Enterprises often suffer from a curse of accumulated risk. The applications they're building are large and complex, and thus need a lot of risk management process, which means releases take longer— which makes them riskier, and we're back where we started. What can we do to break out of this deadly, self-reinforcing feedback loop?
In this talk, John will explore how combining key ideas from Progressive Delivery and observability lead to better, more resilient releases for teams of all sizes. In particular, John will cover how combining intelligent traffic shadowing with metrics-canary phased rollouts can get you back to a happier place for releasing your software. Releases should be boring—let's make them stay that way.
"I love feature flags!" is a quote from Jason McGee, the VP overseeing IBM Kubernetes Service. This single post in Slack is when the IBM Kubernetes Service team knew they had fundamentally changed how they build, deliver, and operate the IBM Kubernetes Service. Since its inception three years ago, the IBM Kubernetes Service has been a leader in IBM by pushing the boundaries for tools and culture. It has evolved from a cumbersome monolith to a true cloud-native service that manages over 20,000 Kubernetes Clusters. They have embraced feature flags, initially for managing deployments, and now for the progressive delivery of new features and operational regions. This talk will cover the team's journey over the past three years, things they did wrong, things they did right, and where they want to go in the future.
Feature flags are a great way to enhance the speed with which you can collaborate with colleagues, develop in production, and also learn from real-world usage. Using feature flags to enable things selectively for users is an established pattern. Now, take it a step further and use them to enable fast iterative shipping for your development team. I’ll talk about how we’ve unlocked the ability to use production to safeguard production at GitHub - as counter intuitive as that might sound!
Tray.io helps customers connect hundreds of tools with ease and simplicity, allowing teams to focus on their primary tasks. We use LaunchDarkly to help us with the same! Learn how LaunchDarkly has empowered teams at Tray.io to progressively deliver value to our customers, while practicing good feature management.
The world’s top-performing organizations have embraced Progressive Delivery to gain a competitive advantage and exceed their customers' expectations. Hear from LaunchDarkly co-founders on how our latest innovations will empower all teams – not just Development – to build, operate, and learn from their software.
LaunchDarkly recently crossed a milestone by serving more than 2 trillion feature flags per day. Want to learn more about the system that powers LaunchDarkly’s well known speed and reliability? In this talk, I'll cover key aspects of LaunchDarkly’s architecture that allow us to seamlessly scale alongside our customers, and that will power us through the next trillion!
Business transformations are daunting tasks and introduce tremendous stress to the organization at multiple levels. This stress can lead to reduced productivity, bureaucratic processes, and reduced accountability, in addition to reducing the psychological safety of team members.
In this session, Dr. Gautham Pallapa talks about his personal transformative journey, focusing on the necessity and benefits of empathy in business transformations. He shares his experience with stress, and how he has combated stress in various workforce settings. Pallapa also shares approaches and experiments that he conducted to increase empathy in enterprises and with partners that he is helping on their transformation journeys.
The principles that allow for Continuous Deployment have been shown to increase product speed and quality. According to DevOps Research and Assessment's State of DevOps Report, high quality, shipping frequency, and mean time to recover (MTTR) are correlated. Those who use Continuous Deployment not only ship frequently but have higher overall software quality and fix bugs faster. One aspect of a good Continuous Deployment pipeline is testing. We want to test all branches of the code, we want to test in production, and we want to test our products efficiently.
There is a myth that Experimentation or A/B testing is only useful for small, incremental changes, and that we are only here to mine for winners. This is not true.
Whether we're trying to optimize the total visitors to an application or the value of the code we just deployed - if we take the time to understand and learn, then we stand to gain more than we lose. In the next 10 minutes, we’ll explore 4 key fundamentals to effective experiment design.
Betway has been following a "test in production" approach to building software for a few years. They test in production for two primary reasons: to validate business hypotheses and gain confidence in technical implementations. In this session, Michael will share tips, experiences, and lessons learned from testing in production.
On-call schedules, constant reorganizations, fast turnover, impossible deadlines, new platform and new frameworks every day… Building modern platforms is becoming unsustainable for the new generation of engineers. In these challenging scenarios, autonomous systems offer a new hope. Not requiring human intervention, these systems offer a more manageable and sustainable approach.
Cloud computing, standardized APIs, common practices, and shared tools are opening the door to building self-healing systems – systems that can react quickly and automatically to the increase in load, recover from unexpected errors even before users can notice it, and run rollbacks and migrations without any human interactions.
While all of this is technically possible, specific practices (feature flagging and error monitoring) and strong discipline are needed to build such systems. And above all, a new engineering culture needs to be defined to make the dream reality.
Gottfried Leibniz believed the only two numbers humans needed were one and zero. He refined the binary number system and paved the way for the machines we work on every day. For ages, one was the only number humans had. And it wasn’t until the concept of zero – a placeholder more than a number – was invented that humans made massive jumps in mathematical thinking.
But perhaps one and zero have more to teach us than the bits in our machines. Perhaps this story of two numbers has more to do with the creation of something from nothing. And how you, as a single engineer, can have massive impact on your systems and organization. In this talk, Emily Freeman tells the evolutionary story of early math, dives into the different styles of changemakers throughout history, and how any movement in tech starts with a single engineer.
The world’s top-performing organizations have embraced Progressive Delivery to gain a competitive advantage and exceed their customers' expectations. Hear from LaunchDarkly co-founders on how our latest innovations will empower all teams–not just Development–to build, operate, and learn from their software.
Effective Feature management tools give your team the ultimate control required to plan, manage, and safeguard all of your releases. With LaunchDarkly's Feature Workflows, we are introducing 3 new core capabilities -- Scheduling, Approvals, and Metric checks, that further extend the power of feature flagging to support complex workflows. Learn more about how our teams at LaunchDarkly plan to use feature workflows to better manage our product for our customers.
What’s your Launch Day Workflow? Have you automated anything to improve that workflow? Do you use any LaunchDarkly integrations or have you built any yourself?
This session will dive deep into what happens before and after you flip that toggle. We’ll explore integrations we’ve built to help you keep track of that feature once it’s out the door. And we’ll also discuss workflows and integrations that we’re working on to improve the quality, accountability, and compliance of rolling out features.
Feature flags make it easy to experiment in production, but they can also be a source of confusion during debugging, profiling, and root-cause analysis. Observability tools like distributed tracing offer exciting new approaches to understanding the performance impact of your features in production. In this talk, you’ll learn about how open source observability tools, such as OpenTelemetry, can be integrated into your services to provide critical performance data and insights about how your features are performing in different scenarios. You’ll also learn how to easily get started adding OpenTelemetry into your services to monitor your flags.
In this talk, you'll learn about the architecture and operational practices used by an engineering team of less than a dozen people to run a real-time event analytics platform that persists billions of events per day, with search over the telemetry performed near-real time. Liz will cover Honeycomb's use of Apache Kafka, Terraform, LaunchDarkly, Chef, EC2 auto-scaling, Lambda, S3, and how they use boring VMs rather than Kubernetes for their main serving pipeline.
You'll learn how streaming data solutions can work at less than Facebook scale, and with few engineers. Yes, even your frontend-y full stack devs can be on call; they just need the right tooling and help. Design systems to be robust with degrees of redundancy. And check them!
sweetgreen is quickly transforming into a food brand enabled by tech. As part of our mission to connect people to real food, we are one of the first food brands to build a delivery channel on our own platform. When initially thinking through the release process, not only did we want to position the native delivery feature for success, but also every subsequent, "big-impact" feature. Ultimately, our delivery team decided to leverage feature flags through LaunchDarkly, in order to support a canary rollout and take advantage of numerous other benefits arising from continuous delivery.
As we started to build out the delivery feature, our team had a variety of business and technical requirements, from geographical user segmentation to a flexible development workflow. Through feature flags, we were able to satisfy these requirements and power a beta launch in Los Angeles. Over December 2019, our beta launch allowed us to gather important customer feedback and implement improvements, which paved the way for a successful fleet-wide roll out in January.
Today, our delivery channel is an important contributor towards our overall business, allowing us to connect people to real food, wherever they are. Through building the delivery feature, our organization embedded important practices around feature flags and canary releases, which will support sweetgreen for years to come.
Engineering Pathways for Systemic Change – What Happens When You Apply Coding Principles to Social Issues?
In this talk, I will connect coding principles to social issues. You will learn how to leverage your knowledge of architecting and designing models for systems, and apply your experience to social issues that plague our country. I will bridge my professional experience in tech with those of the systems impacting people in our communities.
James Governor, Principal Analyst and Co-Founder, Redmonk
Adam Zimman, VP of Platform & Partnerships, LaunchDarkly
Progressive Delivery is the latest evolution in software delivery life cycles. James and Adam will share how the term came to be and what was driving the need for something new. They will present a shared vision on where things are headed and how teams can start to gain the benefits. Learn about the new tools, better ways to use data, and how to empower more individuals and teams in the software delivery process.