Combining development and operations is at the core of successful software delivery. A well-working DevOps pipeline is one of the first things we set up for any solution. It forms the foundation for the capability to iteratively deliver features.
DevOps is not just about writing code and deploying it. We take care of the whole application lifecycle with a seamless transition to our suite of 24/7 continuous services. The same toolchain can remain in use as our developers and support personnel are both intimately familiar with them.
We use the template below to start the pipeline design process on AWS. Customisations and additions to this are done based on a case by case basis, for example when code quality tooling or other SaaS services are added to the architecture.
We see DevOps as a continuous cultural trait that can start small and grow with a software product and the organisation developing it. It's essential to get the basics right to stay competitive. Further investments in DevOps will bring returns as delivery times shorten and the reliability of the solution increases.
Code, review, deploy, test & monitor. Infrastructure as code. Live deployments without service breaks. Automated testing and deployment up to staging environments.
Automated feature branch environments, automated production deployment as trust in automation increases.
Canary deployments, Blue/Green deployments, global cross-region solutions.
We use AWS platform services for DevOps setup, including CodeCommit as a source code repository and CodeBuild, CodePipeline, and CodeDeploy for various stages of Continuous Integration and Continuous Delivery pipeline. For hybrid environments, we evaluate the tools with the best fit for the use case. Our toolbox is extensive, including Gitlab, GitHub, CircleCI, CodeShip & Jenkins, as well as various quality and security assurance tools (SonarQube, Snyk, etc.).
Our 24/7 services support DevOps teams in making sure services are always available. We use AWS CloudWatch logs and metrics with other tools as needed to make sure we're always aware of the state of our solutions. Automated alarms based on threshold triggers or behavioral anomaly detection alert our specialists to have a closer look and make sure automated recovery actions are in place for the issue. Our skilled Cloud Ops specialists will investigate and fix issues in cooperation with DevOps teams to get to the root cause of the issues.
