October 2022 Summaries

The Harness Developer Hub (HDH) has been launched as a centralized platform to streamline and enhance the software delivery process for developers and engineers by consolidating technical materials, tutorials, and reference documentation. This open-source initiative encourages community contributions and offers a verb-based approach to common software delivery tasks such as deploying applications and building code. The HDH also provides resources on reliability and managing service level objectives, with module documentation gradually being incorporated. As a beta release, the platform is open to feedback and contributions through GitHub or the on-site feedback form, with plans to expand its offerings and align more closely with users' Harness Accounts as it evolves towards general availability.

Continuous Integration and Continuous Delivery (CI/CD) pipelines significantly enhance software delivery by reducing deployment time, improving governance and compliance through secrets management and role-based access control (RBAC), and elevating service quality with integrated testing and feedback mechanisms. While they accelerate time to market by cutting down the average 66-hour pipeline creation time and boosting operational efficiency, they also present challenges such as debugging failures and managing the increased complexity in microservices and cloud migrations. Effective CI/CD implementation requires attention to governance, risk management, and compliance (GRC), as well as ensuring service performance through robust testing and infrastructure management practices. Harness offers solutions to these challenges, providing features for auditing, visibility, and policy enforcement, thus enabling organizations to innovate and improve their software delivery processes efficiently.

Harness's summer internship program highlighted the contributions of Siddhi Bansal and Ram Potham, who worked on projects related to software delivery insights and AI/ML models for cloud cost anomaly detection, reflecting the company's emphasis on nurturing talent and driving innovation. Siddhi, a computer science major from the University of Texas at Austin, developed dashboards to help customers identify bottlenecks in their development cycles, gaining valuable insights into software development lifecycles and corporate environments. Ram, an artificial intelligence and computational biology major from Carnegie Mellon University, focused on building AI/ML models for forecasting and anomaly detection, underscoring the importance of managing cloud costs through proactive anomaly detection. Both interns appreciated Harness's supportive and collaborative atmosphere, which facilitated their professional growth and affirmed their career interests.

Eric Minick's article identifies common pitfalls that suggest a team might not be properly implementing continuous delivery (CD) practices and offers strategies to improve. It highlights issues such as lengthy deployment times, excessive planning, and reliance on dedicated release teams, which can hinder the efficiency and speed of software delivery. The article stresses the importance of committing to trunk, rapidly fixing build and deployment failures, and minimizing manual interventions in deployment pipelines. It also underscores the need for developers to deploy and debug their own code to enhance efficiency and highlights the importance of understanding the business impact of deployments. Minick clarifies that DevOps is a cultural approach, while CD is a set of practices aimed at delivering software safely and quickly, and he advises organizations to adopt CD principles to stay competitive.

Harness Continuous Integration (CI) has introduced new features including Cache Intelligence, enhanced Test Intelligence™, and Starter Templates aimed at reducing build times and simplifying the developer onboarding process, promising up to 90% faster build cycles. These enhancements are part of a fully managed cloud service offering that allows developers to create and run pipelines swiftly without the burden of managing infrastructure. By leveraging machine learning, Harness CI optimizes pipeline execution through strategic caching and concurrent test executions, ensuring developers spend more time coding rather than waiting for builds. The platform also offers a flexible pricing model with credits for build execution, starter templates for various programming languages to expedite onboarding, and an intelligent YAML editor for streamlined configuration. The system is built on the open-source Drone framework and supports running pipelines on multiple operating systems, including Mac OS, Windows, and Linux, catering to developers' diverse needs and enhancing the overall developer experience.

Integrating JFrog Artifactory with Harness CI enhances the software development lifecycle by facilitating the storage and reuse of build artifacts, thereby boosting developer productivity and expediting software delivery. JFrog Artifactory serves as a centralized repository management solution that allows developers to store, track, and share binaries and packages without needing to delve into their complexities. By using Artifactory in conjunction with Harness, a modern software delivery platform, developers can seamlessly test, build, and store artifacts in a collaborative environment. The integration process includes setting up Harness with Artifactory as a connector, configuring project settings, and utilizing Kubernetes for infrastructure management. This setup enables teams to streamline their workflows, ensuring that build artifacts are efficiently pushed to Artifactory, thus enabling faster and more reliable software delivery. The article provides a step-by-step guide to this integration, illustrating the process with a simple Java project and emphasizing the robust capabilities of both Harness and Artifactory in optimizing DevOps practices.

The article explores the integration of feature flags in Next.js applications to optimize deployment and feature release processes, highlighting challenges and solutions for maintaining high performance. It discusses the problem of delayed page startup due to network calls needed to retrieve feature flag settings and presents a solution using Split's SDK, which enables instant page loading and feature flag evaluation without additional network calls. By caching feature flag definitions in local storage during the initial server-side rendering, applications can avoid network delays, ensuring a fast and seamless experience for users. This approach not only enhances performance but also provides privacy benefits by keeping flag information on the client side, preventing interference from ad-blockers. Additionally, the article underscores the advantages of feature flags, such as decoupling deployments from releases, allowing gradual rollouts, and enabling precise control over feature activation and analysis of their impact, ultimately supporting continuous improvement and innovation.

Effective GitOps secrets management is crucial for secure software deployments, addressing challenges through strategies like encrypted secrets and external secrets managers to mitigate risks. GitOps, which uses a system's desired configuration stored in Git for software deployment, relies on Git repositories as the single source of truth, posing security risks if sensitive information like API keys and passwords are inadequately managed. Tools like Bitnami's Sealed Secrets and Mozilla's SOPS provide methods to encrypt and manage secrets, though they can be complex to scale and require manual intervention. Alternatively, using external secrets management systems, such as ExternalSecrets and Kubernetes Secrets Store CSI Driver, allows for safer integration of secrets into Kubernetes clusters by storing only references in Git. Despite these solutions, handling secrets securely in GitOps remains challenging, necessitating thoughtful strategy and tools like Harness GitOps-as-a-Service, which provides enterprise controls to ensure secure and compliant deployments with minimal engineering overhead.

Improving developer experience through modern CI/CD platforms is crucial for reducing toolchain maintenance time and addressing high turnover rates by minimizing unnecessary toil and consolidating tools. Many companies do not adequately focus on developer experience, resulting in dissatisfaction and high turnover, with 42% of developers considering leaving their jobs, according to a DigitalOcean survey. Homegrown toolchains, common with tools like Jenkins, can create fragmented systems and technical debt, leading to maintenance burdens and lost time. Developers often face significant challenges with poorly integrated deployment toolchains, which can detract from their productivity and job satisfaction. Implementing full CI/CD platforms can streamline toolchains, reduce toil, and improve the overall developer experience, ultimately aiding in recruitment and retention of talented developers.

Harness is actively participating in Hacktoberfest 2022 by supporting OpenFeature, a CNCF project focused on standardizing feature flag functionality to improve interoperability across programming languages. Hacktoberfest, founded by DigitalOcean in 2014, is an annual event encouraging open-source contributions, and Harness is contributing by promoting the development of OpenFeature providers for its feature flag SDKs. As a proponent of open-source software, Harness sponsors initiatives like the Open Source Initiative, CD Foundation, and CNCF, and offers several free tools under open-source and source-available licenses. During Hacktoberfest, Harness is providing resources and guidance to contributors through Slack discussions and GitHub repositories for languages such as Java, Go, .NET, and JavaScript, encouraging community engagement and collaboration in building OpenFeature providers.

Chaos engineering tools such as Chaos Mesh, Gremlin, and LitmusChaos are employed to enhance system resilience by simulating failures and testing responses, which helps identify weaknesses and proactively improve software reliability in cloud-native environments. With the shift towards cloud-native deployments, businesses face increased complexity and potential failure modes, making chaos engineering essential to minimize unplanned downtime that can severely impact business operations. Chaos engineering involves testing a software’s ability to handle failures without affecting functionality, allowing development teams to pinpoint vulnerabilities and improve system reliability. Tools like Chaos Mesh, Chaos Monkey, and Gremlin offer various features and capabilities to perform chaos experiments and ensure systems can withstand issues like network latency and infrastructure performance challenges. These tools are used by leading tech companies to better understand system behavior and improve reliability, with Chaos Mesh and LitmusChaos being open-source and Gremlin offered as a SaaS solution. Each tool has its unique strengths and limitations, offering different levels of configurability, automation, and integration into existing DevOps workflows to support continuous system validation and reliability testing.

Choosing the right feature management architecture is crucial for ensuring speed, privacy, and security at scale, with Split's rules engine approach offering local evaluation of feature flags within client applications, reducing the need to send sensitive data to cloud servers. This method accelerates decision-making, enhances resilience, and fortifies data protection, especially in high-stakes environments like banking and healthcare, by evaluating flags locally and retaining sensitive information within the application. Unlike traditional platforms where client-side SDKs serve as mere proxies necessitating cloud-based evaluations, Split's robust and intuitive SDKs function as autonomous rule engines, performing local evaluations swiftly and securely. The platform's architecture not only minimizes the risk of data breaches but also ensures real-time data processing and resilience through cached rulesets and multi-region hosting, making it suitable for mobile applications with unreliable connectivity. Split supports a wide range of programming languages and multi-tenancy, offering comprehensive feature management and experimentation solutions that facilitate faster feature releases with data-driven insights while maintaining a strong partnership with clients to foster continuous improvement and innovation.

Installing a Harness Delegate on Kubernetes using Helm simplifies the process of pipeline creation and execution, requiring minimal hardware resources and enhancing software development workflows. Delegates function like self-hosted runners and perform tasks assigned by the Harness Manager upon installation. The tutorial utilizes Helm, a package manager for Kubernetes, to install the Delegate by modifying the values.yaml file to fit specific resource requirements such as 2 vCPUs, 16 GB memory, and 50 GB disk space. The installation process involves ensuring necessary permissions on the Kubernetes cluster and linking the Helm repository to download and apply manifests. Successfully installing the Delegate allows users to create pipelines and improve their software development processes, with community support available through forums and Slack channels.

Helm is a package manager for Kubernetes that simplifies application deployment and management by packaging manifests into charts, allowing for consistent deployments across environments and facilitating rollbacks in case of errors. Initially launched by Deis in 2016 and now overseen by the Cloud Native Computing Foundation (CNCF), Helm has been influenced by other package managers like apt and Homebrew. The text describes how Helm, especially its third version, integrates with Harness to enhance deployment processes through seamless rollbacks to stable versions during failures. It highlights the complexity of manually deploying Kubernetes manifests and how Helm streamlines this by allowing users to define, install, and upgrade applications using Helm charts, which consolidate resources for easier management. Additionally, the article provides a step-by-step guide on deploying a Helm application using Harness, emphasizing the importance of understanding Helm charts and using the right configurations in the values.yaml file to ensure efficient and error-free deployments.

The text explores the integration of Split with Kubernetes to enhance A/B testing and feature management for containerized applications. Split offers three containerized applications designed for feature flagging and experimentation, which can be run on personal infrastructure using NodePort services for external communication. The document highlights the use of the Split Evaluator for languages without a Split SDK, the Split Proxy for improved SDK initialization in microservices environments, and the Split Synchronizer for coordinating data sharing to a remote datastore like Redis. It provides detailed instructions and sample YAML files for setting up these components, emphasizing the importance of proper sizing and scaling to handle production-level traffic efficiently. Additionally, the text underscores the value of feature flags in safely deploying features and conducting experiments, as well as the benefits of using the Split Feature Data Platform for comprehensive feature management.

Harness has launched a comprehensive Resource Center aimed at enhancing user support and improving product understanding by offering updates, documentation, community engagement, and direct feedback channels. Accessible from the help panel in the Harness dashboard, the Resource Center provides users and developers with valuable resources such as documentation, API docs, and release notes. It also facilitates community interaction through features like Community Slack and offers learning tracks via Harness University for a holistic view of the product. Users are encouraged to submit tickets for feature requests or issues and can track these through the platform. Additionally, Harness maintains an active community for sharing best practices and provides round-the-clock support to ensure a seamless user experience.

Harness Triggers automate CI/CD pipelines in response to Git events such as PUSH, CREATE, and ALL, enabling event-driven deployments and the practice of building or deploying to a target environment with every commit. This automation is achieved through predefined conditions and GitHub webhooks, which streamline the deployment process. Triggers act as a mechanism to initiate a pipeline build or deployment based on Git events, and are categorized into conditions related to Pull Request, Issue Comment, and Push events. These conditions can be further specified to trigger pipelines when certain patterns are matched, such as branch names or file changes in specific directories, which is especially useful in a monorepo setup. Harness utilizes the GitHub events API to configure webhooks, ensuring payloads are sent only for specified events, thus allowing for customized pipeline triggers based on Git events and payload conditions. Users can seek additional support and information from the Harness community forums or Slack channels dedicated to continuous delivery and integration.

Transitioning from multistep Dockerfile builds to a Drone CI pipeline simplifies application deployment, particularly in cloud-native environments. Kamesh Sampath uses a Next.js application to illustrate this process, highlighting that traditional methods of building and deploying applications may hinder optimal cloud-native results. By containerizing applications and utilizing CI tools like Drone, which is known for its simplicity and adaptability to cloud-native architectures, developers can efficiently manage the deployment process. The tutorial guides users through setting up a Drone pipeline, converting Dockerfile commands into Drone steps, and organizing these steps into a coherent deployment process. This approach not only streamlines building and deploying applications but also adheres to best practices in software delivery, such as leveraging a single responsibility principle and using continuous integration for scalability and reliability.