September 2023 Summaries

Harness Smart Feature Flags enhance software reliability by enabling controlled feature rollouts, real-time monitoring, and seamless rollbacks, ensuring high velocity and robust performance for a smooth user experience and operational resilience. In a landscape where digital transformation demands reliable software as a cornerstone for growth and customer trust, feature flags provide developers with a strategic toolset that goes beyond simple on/off toggles. By integrating reliability considerations early in the Software Development Life Cycle, developers can release software confidently, manage feature flags with a Git-like experience, and employ automated lifecycle management. The case study of Entur showcases how their deployment frequency increased significantly while reducing failed change rates, thanks to Harness's approach. Site Reliability Engineering (SRE) complements this by fostering a culture of continuous learning and adaptation, with smart feature flags serving as a pivotal component in managing the complexities of fast-paced software delivery environments. By facilitating adaptive deployments, these flags help mitigate potential issues, preserve the integrity of live environments, and enhance user experiences, ultimately leading towards autonomous ecosystems where reliability is seamlessly integrated into the development process.

Rollback plans are essential components of the software development process, serving as a contingency strategy to revert deployments to a previous state when issues arise, thereby minimizing disruption and downtime. Feature flags, also known as feature toggles, enhance rollback plans by allowing developers to quickly toggle specific features on or off, providing granular control that minimizes the impact on the overall system and automates parts of the rollback process. This integration not only streamlines rollback strategies but also reduces manual intervention and increases the efficiency and reliability of software deployments. Platforms like Split offer robust feature flag solutions that facilitate effective rollback plans by enabling developers to test features in controlled environments, gather real-time feedback, and execute rollbacks swiftly if necessary. By incorporating feature flags into rollback strategies, teams can ensure smoother operations, protect brand reputation, and enhance their overall development process, making feature flags a crucial tool for managing risk and ensuring stable software releases.

Trunk-based development (TBD) is a branching strategy that enhances software efficiency by minimizing merge conflicts and ensuring continuous integration. Unlike strategies like GitFlow, which can lead to long-lived feature branches and "merge hell," TBD emphasizes keeping code close to the main branch with small, short-lived branches, thus facilitating frequent updates and reducing rebase issues. Feature flags are used to manage feature rollouts safely, allowing unfinished code to be hidden without impacting the release of new features. The main branch or "trunk" serves as the source of truth, with all branches stemming from it and merging back quickly after approval. The strategy supports a culture of continuous integration, making it an attractive option for teams looking to streamline their development process without the complications of extensive merge conflicts. Integrating feature flags with TBD further reduces the complexity of releases by preventing unfinished code from affecting the final product, making it a compelling choice for managing new projects.

Harness Delegates and GitOps Agents offer a streamlined, scalable solution for executing commands remotely across multiple cloud providers and Kubernetes clusters, enhancing operational efficiency and simplifying multi-cloud deployments. By using these tools, organizations can avoid the complexities of local tooling and CLI installations, instead modeling their commands for remote execution through a containerized worker node that can be deployed anywhere on their infrastructure. These agents enable secure, efficient management of tasks across diverse cloud environments, supporting both direct and scoped access to Kubernetes clusters. The GitOps Agent, in particular, facilitates complex multi-cluster deployments by synchronizing manifests across clusters, offering a management layer that can integrate with existing ArgoCD instances or operate independently. Harness's approach allows for flexible scaling and separation of duties, accommodating different team needs while minimizing the administrative overhead associated with managing distributed systems.

Harness has introduced four new modules—Code Repository, Internal Developer Portal, Infrastructure as Code Management, and Software Supply Chain Assurance—designed to enhance efficiency, collaboration, and security across the software delivery lifecycle. The Code Repository module, based on the open-source platform Gitness, offers improved governance and seamless integrations, while the Internal Developer Portal centralizes documentation and accelerates onboarding. Infrastructure as Code Management provides advanced automation and security features, and the Software Supply Chain Assurance module enhances open-source governance and artifact integrity. Additionally, Harness announced updates to existing modules, such as support for GitOps with Flux and automated feature flag management, and launched two communities: the Engineering Excellence Collective™ and FinOps Excellence, aimed at addressing industry-wide challenges. These innovations reflect Harness's commitment to transforming software delivery through intelligent AI automation.

Kubernetes has become the industry standard for container orchestration, offering robust features for managing and scaling applications, but its full potential can only be realized by adhering to certain best practices. It is essential to define clear objectives and scope before adopting Kubernetes, ensuring that goals like improving scalability and enhancing deployment flexibility are achieved. Designing for scalability and resilience is crucial, involving the use of Kubernetes features such as pods, replicasets, and deployments, along with implementing horizontal pod autoscaling for workload efficiency. Effective resource management can be achieved through the use of namespaces, resource limits, and requests to optimize utilization and prevent contention. Regular updates and patches are necessary to stay protected against vulnerabilities, while best practices for container images, including using lightweight base images and image scanning, help maintain security. Implementing robust logging and monitoring with tools like Prometheus and Grafana, as well as regular backups and disaster recovery testing, ensure observability and resilience. Embracing GitOps and Infrastructure as Code simplifies operations, and investing in team training and comprehensive documentation fosters knowledge sharing and efficient Kubernetes management. The Split Feature Data Platform™ complements these efforts by enabling safe feature deployment and experimentation through feature flags, aiming to speed up releases and enhance team productivity.

Harness has launched the Engineering Excellence Collective, an initiative uniting 300 senior engineering leaders from global organizations to develop a comprehensive Engineering Excellence Maturity Model aimed at enhancing engineering practices. This model outlines 11 critical pillars, such as development process and incident management, to drive operational excellence in software development. It supplements existing frameworks like DORA by providing specific guidance on enhancing engineering capabilities. The initiative includes a survey assessment tool that allows teams to evaluate their practices and compare their scores with industry benchmarks. Harness's effort is to foster a collaborative environment for engineering leaders, enabling them to learn from each other and codify best practices, thereby promoting innovation and growth in the industry.

Harness Software Supply Chain Assurance (SSCA) is a new platform that enhances DevSecOps practices by extending security measures beyond application code to the entire software supply chain. It focuses on ensuring software integrity and open-source governance through the generation and validation of Software Bills of Materials (SBOMs) and compliance with SLSA (Supply-chain Levels for Software Artifacts) and Executive Order 14028. SSCA enables organizations to monitor and control third-party artifacts and open-source components, providing deep visibility and the ability to enforce policies on their usage. It supports SBOM orchestration in both CycloneDX and SPDX formats, and offers policy management capabilities to restrict or permit components based on defined criteria. The platform helps maintain software integrity by generating SLSA provenance and verifying it throughout the build pipeline, thus preventing the deployment of risky components and ensuring compliance with security standards.

Harness has introduced seven new AI-driven features for its AI Development Assistant, AIDA™, aimed at enhancing efficiency and reducing developer toil across the software delivery lifecycle. These features include semantic search, automated onboarding for feature flags, real-time support, advanced build and deploy troubleshooting, dynamic OPA policy generation, ChaosGuard™ for security governance, and custom dashboard configuration. AIDA leverages generative AI and natural language processing to streamline processes and empower developers by simplifying tasks such as code discovery, feature flag integration, and policy writing. The platform's capabilities aim to eliminate manual efforts, enhance productivity, and foster innovation in development workflows, setting a new benchmark for developer empowerment and efficiency in the tech industry.

Trunk-based development is a software development model that emphasizes frequent code commits to a single mainline, known as the trunk, which serves as the single source of truth. This approach reduces merge conflicts, accelerates release cycles, and enhances productivity through continuous integration, automated testing, and rapid deployment. Unlike other models such as Gitflow or feature branching, trunk-based development encourages developers to integrate code changes into the trunk multiple times daily, using short-lived branches only when necessary. This method is favored for its ability to maintain a high-quality, manageable codebase by ensuring all team members work from the same central repository, facilitating quick feedback, early error detection, and more frequent releases. Best practices for implementing trunk-based development include fostering a culture of continuous integration, automating testing processes, maintaining effective code review practices, and managing feature toggles to integrate code for incomplete features without exposing them prematurely. The approach supports efficient workflows, quick time-to-market, and improved coding practices, making it suitable for teams of any size or project complexity. Additionally, tools like Harness offer feature management and experimentation capabilities to further enhance trunk-based development by providing control over feature deployments and enabling A/B testing.

Harness, a prominent Modern Software Delivery Platform® company, announced four new product modules designed to enhance software delivery and developer experience at the { unscripted } conference in San Francisco. The new offerings include the Harness Code Repository, which is a premium module built on the open-source Gitness platform, the Harness Internal Developer Portal (IDP) for streamlining developer experience and onboarding, the Harness Infrastructure as Code Management (IaCM) for efficient and secure infrastructure management, and the Harness Software Supply Chain Assurance (SSCA) for enhancing security in software supply chains. These modules aim to address key challenges in software development by improving collaboration, governance, security, and operational efficiency while reducing risks associated with manual processes. Additionally, Harness unveiled two industry-wide communities, FinOps Excellence and Engineering Excellence Collective™, to foster collaboration on cloud cost management and engineering challenges.

Harness has released its Internal Developer Portal (IDP) in public preview, following a successful beta phase, to enhance developer self-service and streamline service management. Built on Backstage, the portal features robust security measures, powerful pipelines, and an intuitive software catalog to accelerate development workflows. Key benefits include improved onboarding times through developer self-service flows and reduced cognitive load via a comprehensive software catalog. The IDP also offers scorecards for measuring service maturity, helping platform engineers enforce best practices and track progress. As part of its evolution, Harness plans to introduce custom plugins, auto-discovery of services, improved self-service flows, and deeper integration with the Harness platform. The company emphasizes collaboration with early adopters and ongoing innovation, inviting users to influence its roadmap and explore its capabilities further.

Harness has unveiled Gitness, a fully open-source Git platform designed to transform software development by integrating source code management and CI pipelines into a single, cohesive platform. Announced at the {unscripted} conference in San Francisco, Gitness aims to overcome the challenges of existing software development tools, such as high costs, feature overload, and integration complexities, by offering a streamlined, secure, and efficient solution. The platform enhances collaboration through automated code reviews and pull requests, bolsters security with branch protection and user access management, and improves efficiency with AI-powered code search and automated CI pipelines. Notably, Gitness facilitates an easy migration process for existing repositories from platforms like GitHub and GitLab, enabling quick adoption and immediate benefits. In addition to Gitness, Harness introduced the Harness Code Repository, a premium module with enterprise-focused features like governance and seamless integrations, set to enter beta testing soon. Jyoti Bansal, CEO and co-founder of Harness, highlighted Gitness as a pivotal development for the software community, underscoring its ability to streamline workflows and enhance code quality. Harness, headquartered in San Francisco, emphasizes intelligent AI automation to expedite and secure software delivery, providing engineering and DevOps teams with a reliable way to deploy applications.

Gitness is an innovative open-source Git solution designed to enhance the development process by integrating code hosting with automated DevOps pipelines, offering a comprehensive platform for writing and delivering code efficiently and securely. It aims to streamline collaboration through automated status checks and mandatory reviews, improve code navigation with AI-powered semantic search, and simplify the creation and management of automated pipelines using various plugin steps. Gitness emphasizes security with branch protection and user access management, ensuring robust code management. It is user-friendly, with an easy setup via a Docker command and supports seamless migration from platforms like GitLab and GitHub, converting existing pipelines to Gitness YAML for immediate use.

Harness has launched a new module called "Infrastructure as Code Management" (IaCM), currently in its Beta phase, to address challenges associated with using infrastructure as code (IaC) solutions like Terraform, particularly at scale. This module aims to streamline infrastructure changes through advanced pipelines that integrate with tools such as Checkov and tfsec, provide resource visibility, and offer cost estimation capabilities. The module also includes features like an approval step for reviewing changes, PR automation for integrating changes into the git workflow, and OPA rules for policy compliance. Additionally, it offers managed state management to ease the burden of state file handling. Existing Harness customers benefit from the ease of integrating IaCM with their current setup, leveraging existing components like pipelines and connectors. Harness plans to expand support to other IaC solutions and is committed to supporting open-source versions of Terraform and OpenTofu.

Harness has expanded its certification program, offering three levels—Developer, Administrator, and Architect—across its Continuous Integration (CI), Continuous Deployment (CD), and GitOps modules, to help professionals validate and enhance their skills through structured and hands-on learning. The certification path is designed to accommodate various skill levels, beginning with the Developer level, which focuses on mastering module usage and is available for free. The Administrator and Architect levels require a deeper understanding of module administration and strategic implementation, respectively, and each includes both a theoretical and practical exam component. These exams are accessible asynchronously, allowing candidates to complete them at their convenience. In addition to the certification program, Harness provides learning materials on its Developer Hub, where users can explore study guides and register for exams, with successful candidates receiving a digital Credly Badge. Harness aims to continue expanding its certification offerings to cover all its modules, encouraging continuous learning and skill development in CI, CD, and GitOps practices.

Quizlet, a global learning platform, faced challenges in managing cloud costs and empowering engineers to act on cost-saving recommendations. To address this, the company implemented Harness Cloud Cost Management and developed an automated workflow that creates pull requests (PRs) for cost-saving recommendations related to their microservices, integrating these with Slack for better visibility and engagement. This approach increased the engagement rate from 0% to 75-80% among engineering teams, resulting in an average 40% reduction in Kubernetes workload costs. The automated system uses Google Cloud Scheduler and Functions to pull recommendations from the Harness API, edit configuration files in GitHub, and notify service owners via Slack. The process also includes a well-defined repository structure and utilizes Helm and Kustomize to manage configurations, allowing for efficient navigation and editing of files. To ensure PRs are assigned correctly, the CODEOWNERS file is maintained and updated as needed. The initiative has provided significant cost savings and improved the efficiency of cloud cost management at Quizlet.

Harness CD & GitOps has introduced a "Get Started" UI aimed at simplifying the onboarding process for new users by offering guided tutorials and automated setup through the Harness CLI. This new interface, available to users across all plans, utilizes sample applications from a repository forked from the popular argocd-example-apps repo, easing the learning curve by eliminating the need to understand new sample apps. Harness CD & GitOps supports a wide array of deployment types, including Kubernetes, serverless, and traditional applications, automating complex application deployments at scale while adhering to organizational governance standards. The platform's comprehensive features might appear daunting to beginners, but the new UI helps demystify continuous delivery and deployment practices. The introduction of this tool marks a step towards making Harness accessible to new users, with future enhancements expected to incorporate GitOps deployments and advanced pipeline features, ultimately aiming to transform beginners into proficient users capable of implementing continuous deployment efficiently and with robust governance measures.

Harness has introduced a new command-line interface (CLI) in public preview to streamline the onboarding process for new users of its platform, addressing the complexities found in its UI, API, and Terraform Provider. The CLI, which is open-source under the MIT license, aims to simplify access to Harness resources by offering a more intuitive and developer-friendly approach that can be integrated within familiar development environments. It manages essential resources for the CD & GitOps module and is inspired by popular cloud-native CLIs, allowing inputs via YAML files with command-line overrides. The CLI, available for Mac, Linux, and Windows, is easy to install and upgrade, with comprehensive documentation and tutorials to assist users. It leverages the existing Harness API for operations and facilitates both push-based CD pipeline deployments and pull-based GitOps deployments, with plans to expand its capabilities across all Harness modules based on user feedback.

Chaos Carnival 2024, organized by Harness, is a prominent virtual conference uniting global experts and innovators in chaos engineering, aiming to enhance software resilience and foster a community of forward-thinking developers. With over 30 talks and workshops, the event offers tailored tracks for beginners and experts, featuring insights from industry leaders like AWS and Google, and highlights the practical application of chaos engineering to navigate today's unpredictable digital landscape. The conference emphasizes community engagement through opportunities to participate as speakers, contribute to discussions, and submit papers, while also offering sponsorship opportunities for brands looking to align with cutting-edge innovation. The event has gained traction since its inception in 2020, drawing over 3,200 participants last year, and promises to deliver transformative insights through keynotes, real-world use cases, and hands-on learning experiences.

The study conducted by Harness SEI revealed that integrating GitHub Copilot into the development workflow significantly enhances developer productivity by increasing collaboration and expediting development cycles. GitHub Copilot, powered by AI models from GitHub and OpenAI, assists developers by generating code snippets and providing coding insights, resulting in a 10.6% increase in pull requests and a reduction of cycle time by 3.5 hours on average. Prior to adopting GitHub Copilot, development teams faced challenges with limited pull request activity and extended cycle times, leading to inefficiencies in collaboration and delayed feature releases. The study's findings underscore the transformative potential of AI-powered tools like GitHub Copilot in software development, highlighting their role in improving coding quality and efficiency. Further experiments are planned to explore the broader impact of GitHub Copilot, including its effects on code quality and other heterogeneous outcomes.

Managing the Software Bill of Materials (SBOM) lifecycle is crucial in ensuring transparency, compliance, and security in software development, especially in the context of increasing software supply chain attacks and regulatory requirements like Executive Order 14028. SBOMs offer a machine-readable inventory of software components, providing essential transparency and security by identifying libraries, modules, and dependencies along with their metadata. Prominent SBOM standards include CycloneDX and SPDX, which facilitate compliance and security management. The Software Supply Chain Assurance (SSCA) module enhances SBOM management by offering orchestration capabilities, policy enforcement, and seamless integration with CI/CD pipelines, ultimately supporting organizations in maintaining robust cybersecurity measures and adhering to regulatory demands.

In the context of modern software development, trunk-based and feature-based (GitFlow) development strategies offer distinct approaches, each with unique advantages and challenges that impact Continuous Integration and Continuous Deployment (CI/CD) practices. Feature-based development involves creating long-lived branches for new features, allowing developers to work in isolation, which is beneficial for managing multiple software versions but can lead to complex merges and outdated branches if not regularly integrated. In contrast, trunk-based development promotes frequent integration into a single "trunk" branch, aligning well with CI/CD by ensuring rapid error detection and a consistent, deployable codebase, though it requires disciplined integration and robust automated testing to prevent instability. Feature flags complement trunk-based development by allowing features to be toggled on or off without code changes, facilitating safe deployments and incremental updates. The choice between these strategies should consider factors such as project requirements, team size, experience levels, and desired release frequency, with trunk-based development favoring rapid iterations and GitFlow offering structured management for larger or more diverse teams. Understanding these methodologies enables teams to optimize their development and delivery processes for maximum efficiency and quality.

HashiCorp's recent change from MPL to BSL licensing for its open-source projects, including Terraform and Vault, does not impact Harness CD’s integrations with these tools for provisioning, onboarding, and secret management. Harness CD automates the software release process, while Terraform and Vault handle infrastructure and secrets management, respectively, meaning their functionalities do not significantly overlap with Harness's offerings. Harness customers continue to use Terraform in pipelines by providing their own binaries and configurations, while the Harness Terraform Provider facilitates resource onboarding, and HashiCorp Vault integrates for secure secrets management during deployments. Harness remains committed to supporting these integrations and plans to further invest in enhancing compatibility with Terraform and Vault features.

Harness Cloud Asset Governance is a tool designed to optimize cloud resource management by automating the elimination of cloud waste, ensuring compliance, and facilitating innovation. Built on top of the open-source Cloud Custodian, it addresses common challenges such as lack of a GUI, reporting, and security governance, and operational overhead. By offering preconfigured governance-as-code rules and a natural language interface powered by Harness AI, it simplifies policy creation and implementation, enhancing visibility into cloud spend and efficiency. The platform supports all major cloud providers and provides out-of-the-box policies to optimize resources and set guardrails. Users have reported significant cost savings and improved governance, with features like "Dry Run" for testing rules and comprehensive action logs supporting informed decision-making.

Effective process management in scaling teams is crucial for achieving optimal business outcomes, requiring a balance between consistency and innovation. Processes are indispensable for growth but can become burdensome if not continuously adapted to match evolving business needs. The text outlines the "9 Stages of Process Failure," highlighting how processes can become obsolete and result in suboptimal outcomes if not reevaluated. The distinction between mismatched and amplifying processes is emphasized, with the latter having the potential to significantly enhance efficiency and output. Introducing new processes often triggers skepticism, especially if past processes were met with resistance and proved ineffective. The key to successful process management lies in continuous adaptability driven by actionable data, which helps determine whether a process is a 10X enhancer or a misfit. Tools like Harness Software Engineering Insights (SEI) can provide valuable insights for testing process effectiveness and ensuring teams remain adaptable and iterative in their approach.

The Harness Developer Hub has undergone significant updates to enhance user experience by incorporating customer feedback, resulting in clearer explanations and an intuitive user journey. The updates include the addition of new tutorials and guides for various Harness modules such as Continuous Integration, Continuous Delivery & GitOps, Cloud Cost Management, Service Reliability Management, and Security Testing Orchestration, each designed to facilitate hands-on learning and detailed understanding of specific use cases. Documentation has been reorganized, refreshed, and expanded to reduce redundancy and improve accessibility, with new topics covering areas like Asset Governance, cost optimization, service reliability, and security testing. The platform also offers Harness-Certified Expert Certifications, with expanded levels and hands-on portions to validate practical skills. Users are encouraged to regularly visit the Harness Developer Hub for the latest content and to contribute suggestions for improvements.

Harness Cloud Asset Governance introduces a framework for cloud cost optimization and compliance, drawing parallels between software development and air travel logistics. The text explores two approaches to organizing software repositories, akin to airline route models: the monorepo approach, similar to a hub-and-spoke system, centralizes all source files in one repository, offering ease of refactoring and centralized tools, but can struggle with scaling issues and tooling limitations. Conversely, the polyrepo model resembles a point-to-point system, with each application component version-controlled separately, allowing for isolation and faster, more targeted deployments, though it requires careful management of dependencies and coordination. Ultimately, the choice between monorepo and polyrepo depends on organizational needs, with monorepos suiting centralized core applications and polyrepos fitting federated service models. Harness provides tools to manage GitOps complexity, supporting both models while ensuring security and access control, and offers tutorials to aid teams in adopting continuous delivery practices.