July 2023 Summaries

Harness has introduced the Harness Continuous Integration (CI) Migration Utility, a free, open-source tool designed to automate up to 80% of the migration work from various CI providers to Harness, thereby significantly reducing the time and effort involved in switching providers. The utility supports transitions from popular CI platforms such as GitLab, CircleCI, GitHub Actions, and others, streamlining what has traditionally been a cumbersome process. Harness CI, part of the Harness Modern Software Delivery Platform, accelerates builds by up to 4X and integrates seamlessly with any Git provider, offering versatile workload support and thousands of integrations to enhance developer productivity and enterprise readiness. The platform also provides robust features like Proprietary Test Intelligence, Cache Intelligence, and Docker Layer Caching to optimize build times, along with enterprise-ready capabilities such as Role-Based Access Control and Policies-as-Code. By consolidating the software development lifecycle under the Harness umbrella, the platform aims to increase collaboration, standardization, and efficiency in DevOps processes, while offering open-source contributions to further its development.

Harness CI enhances software delivery by integrating a wide range of open-source community automation tools, offering thousands of integrations and simplifying the process of creating CI pipelines. It allows teams to tap into the collective wisdom of the global developer community by using proven community plugins, such as Drone plugins, Bitrise steps, and GitHub Actions, within their CI workflows. This approach not only provides unmatched flexibility and choice but also streamlines pipeline authoring, enabling developers to focus more on coding. Harness CI makes it easy to transition from existing CI solutions while maintaining the use of favorite open-source integrations. Through its user-friendly features and the fastest CI solution available, Harness CI aims to supercharge the software delivery process, encouraging users to experience its benefits by signing up for a free plan or requesting a demo.

GitOps is a modern approach to software delivery that positions Git as the single source of truth for application deployment, offering both push and pull methodologies to enhance deployment reliability and efficiency. The push approach automates deployment pipelines triggered by code changes, while the pull approach employs agents to sync changes from Git repositories to ensure the desired and actual states of the system are aligned. This approach is gaining traction, particularly with Kubernetes deployments, due to its ability to simplify complex software delivery processes, streamline workflows, and enhance collaboration by integrating development and operations. Tools like ArgoCD and FluxCD have contributed to its growing popularity, and platforms like Harness offer practical solutions to implement GitOps strategies. As organizations continue to seek increased agility and scalability, GitOps is emerging as a key player in the evolving software delivery landscape, promising remarkable speed, reliability, and adaptability benefits.

Integrating Jira Releases with lead time metrics in Software Engineering Insights advances release management by offering detailed analysis, enhancing predictability, and supporting continuous improvement, thereby optimizing efficiency and enabling realistic project timelines. This integration provides a comprehensive view of the release process by allowing seamless lead time measurement for each release, helping identify bottlenecks that might impede delivery timelines. The enhanced capability offers in-depth insights into the time taken at each release stage, improving predictability and allowing for data-driven decisions to enhance efficiency. The combination of Jira Releases and lead time measurement fosters a mindset of continuous improvement by analyzing lead time data to identify patterns and inefficiencies. The insights from the release cycle are available through two reports: the Jira Releases Report, which gives an organizational overview, and the Lead Time Report, offering detailed information on the software development lifecycle workflow until release. Additionally, the upcoming feature to measure deployment frequency will further augment these capabilities, and configuration on SEI dashboards can be achieved through simple workflow profile modifications.

Harness has successfully utilized Looker's aggregate tables to optimize dashboard load times and reduce data processing costs for large customers, achieving efficiency gains of up to 1000 times. Aggregate tables, a type of Persistent Derived Table (PDT), allow the creation of smaller rollup tables that enhance data engineering flexibility and scalability by intelligently selecting the smallest table to satisfy queries. This approach significantly reduces the amount of data processed, resulting in faster dashboards and cost savings. Specifically, Harness applied this strategy to its Cloud Cost Management module, improving dashboard performance from minutes to seconds and reducing data processing to a fraction of its original size, thus providing a more efficient and cost-effective solution for its customers.

Harness has rolled out a series of updates across its product suite, enhancing capabilities in Continuous Delivery, Feature Flags, Chaos Engineering, and more to improve software delivery and user experience. Key updates include improved Git branch switching, enhanced Kubernetes deployments, and new features such as scheduled automatic approvals and expressions for looping strategy status. Feature Flags have received user interface improvements and SDK upgrades for multiple programming languages, while Cloud Cost Management and Service Reliability Management have seen user experience enhancements. Security Testing Orchestration now supports Typescript in Sonarqube scans and improved Checkmarx configurations. Chaos Engineering updates offer better infrastructure management and experiment reporting, while Software Engineering Insights now provide valuable information on Jira Releases. Platform enhancements include better Git branch management, security updates in username handling, and new capabilities in SMTP configuration. The Early Access program introduces an Internal Developer Portal to boost developer productivity through self-service portals and real-time software catalogs.

Engineering intelligence extends beyond mere metrics, requiring a combination of measurements and insights to achieve excellence. Common myths such as the belief that engineering intelligence is solely about metrics, that outcome metrics are sufficient, and that measuring developers equates to measuring productivity are debunked, emphasizing the importance of both input and outcome metrics as well as the insights derived from them. While metrics can be manipulated, frameworks often include counter metrics to ensure accuracy, and even imperfect data can provide valuable directional guidance. Surveys, although useful for capturing subjective insights, should be complemented by data sources to provide a comprehensive view. Overall, achieving engineering excellence demands an integrated approach that combines metrics, insights, and tools to measure both individual performance and overall productivity.

In the exploration of Test-Driven Development (TDD) and Feature Flag-Driven Development (FFDD), both methodologies present unique advantages and challenges that cater to different aspects of the software development lifecycle. TDD emphasizes writing tests before code to ensure high-quality, bug-free software, fostering a disciplined approach to development where testing is integral rather than an afterthought. This method encourages modular, maintainable code but can be challenging for teams new to the discipline. In contrast, FFDD offers flexibility by allowing features to be toggled on or off without redeploying code, enabling real-time testing and safe experimentation in production environments. This approach supports incremental releases and user-driven development but requires careful management to avoid complex codebases and technical debt. Additionally, the integration of artificial intelligence (AI) in development processes, including FFDD, enhances the ability to analyze feature performance and automate repetitive tasks, offering data-driven insights for better decision-making. Ultimately, the choice between TDD and FFDD depends on a project's specific needs, with many successful teams combining both methodologies to balance code quality and feature release agility.

GitOps, a paradigm inspired by factory automation, utilizes Kubernetes and Git to automate and scale software deployment processes, ensuring efficient and consistent results with enhanced traceability and collaboration. In this framework, Kubernetes acts as a goal-driven robot, executing desired state configurations, while Git serves as a centralized logbook for change management. Harness Pipelines extend GitOps by offering comprehensive end-to-end pipeline automation, integrating both human and automated processes, thus allowing for governance, compliance, and seamless orchestration across multiple environments. With built-in observability and monitoring, Harness provides a unified interface for improved management and collaboration, complementing tools like ArgoCD and Flux by offering a platform for pipeline orchestration and enhanced deployment capabilities. Overall, GitOps promotes traceability, scalability, reproducibility, and efficient collaboration, bringing the advantages of industrial-scale automation to the software delivery process.

The discussion explores the burgeoning role of AI in software delivery, highlighting its potential to automate repetitive tasks, enhance quality assurance, and personalize user experiences, thereby streamlining development cycles and providing intelligent analytics. It underscores the advantages, such as improved efficiency, predictive maintenance, and data-driven decision-making, while also addressing significant drawbacks like ethical concerns, data privacy issues, and the need for skilled professionals. The text emphasizes the necessity of balancing AI automation with human oversight to mitigate risks associated with overreliance and bias, and it suggests that organizations should consider the initial investment and infrastructure requirements carefully. Additionally, the importance of a robust AI strategy and responsible AI development practices is stressed to maximize benefits while minimizing potential pitfalls. The text concludes by introducing Split, a feature management and experimentation platform that aids in the safe introduction of AI features, offering tools for efficient deployment, risk reduction, and enhanced team collaboration.