March 2021 Summaries

Shawn Sesna's post from March 31, 2021, provides a detailed guide on deploying the Java application PetClinic to a Payara web server using Octopus Deploy, which supports various web servers including Tomcat and Wildfly. The deployment process involves setting up infrastructure with a MySQL PaaS server on Azure and an Ubuntu VM for Payara, automated through a runbook. The article describes the necessary steps for configuring the Azure Resource Manager template and Bash script to install the Octopus Tentacle and Payara server, and addresses potential issues with MySQL PaaS, such as firewall configurations and permissions for stored procedures. It also highlights Payara's autodeploy feature, which simplifies deployment by allowing the .war file to be placed in a specific folder for automatic deployment. Additionally, the post covers the use of Octopus Deploy's Structured Configuration Variables to update database connection information and concludes by emphasizing that, despite the absence of a specific template for Payara, Octopus Deploy effectively supports deploying applications to it.

Phil Stephenson's blog post provides an in-depth guide to automating the creation of API keys for the Octopus REST API using scripting techniques, particularly with PowerShell. The post underscores the significance of Octopus's API-first design, allowing users to perform any action available on the Octopus Web Portal through the API. It details the process of utilizing the Swagger API documentation and Chrome DevTools to understand and replicate browser interactions for API key creation, emphasizing the importance of service accounts for assigning appropriate permissions to API keys. Stephenson also discusses leveraging browser tools to inspect request and response headers during login, enabling users to simulate these actions programmatically without manual intervention. The blog concludes by highlighting the utility of automating these processes for service users, provisioning Octopus instances, and integrating with tools like the Octopus Terraform Provider, while encouraging further exploration of the Octopus REST API through additional resources such as webinars.

The blog post by Shawn Sesna discusses the integration of TeamCity and Octopus Deploy to enhance the deployment process by including build information. It highlights the importance of continuous integration (CI) and continuous delivery (CD) in software development, emphasizing the role of source control, issue tracking, and build servers. The post specifically details how to log issues in Jira for an application, tag commits with issue keys, and configure Octopus Deploy to push build information. It also covers setting up issue tracking integration in Octopus Deploy, particularly with Jira, to ensure issue tracking works effectively. Additionally, it explains how to use the Release Note Prefix feature for better release notes and provides methods for sharing build information with stakeholders via release notes templates or email notifications. Overall, the post serves as a practical guide for utilizing Octopus Deploy to improve communication and transparency in CI/CD pipelines.

Managing Kubernetes clusters can be challenging, especially when dealing with multiple clusters, but tools like Rancher simplify this process by providing a centralized interface to create and manage clusters. Rancher, which can be installed anywhere Docker is installed, allows users to create on-premises or cloud-based Kubernetes clusters, such as those on Amazon EKS, Azure AKS, and Google GKE. The process of setting up clusters is user-friendly, with Rancher providing an intuitive UI for configuration. After setting up clusters in Rancher, users can integrate them with Octopus Deploy by creating an API key for authentication and using Rancher's proxy communication features to manage connections. This integration enables streamlined deployments to both on-premises and cloud-based Kubernetes clusters. By following the steps outlined, users can effectively deploy projects using Octopus Deploy to clusters managed by Rancher, taking advantage of centralized management and simplified deployment processes.

Microservices enable large development teams to deploy code independently without needing a unified codebase or release schedule, but managing their deployment can be complex due to the orchestration of Kubernetes resources and environment promotion. Octopus Deploy offers features to streamline microservice deployments, demonstrated through a process using the sample application Online Boutique by Google. By utilizing step templates, Octopus abstracts common boilerplate definitions and allows the quick definition of deployments for unique microservices through raw YAML editing. It also uses channel rules to manage Docker image tag selection and the Deploy a release step to create a meta-project for sequential deployments, ensuring microservices are deployed in the correct order across environments. Overall, these features facilitate efficient and reliable deployment of microservices, reducing manual effort and enhancing coordination.

The 2021 Q1 release of Octopus Server introduces several notable updates, including the ability for Linux users to self-host the Octopus Server and run the Tentacle agent on ARM and ARM64 hardware, marking a significant step in enhancing DevOps automation across platforms. Following Octopus's membership in the Linux Foundation in 2020, these features are aimed at improving the experience for both Windows and Linux users, with Octopus Cloud already utilizing a Linux Docker image for deployments. This release, Octopus Server 2020.6, also includes enhancements such as Global Search and improved API key management, which benefit all users by streamlining navigation and strengthening security. Additionally, the update presents long-term support for certain versions and includes breaking changes like the requirement of ProjectView permission for channel operations and discontinued support for Windows Server 2008. The release is part of a new quarterly schedule, with self-hosted customers encouraged to upgrade and feedback invited from users to refine future releases.

Matthew Casperson's blog post discusses the integration of Cypress end-to-end testing into the deployment pipeline, particularly within an Octopus deployment environment. The article highlights the challenges and solutions for running Cypress tests, emphasizing the convenience of using Docker containers over VMs for this purpose due to the pre-configured nature of Cypress Docker images. It also addresses the difficulty of mounting test scripts in Kubernetes environments and provides a workaround through Octopus worker containers, which can execute deployment steps inside Docker containers. Additionally, the post outlines the creation of a custom Docker image to run Cypress tests and demonstrates the process of executing these tests and collecting results using a Bash script. The approach allows for efficient, reusable testing across various platforms, including Kubernetes clusters, ensuring that web applications are thoroughly verified at each stage of their lifecycle before reaching end users.

Phil Stephenson's blog post provides a detailed guide on deploying a Next.js blog to AWS S3 using GitHub Actions and Octopus Deploy. The process begins with packaging the site, which involves checking out the source code, installing dependencies, generating static assets, and tagging commits with semantic-release to automate versioning. The packaged assets are then pushed to an Octopus repository using octopackjs. For deployment, an AWS S3 bucket is set up to host the static site, leveraging Octopus Deploy's built-in step template for uploading packages to S3. The post highlights the benefits of using AWS S3 for static sites, such as minimal configuration and cost-effectiveness, and suggests using environment variables for secure API key management. Additionally, it briefly touches on options for serving larger Next.js sites dynamically and securing the site with CloudFront for SSL support.

In an effort to enhance workflow efficiency and reduce feedback time, the team at Octopus has implemented significant updates to their deployment processes, including the adoption of Continuous Deployment for internal environments and Continuous Delivery for external ones. This transition allows automatic release creation from commits to releasable branches, facilitating faster and higher-quality product delivery. To combat "bit rot" in long-term support releases, nightly builds for potentially releasable branches have been introduced, which help identify and address issues early. The change in versioning strategy to a major.minor.build format accommodates nightly builds and resolves conflicts with downstream systems. Moreover, the adoption of the Git revision graph for generating release notes offers more accuracy and helps in identifying valid upgrade paths, preventing potential bugs related to database structural changes. These updates aim to improve the overall customer experience and expedite the delivery of improvements to Octopus Cloud and self-hosted installations.

Octopus Deploy is a comprehensive product that requires extensive documentation and resources, such as videos and real-world scenario examples, to assist users in navigating its vast functionalities. Recognizing that some users prefer hands-on learning, the Octopus Samples instance was created to offer experienced users a platform to explore solutions to specific deployment challenges through modeled scenarios. While the Samples instance allows guest access to view project details, such as deployment processes and runbooks for Infrastructure as Code implementations across various cloud services, it restricts actions like release creation and deployment execution. The examples in Samples are categorized into Pattern and Target spaces, demonstrating deployment patterns and technology-specific deployments, respectively, and are supported by build servers including Azure DevOps, Jenkins, and GitHub Actions.