September 2022 Summaries

A good development environment should closely resemble the production environment to ensure consistent application behavior, yet achieving this can be challenging due to increased infrastructure costs and workflow complexities. Kubernetes offers significant benefits in development environments such as faster release cycles, improved collaboration, greater autonomy for development teams, and reduced downtime and bugs in production, but it also presents challenges like complexity, a steep learning curve, limited resources, configuration discrepancies, and vendor differences. While Kubernetes is advantageous for scalable and modern microservices-based applications, it might not be suitable for small teams or simpler applications. Solutions like Qovery can simplify the use of Kubernetes by managing clusters and providing features like clone environments and preview environments to streamline deployment and management processes.

An upcoming HUG meetup in Paris on October 11th will explore the present and future of Infrastructure as Code (IaC) featuring experts and community-driven lightning talks, deliberately avoiding sales pitches. The event, conducted in English, will delve into topics such as Pulumi, Terraform, CDK, and related areas, focusing on components, higher-level constructs, and developer experience. Attendees will have the opportunity to engage in a round table discussion with industry leaders, including Paul Stack from Pulumi Corporation, AWS Heroes Matthieu Napoli and Sebastian Korfmann, and moderator Anton Babenko, an influencer in Terraform. The informal gathering aims to foster curiosity and insights in DevOps and IaC, with snacks and networking opportunities provided by sponsor Couchbase. Registration is required for those interested in attending.

Qovery's DevOps automation tool offers advanced settings for users who require more detailed customization beyond basic parameters like RAM and storage, particularly for specific industries or large companies. These advanced features, available in the API V2 and the interface V3, include IP whitelisting, which restricts application access to specified IP addresses, and enhancements in application deployment such as setting build timeout intervals and disabling custom domain checks. Additionally, Kubernetes probes are available for conducting health checks; liveness probes restart failing containers, while readiness probes determine if a container is ready to handle traffic. As Qovery is preparing to release V3 in Beta, more features are anticipated, with the frontend team making significant progress toward the launch expected in early Q4.

Qovery is a cloud deployment platform designed to balance autonomy and control for developers and DevOps teams by providing an integrated user experience for managing cloud environments. It enables developers to swiftly test and release features by creating on-demand environments, while allowing DevOps teams to maintain control and ensure best practices in infrastructure provisioning. Qovery achieves this through an abstraction layer over infrastructure components such as Kubernetes and security groups, inspired by platforms like Heroku and Rancher. The system consists of a control plane that manages business logic and a Qovery Engine that handles infrastructure deployment, both of which work in tandem to ensure secure and efficient application deployment. Qovery integrates with popular development tools and services, and employs a zero-trust security model to safeguard infrastructure, promoting resilience and autonomy by functioning independently of the control plane once infrastructure is deployed.

Usage AI is a tool designed to help companies reduce their AWS cloud expenditures by efficiently managing Reserved Instances (RIs) with zero downtime and no need for code changes. By leveraging a Guaranteed Buyback Clause, Usage AI allows customers to achieve up to 57% savings on their AWS bills without committing to long-term AWS contracts. The tool operates by dynamically buying and selling RIs in response to a company's changing compute needs, ensuring optimal instance coverage while minimizing costs. This process is quick and straightforward, requiring only the setup of an IAM policy and role, and can be completed in under five minutes. Usage AI's effectiveness is demonstrated through case studies like FabFitFun, which reportedly saves $1 million annually. Offering a similar advantage to AWS Spot Instances, Usage AI focuses on RIs, providing consistent savings without the volatility of Spot prices. The service is available to any Qovery customer using AWS, and integration with Qovery’s platform aims to enhance the deployment of production-like environments while optimizing cloud spend.

DeveloperWeek Cloud Conference in Austin, Texas, is a significant international event for cloud computing where various stakeholders, including engineers, developers, and IT managers, gather to explore new innovations. The team behind the application Qovery attended as exhibitors for the first time and shared their experiences, noting the challenges and learnings from setting up their booth and distributing promotional materials. Despite anticipating around 3,000 attendees, actual turnout was lower than expected due to a perceived marketing oversight by the event organizers. However, the team still engaged with several notable companies like Cisco, Apple, and the US Government, showcasing Qovery's features such as Preview Environments and the Qovery Terraform Provider, which were well received for their potential to streamline DevOps processes in large organizations. The event, while not as crowded as hoped, offered valuable networking opportunities and insights, with plans for future participation in larger events like AWS Summits and KubeCon.

A staging environment serves as the final testing ground before changes are implemented in a production environment, resembling the production setup to ensure consistency and reliability. However, traditional staging environments are increasingly seen as inadequate for modern Continuous Integration/Continuous Deployment (CI/CD) practices due to their lack of isolation, inflexibility, and hindrance to team collaboration. To address these limitations, on-demand and preview environments have emerged as agile solutions that can be created and dismantled easily, allowing for isolated testing of features, fostering rapid collaboration, and reducing infrastructure costs. These environments enable developers to test new features without affecting others, incorporate feedback swiftly, and maintain control over expenses by using services like Qovery's Environment as a Service (EaaS), which enhances team velocity and shortens time-to-market by providing dynamic, full-fledged environments on a user's AWS account.

Papershift, a company specializing in workforce planning software, was founded by Florian after experiencing a business failure due to poor organization. Initially hosted on Heroku, Papershift faced escalating costs as they grew, prompting a transition to AWS and Cloud 66 for more flexibility. However, the solution's lack of autoscaling led to the adoption of containerization, which was managed by a single infrastructure engineer, resulting in limited troubleshooting capabilities for the team. To address this, Papershift transitioned to Qovery, a platform offering a combination of good UI and containerized infrastructure management, which facilitated a seamless migration completed in under a month. This move allowed the entire team, including engineers and QA, to efficiently manage deployments and testing, while ensuring effective autoscaling. The company is now looking forward to future developments, including a new front-end interface and additional features, while continuing to contribute feedback and suggestions to Qovery.

The text outlines common pitfalls encountered when deploying applications in Kubernetes, emphasizing the importance of correctly using labels and selectors, performing health checks, and avoiding the default namespace for all objects. It warns against relying on the "latest" tag for image versions, highlighting the need for explicit version tags to enable rollbacks. The document stresses the significance of robust monitoring and logging systems, recommending third-party tools for enhanced functionality beyond native Kubernetes capabilities. It also addresses the issues of incorrect container port mappings and the crashloopbackoff error, providing guidance for troubleshooting and prevention. The text concludes by introducing Qovery as a solution to simplify Kubernetes deployment on AWS, offering automated environments with built-in security and cost optimization features.