@Scale 2016 lineup announced!

Erin Green

@Scale 2016 is coming soon! Engineers from Airbnb, Amazon, Databricks, Dropbox, Facebook, Flipkart, Google, Instagram, LinkedIn, Netflix, NVIDIA, Oculus, Paypal, Pinterest, Slack, Spotify, Uber, and more will be discussing their newest solutions for addressing engineering challenges and building for scale.

This year, we’ll have four tracks: Data, Dev Tools & Ops, Mobile, and a new addition, Hot Topics (like machine learning and video). Also, Facebook’s head of engineering and infrastructure, Jay Parikh, will kick off the event with a keynote.

Register for the 2016 @Scale conference here. If you do not have an access code yet, message us on the @Scale Community page with your name and email. The event will take place at the San Jose Convention Center on August 31, from 10 a.m. to 6:30 p.m. (registration starts at 8 a.m.; keynotes begin at 10).

For a deeper dive on what’s in store for this year’s show, check out the summary of talks below. Or to see past talks, check out previous events on the @Scale web site.

Data Track

  • Shan He from Uber will provide examples of the available standards, libraries, and the open source stack that Uber developed for visual analytics, mapping, and public-facing data visualizations ranging from SVG and D3 to WebGL and beyond.
  • Facebook’s Yann Collet discusses how the computing world has changed dramatically since the days of Phil Katz’s hugely successful compression algorithm, deflate, commonly known as Zip, and the ways in which Facebook compresses data at scale.
  • Matei Zaharia will cover how having a unified engine enabled new types of applications based on Spark, and how Databricks designed Spark’s API to enable efficient composition.
  • Project Infinite is one of the biggest evolutions in the Dropbox product since the shared folder. Ben Newhouse’s talk will explore how a hack week project turned into a major company initiative, plus the technical decisions and trade-off in its design, and how Dropbox keeps the quality bar high despite deeper integrations across hundreds of millions of devices.
  • Yoshinori Matsunobu will introduce Facebook’s open source, next-generation OLTP SQL database on modern Flash storage called MyRocks — a RocksDB storage engine for MySQL.
  • Amazon will give us a peek under the hood of a cloud-scale relational database service, with Debanjan Saha discussing the log structured distribute storage architecture that helps the company achieve near instantaneous crash recovery, zero-impact continuous backup, and point-in-time restore irrespective of the size of the database and the associated storage volume.
  • Dan Halperin from Google will detail how various APIs and techniques enable an Apache Beam data processing pipeline runner to dynamically rebalance work at runtime, to work around stragglers, and to automatically scale up and down cluster size as a job’s workload changes — efficiently using computing resources without compromising on performance or correctness.
  • LinkedIn’s Priyesh Narayanan and Holla Gopalakrishna are going to focus on the architecture of open source, geodistributed, highly available, and horizontally scalable object store Ambry, built at LinkedIn. They’ll discuss how the design decisions help it scale for both large and small objects, and how it helped solve the main pain points of LinkedIn’s legacy media store.
  • Facebook has developed a family of tools to manage the entire process of training, testing, and deploying machine learning models; Pierre Andrews and Aditya Kalro will explore FBLearner Flow, a pipeline management system that facilitates experimentation, training, and comparison of models, and discuss why it’s used by more than 25 percent of Facebook engineers every month and how it is used to train more than a million models, ingesting 2.3 billion data entries per second, to then be used in production at Facebook.

Dev Tools & Ops Track

  • PayPal’s adoption of Docker as containerization technology has enhanced the company’s developer experience, and Jigar Desai will highlight the company’s journey to docker-ize 2,500 apps and hundreds of thousands of container instances — in a nutshell, developing and running containers at scale.
  • Adrian Zgorzalek describes Facebook’s approach to dealing with resource creep at scale, and the tools and approaches that are used to identify and correct resource creep for Facebook’s PHP developers with regressions on the online data storage and access infrastructure.
  • Lisa Guo will detail a brief history of Instagram’s infrastructure evolution; overall architecture and multi-data center support; tuning of uwsgi parameters for scaling; performance monitoring and diagnosis; and Django/python upgrades.
  • Netflix’s Dianne Marsh will explore how decisions at Netflix, especially those made early on, have built upon one another and led to a highly resilient service.
  • In the past year, Uber’s Mobile Platform team has made major changes to improve development speed and agility for its engineers, with a focus on stability and meeting a high quality bar for its software; Alan Zeino will outline tooling and infrastructure changes that help its engineers check their code into master faster than ever before, and the recent effort to migrate all code into a monolithic repository using the Buck build tool — laying a foundation for future experimentation and growth in Uber’s toolchain.
  • Facebook’s Jeffrey Dunn will cover BrowserLab and techniques that Facebook applies to handle highly dynamic web content, isolate server variance from client-side performance, deal with non-deterministic code, and enable engineers to understand performance implications of their changes.
  • Chunyan Wang will discuss Pinterest’s technical challenges and approaches to A/B testing and experimentation in general, and how, with RocksDB, the company has built stateful online services on replicated RocksDB to create high throughput and low latency online services operating on multiple large online updated and/or offline data sets to power many Pinterest products.
  • Ivan Filho of Google will cover the pursuit of a fair and meaningful Cloud benchmarking framework, PerfKit Benchmarker, from the point of view of a performance engineer.

Mobile Track

  • Facebook’s Anne Ruggirello and Felix Wong will compare the benefits of the approaches — traditional industry approaches, and agile approaches — that Facebook uses in tandem to ship mobile apps in a way that’s both fast and smart.
  • In Valera Zakharov’s talk, he’ll describe why it’s worth having developers write UI tests for their features, as well as present practical information on how the Slack Android team approached the all-too-common issue of being intimidated by UI testing.
  • Quip’s Yunjie Ma will go over the architecture behind Quip’s editor, its evolution as the company adopted React, and how it connects to Quip’s native iOS framework.
  • Igor Canadi and Alex Langenfeld will cover what it takes to get mobile and server developers on the same page when it comes to using GraphQL at Facebook, including the company’s code generation tools that create native classes used for representing GraphQL query responses.
  • Flipkart’s Amar Nagaram will be sharing some of the challenges of building with a mobile-first approach in India and the technology investments it made while adopting a mobile-first approach to everything it did; he’ll also explain why Flipkart became the first Indian app to cross 50 million downloads in India.
  • With HTTP2 push, Facebook has built out a new client/server interaction model, which now makes it possible for the company's Edge/FBCDN servers to push required images and Live streams from the server for a News Feed story or on-going Live stream; Ranjeeth Dasineni and Saral Shodhan speak to how Facebook leverages HTTP2 to achieve lower latencies.
  • Uber’s Bian Jiang and Madhav Srinivasan will go over how Uber implemented network record/replay and scenario-based testing for iOS/Android so its tests would run reliably on CI, covering the challenges faced and the lessons learned on the path to designing Uber's frameworks to ensure that its devs could write predictable and deterministic tests.

Hot Topics Track

  • Facebook wants to ensure a high-quality, consistent viewing experience regardless of the capture device or conditions. Johannes Kopf’s talk will present a new optimization the company has developed to improve one aspect of 360 video viewing quality across Facebook and the Oculus platform, while preserving (and even improving) upload/playback speed and efficiency.
  • Discover Weekly is a personalized mixtape of 30 highly personalized song curated and delivered to Spotify’s 100M active users every Monday. In this talk, Edward Newett dives into the narrative of how Discover Weekly came to be, highlighting technical challenges, data-driven development, and the machine learning models used to power the company’s recommendations engine.
  • Khoi Nguyen of Oculus details how lessons and philosophies learned from React and FLUX are applied to building impactful UI in VR, allowing the team to build a fast, stable, disciplined, and scalable VR product from the ground up.
  • Robert Ober of NVIDIA will present a deep dive in GPU evolution. He'll talk about GPU architectural evolution, changing neural net computation demands, and development flow from training to inference deployment as a utility in the data center. He'll cover data from large-scale GPU deployments, and platform evolution for utility use, where predicting ExaFlop/s (that’s 10^18) capability is no longer a crazy number.
  • Facebook’s Andrew Tulloch will talk about some of the tools and tricks the company uses for scaling both the training and deployment of some of the deep learning models at Facebook, as well as some useful libraries Facebook has open-sourced for production-oriented deep learning principles.

To get a sense of @Scale years past, take a look at the slideshow below:

Bret Taylor (Quip), Selina Tobaccowala (SurveyMonkey), Thuan Pham (Uber), and Mike Krieger (Instagram) discuss scale issues in the keynote panel, moderated by Cade Metz (Wired).
Bret Taylor (Quip), Selina Tobaccowala (SurveyMonkey), Thuan Pham (Uber), and Mike Krieger (Instagram) discuss scale issues in the keynote panel, moderated by Cade Metz (Wired).
In @Scale 2014's data track, Xun Liu delves into Zen, Pinterest's graph storage service.
In @Scale 2014's data track, Xun Liu delves into Zen, Pinterest's graph storage service.
Google's Mikhail Sychev talks about faster images with WebP at YouTube.
Google's Mikhail Sychev talks about faster images with WebP at YouTube.
Jay Parikh from Facebook kicks off @Scale 2014 with a keynote featuring a diverse mix of company attendees.
Jay Parikh from Facebook kicks off @Scale 2014 with a keynote featuring a diverse mix of company attendees.
Jafar Husain closes the web track with a talk about asynchronous programming at Netflix.
Jafar Husain closes the web track with a talk about asynchronous programming at Netflix.
Tamar Bercovici explains how Box is building for scale.
Tamar Bercovici explains how Box is building for scale.
Paul Holden from Uber talks about scaling a mobile-only company.
Paul Holden from Uber talks about scaling a mobile-only company.
Happy hour at the @Scale Conference.
Happy hour at the @Scale Conference.

Keep Updated

Stay up-to-date via RSS with the latest open source project releases from Facebook, news from our Engineering teams, and upcoming events.

Facebook © 2018