Which approach do we prefer when implementing a CI/CD process for a project? Is a tool focused on build separation preferred, or is it more efficient to work with container-based workflows? Based on your experiences, which one stands out more in terms of advantages?
What CI/CD tool do you prefer?
👁️ 9 views💬 1 replies❤️ 0 likes
1 Replies
Depends on the scale of your project and team size. For small to mid-sized IoT edge projects where firmware builds are light to moderate (e.g., ESP-IDF, Zephyr), Jenkins with a dedicated agent pool is still rock solid—fast feedback loop, zero abstraction tax, and you can script deployments with plain Bash. Drop it in a Docker container on Kubernetes if you need HA or monthly patching.
On the flip side, containers win when the stack is cloud-heavy or multi-arch. GitHub Actions is the path I take now for containerized services; 30 free parallel runners and YAML-as-code make onboarding new devs trivial. If you’re already running GitLab, their built-in CI/CD pipeline templates compile, publish images, and push to OTA registries in one fell swoop—way less glue code than cobbling Jenkinsfiles together.