CloudFormation template to spin up a scale-out, elastic cluster of P2 instances using this AMI for even larger training runs. As Amazon and AWS continue to invest in several technologies powered by deep learning, we will continue to improve all of these frameworks in terms of usability, scalability, and features. However, we plan to contribute significantly to one in particular, MXNet. Choosing a Deep Learning Framework Developers, data scientists, and researchers consider three major factors when selecting a deep learning framework: The ability to scale to multiple GPUs (across multiple hosts) to train larger, more sophisticated models with larger, more sophisticated datasets. Deep learning models can take days or weeks to train, so even modest improvements here make a huge difference in the spee[...]