Continuous integration automates the building, testing and deploying of applications. Software projects, whether created by a single individual or entire teams, typically use continuous integration as a hub to ensure important steps such as unit testing are automated rather than manual processes.
When continuous integration (CI) is established as a step in a software project's development process it can dramatically reduce deployment times by minimizing steps that require human intervention. The only minor downside to using CI is that it takes some initial time by a developer to set up and then there is some ongoing maintainence if a project is broken into multiple parts, such as going from a monolith architecture to microservices.
Another major advantage with CI is that testing can be an automated step in the deployment process. Broken deployments can be prevented by running a comprehensive test suite of unit and integration tests when developers check in code to a source code repository. Any bugs accidentally introduced during a check-in that are caught by the test suite are reported and prevent the deployment from proceeding.
The automated testing on checked in source code can be thought of like the bumper guards in bowling that prevent code quality from going too far off track. CI combined with unit and integration tests check that any code modifications do not break existing tests to ensure the software works as intended.
The following picture represents a high level perspective on how continuous integration and deployment can work.
In the above diagram, when new code is committed to a source repository there is a hook that notifies the continuous integration server that new code needs to be built (the continuous integration server could also poll the source code repository if a notification is not possible).
The continuous integration server pulls the code to build and test it. If all tests pass, the continuous integration server begins the deployment process. The new code is pulled down to the server where the deployment is taking place. Finally the deployment process is completed via restarting services and related deployment activities.
There are many other ways a continuous integration server and its deployments can be structured. The above was just one example of a relatively simple setup.
There are a variety of free and open source continuous integration servers that are configurable based on a project's needs.
Note that many of these servers are not written in Python but work just fine for Python applications. Polyglot organizations (ones that use more than a single language and ecosystem) often use a single CI server for all of their projects regardless of the programming language the application was written in.
BuildBot is a continuous integration framework with a set of components for creating your own CI server. It's written in Python and intended for development teams that want more control over their build and deployment pipeline. BuildBot source code is on GitHub.
TeamCity is JetBrains' closed source CI server that requires a license to use.
Jenkins is commonly used as a continuous integration server implementation for Python projects because it is open source and programming language agnostic. Learn more via the following resources or on the dedicated Jenkins page.
My book on deploying Python web applications walks through every step of setting up a Jenkins project with a WSGI application to enable continuous delivery. Take a look if you're not grokking all of the steps provided in these other blog posts.
Assembling a continuous integration service for a Django project on Jenkins shows how to set up a Ubuntu instance with a Jenkins server that'll build a Django project.
Setting up Jenkins as a continuous integration server for Django is another solid tutorial that also shows how to send email notifications as part of the build process.
What is continuous integration? is a classic detailed article by Martin Fowler on the concepts behind CI and how to implement it.
Continuous Deployment For Practical People is not specific to Python but a great read on what it entails.
Continuous Integration & Delivery - Illustrated uses well done drawings to show how continuous integration and delivery works for testing and managing data.
Diving into continuous integration as a newbie is a retrospective on learning CI from a Rackspace intern on how she learned the subject.
6 top continuous integration tools gives a high level overview of six CI tools from a programming language agnostic perspective.
Updating the GOV.UK Continuous Integration environment explains the UK's Government Digital Service continuous integration configuration that relies on Jenkins.
StackShare's Continuous Integration tag lists a slew of hosted CI services roughly ranked by user upvotes.
Good practices for continuous integration includes advice on checking in code, commit tests and reverting to previous revisions.
Scoring Continuous Integration gives an interesting perspective on ways to rank the effectiveness of how teams use their CI tooling.
Why Continuous Integration Is Important is a high-level overview of how CI can build trust both among developers and between developers and non-technical people in an organization. The post also discusses tasks related to setting up reliable CI such as test environments, integration testing and visibility into the CI results.
Continuous Intrusion: Why CI tools are a hacker's best friend (PDF) strongly advises securing your continuous integration server just as you would every other part of your production application, unless you want your environment to be vulnerable to malicious actors.
Measuring and Improving your CI/CD Pipelines provides metrics for what you should measure with your CI/CD setup to improve the process for helping your development teams ship code.
Six rules for setting up continuous integration systems has some solid general advice for culling problematic tests, ensuring the integration speed supports the development culture you are building and keeping all code in source control instead of having complicated logic configured within the CI server.
How to Identify Major Blockers in a CI/CD Pipeline gives a high level overview of concepts such as shipping velocity, test execution and environment provisioning with regards to CI configurations.
Fix errors in your Python code before your users see them by monitoring with Rollbar.