September 2020 Summaries

Harness has launched the "Heroes of Continuous Delivery" program to recognize professionals excelling in enabling continuous delivery (CD) within their organizations. This initiative celebrates their achievements and insights to inspire the broader community, emphasizing empowerment through autonomy and collaboration. Harness, a CI/CD platform, simplifies software delivery processes, allowing developers to innovate efficiently. The 2020 program highlighted eight industry leaders, including professionals from companies like GoSpotCheck, Discover, and LogMeIn, who have successfully implemented CD practices, ensuring scalability and security. Selected heroes receive various recognitions, including branded merchandise and promotions on Harness's platforms, and are featured in a Continuous Delivery ebook. The program also involves artistic collaborations to visually represent the benefits of collaboration and improvement, with contributions from artists like Ashton Rodenhiser. Harness encourages nominations for future recognition while offering a free trial of its Continuous Delivery Platform to explore how these professionals utilize it.

Harness has introduced a trial for its Cloud Cost Management feature, allowing users to gain insights into their cloud and cluster spending, complemented by platform enhancements such as Google Cloud Build support and custom deployment types. The trial is part of Harness's mission to democratize cloud cost management, making it accessible for users to optimize their cloud expenses. Alongside these updates, Harness has made significant platform improvements, including the ability to manage Custom Resource Definitions (CRDs) and implement roll-back and steady-state checks. In response to the COVID-19 pandemic, Harness conducted a survey of 500 developers, revealing that 74% of organizations plan to maintain remote teams post-pandemic. Additionally, Harness is hosting its first industry conference, {Unscripted}, on October 21-22, which promises an engaging lineup of speakers. The company has also received notable media coverage and is participating in various virtual events, continuing its commitment to enhancing software delivery and developer satisfaction.

A/B testing, also known as split testing, is a method used by companies to test different versions of a web page or product feature against the original to determine which version enhances user engagement and boosts key business metrics. This process involves gathering baseline data, formulating hypotheses, and using feature flags to manage and control the distribution of experimental features to targeted user segments. The implementation of feature flags facilitates the segmentation of users and the seamless rollout of experiments, allowing teams to monitor the impact on metrics through their analytics platforms. Advanced testing methods like A/B/n and multivariate testing enable the examination of multiple variants or combinations to optimize conversion rates and user experience, though they may require more time to achieve statistically significant results. Successful A/B testing not only improves user experiences and increases revenue but also enhances decision-making by correlating experimental data with company KPIs, making it a powerful tool for driving business impact.

Feature flags are a crucial tool in modern software development, allowing for controlled rollouts, production testing, and safer deployments by decoupling code delivery from feature releases. They provide flexibility and innovation throughout the software lifecycle, enabling developers to conduct gradual rollouts, centralize control, and simplify testing by allowing features to be toggled on or off without redeploying code. This approach facilitates testing in production environments, offers a kill switch for immediate issue resolution, and separates code deployment from feature rollout, thus enhancing localization, internationalization, and migration processes. Feature flags also support automation and experimentation, allowing applications to respond dynamically to changes in system performance and enabling robust A/B testing to drive business impact. Platforms like Split offer a comprehensive feature management solution that integrates these capabilities, helping teams deploy features confidently and efficiently while maintaining high standards of safety and performance.

Migrating from monolithic applications to microservices using feature flags is highlighted as a strategic approach to minimize risks associated with deployment and enhance application stability. Monolithic applications, where the codebase is unified, often pose problems such as cascading failures when a single feature malfunctions, impacting user experience and complicating backend management. Microservices, by contrast, fragment the application into smaller, independent units, reducing the risk of a single point of failure. The transition to microservices is traditionally risky due to the complexity of simultaneous upgrades, but feature flags offer a solution by allowing conditional code activation without full deployment, facilitating continuous delivery and safer canary releases. The process involves identifying an "edge" functionality, developing a microservice for it, and using feature flags to control its activation, enabling gradual user exposure and easy rollback if issues arise. Once the microservice proves stable, the monolith code can be removed, ensuring a seamless transition. Feature flags enhance this migration by providing real-time performance tracking and an immediate deactivation option, promoting a low-risk, efficient development cycle.