April 2022 Summaries

Harness Service Reliability Management (SRM) is a tool that enables engineering and DevOps teams to balance the speed of delivering new features with maintaining software stability by defining, measuring, and tracking Service Level Indicators (SLIs), Service Level Objectives (SLOs), and Error Budgets. This approach ensures high software reliability without sacrificing delivery speed and incorporates Continuous Verification to detect anomalies early, reducing the costs of post-production fixes. In today's software-centric world, the need to deliver high-quality, reliable, and innovative IT solutions is critical, with quality ensuring the software meets performance expectations and reliability ensuring it maintains quality over time. Harness SRM helps in creating reliability guardrails within CI/CD pipelines, allowing teams to improve reliability and speed of delivery with confidence. By continuously monitoring and verifying software, organizations can prevent costly rework and potential revenue loss due to unplanned application downtime.

GitOps leverages Git as the central source of truth for cloud-native application deployments, enhancing developer experience through declarative infrastructure, automated change application, and self-healing capabilities. Harness GitOps-as-a-Service builds upon these principles by integrating enterprise-grade security and governance, facilitating scalable and efficient software delivery, and is based on the open-source Argo CD project. GitOps allows for a streamlined and standardized deployment process where desired application states are managed through version control, enabling automated, reliable, and fast deployments. The approach addresses challenges of traditional deployment methods by providing a single source of truth, reducing manual processes, and enhancing developer efficiency. Harness adds value by offering features like pull request pipelines, a centralized management interface, and a comprehensive audit log, helping to manage the complexities and scale of GitOps implementation across multiple environments and applications.

Harness is actively promoting gender diversity in the tech industry through initiatives such as a quarterly speaker series, a women's development summit, and networking events, all aimed at creating an inclusive environment and supporting women's career growth within the company. The company emphasizes the importance of Allyship, Collaboration, and Recognition and Visibility, setting specific milestones for success and allocating necessary resources to achieve their goals. The leadership team of the women@harness employee resource group (ERG) includes members like Karuna, Marie Antons, and Sophie Manum, who are committed to breaking biases and fostering a supportive community for women at Harness. They share personal insights on overcoming biases, the importance of self-awareness, and how embracing diverse perspectives can enhance the workplace. The team also encourages external collaboration and invites interested individuals to join their efforts, emphasizing the significance of elevating the conversation around gender diversity and inclusion in tech.

Harness, a modern software delivery platform company, has secured $230 million in Series D funding, valuing the company at $3.7 billion, with the round led by Norwest Venture Partners and participation from notable investors such as J.P. Morgan and Capital One Ventures. This funding will support the company's expansion and talent growth, following a year in which its annual revenue rate more than doubled and its workforce nearly tripled. Harness's platform, which aims to automate and streamline the software delivery lifecycle, has seen significant expansions, including the addition of new modules and features like Harness Service Reliability Management and Enterprise GitOps for Harness CD. The company is recognized for its contributions to open-source projects and has received accolades such as Forbes's Best Cloud Computing Companies and Glassdoor's Best Places to Work. Harness's mission is to enhance the developer experience by enabling faster, more secure, and reliable code delivery, leveraging machine learning to improve deployment quality and efficiency.

Modern Continuous Integration (CI) systems significantly improve software development by enhancing collaboration, frequent integration, and constant code review, which in turn boosts software quality, security, and development speed. Harness's CI/CD platform exemplifies these advancements by shortening feedback loops and simplifying complexity, thus increasing developer confidence and reducing risk. Historically, build systems have evolved from tools running locally on developer machines to sophisticated CI systems that integrate best practices for software development. The distinction between build systems and CI lies in their focus: while build systems are often isolated to a developer's environment, CI systems facilitate frequent code merges into a central repository where automated builds and tests occur. The evolution of CI, championed by figures like Kent Beck and Martin Fowler, prioritizes collaboration and frequent, automated integration to prevent merge conflicts and enhance software delivery. Modern CI platforms, like Harness, aim to create a seamless developer experience by providing a self-serve pipeline that adapts to both local and cloud environments, reducing the economic and logistical challenges associated with running extensive tests and processes.

Harness Infrastructure Provisioners optimize the deployment of Azure resources using ARM templates, facilitating dynamic, repeatable, and idempotent infrastructure setups while eliminating manual scripting and portal navigation. ARM templates, a feature of Azure Resource Manager (ARM), use declarative syntax to define and deploy infrastructure, offering benefits such as repeatability and idempotency, ensuring consistent environments across development and production with support for parallel deployments. These templates utilize JSON syntax and include parameters, variables, user-defined functions, resources, and outputs to manage deployment processes efficiently. Harness supports both inline and remote ARM provisioners, allowing templates to be stored locally or fetched from a Git repository, with the deployment process involving a Manager and Delegate to handle template rendering and execution. The blog details the deployment flow for both inline and remote templates, emphasizing the role of the Manager in resolving Harness variables, with future posts promising insights into variable usage and rollback support within the Harness framework.

Harness Chaos Engineering (CE) is a newly launched module within the Harness Software Delivery Platform designed to enhance system resilience by allowing teams to create and manage failure scenarios. This tool provides extensive chaos experiments that simulate real-world failures, helping DevOps and SRE teams identify and mitigate potential reliability issues in their deployments. By integrating with CI/CD pipelines and supporting a wide range of chaos experiments, Harness CE offers insights into system vulnerabilities, enabling proactive resilience-building efforts to prevent costly downtime. The platform ensures data security with enterprise-grade privacy controls and supports private chaos experiment repositories. Harness CE is distinguished by its comprehensive coverage of failure scenarios and its collaboration with the open-source project Litmus, fostering a growing library of chaos experiments. Additionally, it integrates with existing observability tools and supports self-hosted and air-gapped deployments, making it a robust solution for improving system reliability.

Correlating engineering releases with cloud costs can provide significant benefits for organizations, such as early bug detection, fostering a cost-centric mindset, and shortening feedback loops for quicker resolution of issues. When teams analyze cloud cost management and deployments together, they can identify cost-related anomalies that might indicate bugs, such as infinite loops in Amazon CloudWatch or excessive logging in Google Cloud Platform Stackdriver. This practice encourages developers to optimize cloud spending by understanding pricing models and the impact of engineering decisions on costs. Furthermore, integrating deployments with cost analysis helps narrow down the root causes of cost spikes, reducing business impact and enhancing user satisfaction. Harness, an end-to-end software delivery platform, facilitates this process by offering features like anomaly detection and custom dashboards that correlate cost spikes with deployments, simplifying the monitoring and optimization of cloud expenses.

The text explores the evolution and current state of open-source Continuous Integration (CI) tools, highlighting a transition from older systems like CruiseControl and Jenkins to more modern solutions such as GitLab, Tekton, and Harness CI. These contemporary tools are designed with container-native and Kubernetes-native capabilities, offering enhanced efficiency and scalability for CI processes. The piece emphasizes the importance of CI in automating builds and tests, which are triggered by events such as code check-ins, with the ultimate goal of producing a deployable artifact. The text also discusses the community support, flexibility, and challenges associated with these tools, noting that while Jenkins remains popular due to its extensive plugin ecosystem, it faces criticism for being outdated and cumbersome. In contrast, newer tools like Tekton and Harness CI provide standardized, declarative configurations and easy extensibility, appealing to both open-source projects and large enterprises. The article concludes by suggesting resources for those considering both open-source and paid CI tools, aiming to assist in decision-making for selecting the appropriate CI platform.

Harness CI has introduced beta support for AWS virtual machines (VMs), allowing CI builds to be executed on either Kubernetes clusters or AWS VMs, including support for Windows OS and Docker command execution. This enhancement significantly optimizes build times through a VM pooling strategy that pre-initializes Amazon EC2 instances. The design involves running each CI stage on a dedicated AWS VM, which is terminated after the build, eliminating the need for workspace cleanup and ensuring a fresh environment for every build. The implementation requires installing Drone Runner AWS on a delegate VM, which manages the lifecycle of AWS VMs and facilitates the execution of steps during the CI build process. The system is structured into three phases: initialization, step execution, and cleanup, all managed by delegate tasks. The VM pooling strategy aids in reducing startup times by maintaining a certain number of pre-initialized instances, thereby enhancing the efficiency of the CI process on AWS infrastructure.

Harness Intelligent Cloud AutoStopping offers a method for optimizing cloud costs by managing idle resources in non-production workloads on platforms like AWS, Azure, GCP, and Kubernetes clusters. By automatically detecting idle times and shutting down unused resources, it helps companies save significantly on their cloud computing expenses, potentially reducing costs by over 70% for non-production workloads. The system employs a custom load balancer, which works alongside Envoy and proprietary services, to route traffic and manage virtual machines efficiently. AutoStopping Rules can be configured to automatically restart resources upon traffic detection, thus minimizing manual intervention and further reducing costs. This technology is particularly beneficial for managing Google Compute Engine VMs, where users can set up custom load balancers and AutoStopping Rules to optimize resource usage and avoid cloud waste, ultimately contributing to better cloud budget management and overall cost efficiency.

GitOps is an emerging approach to continuous delivery that emphasizes automation, containerization, and orchestration technologies to manage production environments, aiming to deploy software in a reliable and repeatable manner. It leverages Git to store manifest files and Kubernetes for application deployment, allowing developers to focus on feature development rather than infrastructure management. This model enhances developer productivity, provides better visibility into application stacks, and supports faster deployment times through automation and a security-as-code strategy. For organizations new to GitOps, the recommended approach is to start small, gradually implementing GitOps practices such as Continuous Delivery and deployment automation, while ensuring a cultural shift within the organization to understand the significance and goals of this methodology. As GitOps gains popularity, tools like FluxCD, Argo CD, and Jenkins X facilitate its adoption, making it an attractive option for efficient software delivery.

Abigail Boakye's career journey from a computer science student to a Cloud and Unix System Administrator at Ecobank underscores the significance of perseverance, networking, and stepping out of one's comfort zone, especially for women in the tech industry facing biases. Initially aspiring to be an optometrist, Abigail transitioned to tech due to advice from friends and her mother's guidance, eventually finding her niche in computer science despite initial struggles with mathematics and operating systems. Her experience highlights the importance of networking, as she attributes much of her success to the help of friends and professional connections. Abigail's current role involves provisioning and managing servers, with aspirations to further delve into AWS and DevOps. She emphasizes that tech careers are not limited to coding, advising women to explore various tech roles and brace themselves for challenges. Abigail is also pursuing an AWS certification, driven by a deadline, and is eager to continue her education and professional growth. Her story is featured in the "Women of DevOps" series, which aims to amplify the voices of women in technology around the world.