Communications and cooperation between development and operations isn't optional, it's mandatory. Flickr takes the idea of "release early, release often" to an extreme - on a normal day there are 10
full deployments of the site to our servers. This session discusses why this rate of change works so well, and the culture and technology needed to make it possible.
Picnik has been using Amazon Web Services for two years. We love it. And we hate it. It also made us lazy. We'll explain what we mean, and how you can benefit.
Ajax techniques can substantially improve the performance of web applications by replacing heavy pages with zippy data packets. Unfortunately, incorrect application of Ajax techniques will significantly degrade performance. It is a sorrowful thing.
Enabling gzip compression on your webserver is one of the easiest and highest impact ways to speed up your site. However, even when this is done, up to 15% of your visitors will still not receive compressed responses.
This session will cover why this problem occurs and what you can do about it.
OpenDNS performs more than 8 billion DNS queries each day, generating mountains of data which we give back to our users. Processing and storing the interesting bits took a special map-reduce pipeline running on seven servers, backed by MySQL and memcached. I'll share all of the false starts and the techniques used to finally cope with more than half a gigabyte of log files per minute.
Go beyond the case study and watch a real-time crash, analysis, repair and recovery of a web site under load. In this workshop Richard brings an actual web farm to the stage, load tests it to failure, diagnoses the problem and repairs it.
In this session we will discuss improvements made to Twitter over the last year to increase scalability. Changes made through internal process, development methodologies, queuing strategies, and operations will also be discussed.
Facebook integrates features and applications deeply into each page. Such deep integration presents both opportunities and challenges to the frontend performance engineering. In this talk, we discuss some of our work on improving and maintaining frontend performance for web pages that contain a mix of many fast-evolving features.
This talk will present some findings from testing reflow times in a variety of browsers under different conditions. I will discuss the method used to gather the data, explore cross-browser reflow differences, and give developers advice on what to avoid when doing dynamic updates to their pages to avoid costly reflows in the browser.
Do you manage big data for hungry users? Do you, or have you considered, using Hadoop? Come join the Cloudera team as we show you the tips and tricks we use to manage some of the largest Hadoop clusters in the world. We'll go over setup, upgrades, monitoring and optimization. Managing petabytes doesn't have to be hard - we'll show you how.
Having ads is a reality for many Web sites. This expert panel will discuss experiences with minimizing ad impact on page performance.
This workshop provides a simple, working introduction to Puppet. It covers basic design guidelines, best practices for building your own Puppet installation, and all of the tools you'll use in the course of using Puppet.
Managing the technical quality of your website has become extremely complex and the number of metrics you collect has skyrocketed.
Faced with potentially hundreds of candidate metrics, how do you select those that are most meaningful? In this session you will learn which technical quality metrics are key for successfully managing your site.
In 2008, www.aol.com started on a migration path from a web platform developed entirely in-house to one made up of common Open Source components. This talk discusses some of the hurdles faced by the www.aol.com Operations team while building an entirely new architecture from the ground up for AOL's homepages.
Close the gap between browser and desktop application performance with native CPU performance for the web. This talk will provide an overview of this Google research technology and demonstrate a few simple applications.
Page Speed is an open source Firebug add-on that gives suggestions on how to improve web page performance. In this talk, we describe the open development model that we have adopted for Page Speed and the importance of decreasing latency to make web applications more responsive.
Live on the main stage: the latest tools you need to analyze websites, including AOL PageTest, HttpWatch, YSlow 2.0, Visual Round Trip Analyzer and Firebug.
Performance-Based Design (PBD) is a model that starts with target business metrics and works its way backwards to set the corresponding performance goals to meet these targets. The model is iteratively fined tuned as more data comes in.
Internet traffic spikes aren't what they used to be. It is now evident that even the smallest sites can suffer the attention of the global audience. This presentation dives into techniques to avoid collapse under dire circumstances. Looking at some real traffic spikes, we'll pinpoint what part of the architecture is crumbling under the load; then, walk though stop-gaps and complete solutions.
Lots of people talk about scaling and performance. But, are they preparing for all the things that could happen? There are multiple problems and there is not one solution to solve them all.
In 2007, Shopzilla, a leading comparison shopping service, realized that our sites were far too slow for our users and had become difficult to change and more difficult still to support. We were presented with a classic dilemma for an established internet market player: Could we really afford the costs and the risks of re-engineering a site platform responsible for all our revenue?
Why is the Web so slow? Steve Souders talks about the issues that prevent web sites from being faster, and what we should expect to see to address these problems.
Learn about how Daily Kos survived what was essentially an eleven month long Slashdotting, while reducing hardware costs and ultimately the amount of work needed to keep the site running.
You've got a sneaking suspicion that design impacts performance. What next? Your engineers know nothing about design and your designers know nothing about performance. How can you get everyone on the same page? Which design flaws must you absolutely avoid? How do engineers slow designs with poor CSS? This presentation covers the best practices in design and OO CSS for fast, maintainable sites.
AOL's swat team of optimization gurus is challenged with infusing the optimization mindset and skillset across 80+ teams and websites...with 4 people. An inside look at the tools, solutions and best practices that have helped them achieve these goals.
We applied several types of page slowdowns to very small percentages of our users for short periods of time and measured the impact in terms of metrics like clicks, time to click, repeat site usage, queries per visit, revenue, etc. This talk is to share the results.
In this workshop, Steve Souders explains the rules he developed as part of YSlow, as well as new best practices he's developed while at Google. With those in mind, he analyzes several of today's popular websites to see how well they follow, or don't follow, this advice.
The competition amongst browsers has been intense over the past year. Come to this session to hear from the Google Chrome, Firefox, and Internet Explorer teams about what they've done to make their browsers faster than ever before, and what developers can do to take advantage of these features.
The ‘Next Web’ is defined by vastly expanded levels of interactivity driven by technologies such as Flash and Silverlight. The ‘Next Web’ brings consumers a new level of rich online experiences including composite e-business transactions that seamlessly integrate interactive video and voice, and social media and Web-to-mobile transactions.