July 2020 Summaries

Secret management in application deployments, especially with the rise of containers and dynamic services, has led Codefresh to introduce a new integration for managing secrets with Kubernetes. This integration allows users to access Kubernetes secrets and configmaps directly within Codefresh pipelines, offering an easy way to handle secret storage and rotation. Users can choose which secrets to use via existing Kubernetes clusters connected to Codefresh, thus providing an alternative to dedicated secret solutions or using Codefresh's shared configuration capabilities. This method enhances security by isolating secrets through standard Kubernetes service accounts, allowing organizations to restrict access based on roles and namespaces. Additionally, the Codefresh Runner enables users to run pipelines within their own clusters, ensuring that secrets remain secure without Codefresh having direct cluster access. This flexibility in secret management allows organizations to decide how to store and manage their secrets, with the feature being available to all Codefresh Enterprise accounts.

Codefresh has introduced the Codefresh Runner, a solution designed to address the needs of engineering teams that require on-premise CI/CD infrastructure while maintaining security and compliance. The Codefresh Runner allows teams to run pipelines on their own Kubernetes clusters, including private ones behind company firewalls, without compromising security as it communicates with the Codefresh UI through outgoing connections only. This setup ensures sensitive data is not exposed to the public internet, complying with strict security regulations. Additionally, the Codefresh Runner is a Kubernetes-native application, allowing operations teams to leverage existing Kubernetes knowledge for management, monitoring, and scaling. It supports multiple runners across different clusters and can integrate with specialized hardware resources, offering a unified user experience for pipelines regardless of their runtime environment. Initially available to select Enterprise customers, the revamped version of Codefresh Runner is now accessible to all Codefresh accounts, including the free tier, featuring a streamlined installation process and the ability to manage multiple clusters with a single installation.

Developers face increased challenges in debugging due to frequent deployments, micro-services, and global distribution, which complicate identifying issues in complex systems. Observability, as opposed to traditional monitoring, focuses on exploring unknown-unknowns by enabling developers to interactively query raw, high-cardinality data without pre-aggregation, allowing them to form and test hypotheses about their applications' behavior. This approach helps developers understand the reasons behind system behaviors, such as specific user requests failing under certain conditions, which cannot be predicted by dashboards created from prior knowledge. Platforms like Honeycomb provide the necessary tools for developers to achieve this level of insight, integrating seamlessly with CI/CD pipelines like Codefresh to ensure confident and frequent production deployments. Annotating observability platforms during deployment offers real-time insights into changes, ensuring that new code functions as intended.

The webinar discusses the deployment of services to Kubernetes using Helm and highlights the limitations encountered with Helm when deploying to multiple environments, developing code as a team, or automating in a CI/CD pipeline. It introduces Codefresh Pipelines enhanced with Helmfile to address these issues, offering a more organized and flexible approach to managing Helm code and values, compatible with both Helm 2 and 3. The tutorial walks through deploying a simple Hello World application to a Kubernetes cluster using Helm and Helmfile, emphasizing the advantages of using Helmfile for dependency management and templating. It explains the setup of Helm repositories, the use of Helmfile and Helm Hooks, and the integration of a Codefresh CI/CD pipeline for deployment automation. The discussion includes a step-by-step guide on setting up a GKE cluster, configuring Codefresh to communicate with the Kubernetes cluster, and creating and running a Codefresh pipeline, with an emphasis on the benefits of using Helmfile for managing Helm charts and preventing vendor lock-in.