February 2018 Summaries

In a "Will it Deploy?" episode, Rob Pearson demonstrates automating the deployment of a Microsoft SQL Server database using Entity Framework Core migrations to an AWS virtual machine with Octopus Deploy. The featured application, Random Quotes, is a simple ASP.NET Core 2.0 web app backed by a Microsoft SQL Server 2017 database, developed initially by marketing manager Andrew. The deployment targets an AWS EC2 virtual machine running Microsoft Windows Server 2016, utilizing Octopus Deploy steps to manage the process, including copying database scripts, executing migration scripts, and deploying the web application to IIS. The episode's GitHub repository provides all necessary resources and links, inviting viewers to suggest future topics and subscribe for more content.

Andrew Katsivas's article provides a detailed guide on setting up a continuous delivery (CD) pipeline for .NET Core projects using Bitbucket Pipelines and Octopus Deploy. The article highlights Bitbucket as a free, git-based source control platform suitable for personal projects and introduces Bitbucket Pipelines as a lightweight cloud continuous integration (CI) service using Docker containers. The process involves committing code to Bitbucket, triggering a build script, and pushing the app code to an Octopus Deploy server, which automatically creates a release and deploys it. The guide outlines creating necessary accounts, setting up a Bitbucket repository, and configuring a bitbucket-pipelines.yml file to automate builds. It further explains extending the build script to package and push code to Octopus using an API key secured as a pipeline variable. Finally, it instructs on setting up Octopus for automatic deployment to a staging environment by creating a specific lifecycle and project, enabling automatic release creation and deployment upon detecting new packages. The article aims to simplify and encourage the adoption of CD pipelines for personal projects, offering insights into the benefits and possibilities of automated deployments.

Kubernetes has gained significant traction as a leading container orchestration platform, prompting discussions on how to integrate its support into Octopus Deploy. The proposed integration includes introducing a new Kubernetes Cluster target type and two deployment steps: Kubernetes Apply and Kubectl Script, allowing for declarative object management and custom script execution respectively. The Kubernetes Apply step will support template-based deployments with JSON or YAML, similar to AWS CloudFormation, and will feature Octopus's unique capability of substituting container image versions and variables directly into templates. However, the challenge remains in how to best supply variables to these templates due to Kubernetes's lack of parameter file support. The Kubectl Script step will enable users to execute a range of Kubernetes commands by providing a script interface authenticated against the target cluster. Octopus seeks user feedback to ensure these implementations align with typical Kubernetes usage scenarios.

Matthew Casperson's blog post from February 18, 2018, discusses deploying Spring Boot applications as fully executable JAR files on a Linux environment, specifically an EC2 instance created via an AWS CloudFormation template with Octopus Deploy. The post outlines the process, including the configuration of AWS and SSH accounts, creating an executable JAR using the Spring Boot Maven plugin, and setting up machine policies in Octopus to manage dynamic IP addresses for EC2 instances. It details the deployment workflow, which involves transferring the JAR file to the EC2 instance, registering the instance with Octopus, and installing the application as a Linux service. Additionally, the blog covers using CloudFormation to define security groups, configure instance properties, and manage the installation of necessary dependencies like DotNET Core and Java. The deployment involves steps like health checks and package transfers, and concludes with verifying the application as a running service, illustrating the efficiency and convenience of using fully executable JARs with Octopus for managing cloud-based Spring Boot services.

In this episode of "Will it Deploy?", the team successfully automates the deployment of a Spring Boot web application called Random Quotes on the Amazon Web Services (AWS) platform using Octopus Deploy. The deployment process involves cloud infrastructure provisioning with AWS CloudFormation Templates and ensures zero-downtime through a blue-green deployment pattern. Key steps include setting up an Octopus AWS account for secure authentication, transferring the Spring Boot jar package, copying it to an S3 bucket, and deploying it to Elastic Beanstalk environments. The process culminates in swapping environment URLs to achieve zero downtime during production deployment. The episode also encourages viewers to suggest future technologies for exploration and invites them to subscribe to the series' YouTube channel.

Octopus Deploy has evolved from a tool primarily used for deploying .NET applications to servers into a versatile distributed task runner capable of handling a variety of workloads and processes. It is particularly effective for managing PowerShell modules, as demonstrated by the author's use of Octopus for testing and publishing the StatusCakeDSC module. This process involves setting up a standard Octopus project with sensitive information securely stored and deploying steps tailored to different environments. The author utilizes a serverless webhook solution, Takofukku, to trigger deployments based on GitHub commits, ensuring continuous delivery by automating tests with Pester and publishing to the PowerShell Gallery only if all conditions are met. The workflow concludes with Slack notifications to inform the author about deployment statuses, highlighting the efficiency and convenience of using Octopus Deploy for continuous delivery of PowerShell modules.

Matthew Casperson's blog post, published on February 11, 2018, details the deployment of Java applications in a cloud-based environment using Octopus Deploy, focusing on the integration of Maven, AWS CloudFormation, and WildFly. The process involves setting up a Maven feed to source applications, configuring AWS and SSH accounts for authentication, and employing Bitnami AMIs to quickly launch EC2 instances. A CloudFormation template is used to create and configure these instances, installing necessary packages and setting permissions for silent authentication with WildFly. The EC2 instances register themselves with the Octopus server, allowing for seamless deployment of Java applications, such as the gwt-material-demo artifact from Maven Central. The deployment process is further enhanced by health checks and the automatic removal of invalid deployment targets, ensuring reliability. Casperson encourages users interested in automating deployments or building cloud infrastructure to explore Octopus Deploy's capabilities.

Octopus integrates support as a core aspect of its operations, involving all staff, including the CEO, in addressing customer queries. The company, primarily based in Australia, experiences an influx of support requests from overseas, particularly from the US and UK, which are managed through a morning triage call to prioritize and assign issues efficiently. Utilizing tools like Supportflow and Tendersmash, Octopus ensures quick assignment and tracking of support requests via Trello and Slack integrations. Teams conduct morning stand-ups to review assigned tickets, aiming for swift and comprehensive resolutions that benefit both current and future customers by addressing root causes and updating documentation or interfaces. While striving to resolve issues on the first contact, Octopus keeps customers informed and adjusts their response based on the severity of the issue, especially in critical situations like production outages.

Octopus Deploy 2018.2 introduces several new features, including enhanced AWS support with steps to deploy CloudFormation templates, delete existing stacks, and execute custom scripts using AWS CLI, all facilitated by Octopus providing the necessary credentials and CLI tools. The release also unveils the "Deploy a Release" step, allowing for better coordination of multiple projects by enabling one project to trigger another's deployment. Additionally, the update introduces external workers, which allow scripts to run on a Tentacle instead of the Octopus Server, enhancing security and performance. Improvements have been made to dashboards for better readability in large instances and to audit logs for more detailed tracking of changes, especially regarding Project Variables. The release also includes changes to the Packages API to accommodate new feed types, requiring updates to the Octopus.Client.dll for compatibility. Users are encouraged to follow standard upgrading procedures and can find more details on specific changes in GitHub tickets.

Shane Gill describes his experience configuring NGINX as a reverse proxy for two web applications and outlines how he automated this process using Octopus Deploy. By setting up a project in Octopus, he simplified the deployment and management of configuration changes and SSL certificates, enabling these tasks to be completed with a single button push. The setup involves installing NGINX, managing site configurations through source control, and using Octopus to handle SSL certificates, with steps for copying configuration files to the correct directories and maintaining certificate chains. This automation allows easy updates and scalability, enabling others to quickly understand and contribute to the project without delving into extensive documentation.