How was deployment organized in [company]?

The deployment in the company was organized through a CI/CD pipeline using GitLab CI and Kubernetes. After successful testing and artifact building, the new container image was automatically rolled out to the staging environment for integration testing.

For production, a canary deployment was used: the new version was deployed to a small portion of instances, and if no errors occurred, traffic was gradually switched entirely. This reduced risks and allowed for quick rollback in case of problems.

All infrastructure was described in Helm charts, which ensured repeatability and ease of configuration management.