September 2023 Summaries

Database migrations can be managed using GitOps principles, similar to application source code, by employing tools like Atlas and Kubernetes operators to automate and control database changes. The Atlas Kubernetes operator enables the GitOps approach by introducing Custom Resources Definitions (CRDs) that describe the actions required for database migrations and apply these changes in a controlled manner. Integrating the operator with Atlas Cloud provides enhanced visibility and management capabilities, such as viewing migration histories and affected databases, while also supporting the encryption and storage of sensitive information like database credentials as Sealed Secrets. Although challenges exist, such as handling pre-existing SQL files within Kubernetes YAMLs, the approach offers a comprehensive method to manage database migrations alongside other Kubernetes applications, using tools like Codefresh to streamline the process and ensure everything is version-controlled and traceable.

Argo CD 2.8 has removed support for creating config management plugins (CMPs) via configmap, a feature deprecated since version 2.4, prompting users to adapt to new methods for integrating configurations like Kustomize with Helm. The process involves creating a CMP by setting up a sidecar container in the `argocd-repo-server`, which manages configuration via mounted yaml files and Docker images containing necessary tools. The example uses the `alpine/k8s` image and demonstrates the steps to configure and integrate the CMP into an Argo CD instance using Kustomize, emphasizing the importance of security and compatibility with different Kubernetes versions. Users must update application manifests to leverage the new plugin, with suggestions for further learning through the Codefresh GitOps with Argo Certification and using Codefresh's platform to enhance CI/CD and GitOps processes.

Argo Rollouts is a Kubernetes controller that facilitates advanced deployment strategies like blue/green deployments and canaries in a Kubernetes cluster, and is integrated with platforms like Codefresh for enhanced user interface support. While it natively manages container image updates through a Rollout resource, challenges arise when configuration changes, stored in configmaps, accompany application updates, as these are not inherently part of a Rollout. To address this, Kustomize's configmap generators can be leveraged to create and manage version-specific configmaps dynamically, allowing for seamless configuration updates without altering the container image. This approach enables progressive delivery for both application code and configurations, ensuring each deployment version operates with its respective settings. Additionally, the technique supports deployments driven solely by configuration changes, offering flexibility in fine-tuning application settings in production environments. Future discussions will explore more complex scenarios involving multiservice deployments and Codefresh's orchestration capabilities.