Recently, I've been planning the deployment strategy for a new project and came across two mainstream containerization approaches: 1) directly packaging the application into an image and running it on the cloud; 2) first setting up a full CI/CD pipeline to automatically build → test → deploy the code. Which approach do you lean towards and why? Any fellow devs who’ve hit snags—share your experiences so I don’t take the long way around!
Struggling to choose a containerization deployment option? Cast your vote and pick your preference!
👁️ 5 views💬 1 replies❤️ 0 likes
1 Replies
A few months ago, I faced the same question when I had to migrate a new microservices architecture for our client. At first, I went with the first approach—just building Docker images and deploying them directly to the cloud—assuming it would be quick. But after the third manual patch and the fourth time I forgot to adjust an environment variable, I realized: without automation, things get messy sooner or later.
That’s why I upgraded to a CI/CD pipeline—using GitHub Actions. Now, every commit is automatically tested, built, and deployed to staging. The biggest advantage? I can focus on the important stuff instead of messing around with `docker build` and `kubectl apply`. If anyone’s still doubting it: start small, don’t try to migrate five services at once.