I think I've tried most of the CI tools out there at some point. It took me a while to get around to Buildkite because at first I didn't see much point given it seemed like you had to run the agent yourself. Eventually it dawned on me why this approach was more ingenious than I realised:

Running my app in a production (or production-like) environment was already a solved problem, because everything was already in some form of "everything as code". Having a test environment where the only difference was adding the Buildkite agent was a trivial addition.

It means that dev/test/prod parity is simple to achieve and maintain. It's also proven to be much easier to support than trying to deal with the problems that come with trying to force an app to fit into the nuances and constraints that are imposed by the containers/runtime of a CI service. When you completely control all of the environment the tests are running in you define those constraints too. It's been a great balance between a managed service and the flexibility of running it yourself.

And while none of my needs have hit the scale of Shopify (I saw one of their engineers speak about it at a conference once, I can't find the video now though 😞) it's good to know I can scale out my worker nodes to hundreds of thousands of workers to reduce the time it takes for my tests to run.

I would recommend you to consider the JFrog Platform that includes JFrog Pipelines - it will allow you to manage the full artifact life cycle for your sbt, docker and other technologies, and automate all of your CI and CD using cloud native declarative yaml pipelines. Will integrate smoothly with all your other toolset.

more configurable to setup ci/cd: * It can provide caching when build sbt, just add this section to yml file * Easy to use, many documentation

Weakness: * Need use gitlab as repository to bring more powerful configuration

I have been working with Puppet and Ansible. The reason why I prefer ansible is the distribution of it. Ansible is more lightweight and therefore more popular. This leads to situations, where you can get fully packaged applications for ansible (e.g. confluent) supported by the vendor, but only incomplete packages for Puppet.

The only advantage I would see with Puppet if someone wants to use Foreman. This is still better supported with Puppet.

If you are just starting out, might as well learn Kubernetes There's a lot of tools that come with Kube that make it easier to use and most importantly: you become cloud-agnostic. We use Ansible because it's a lot simpler than Chef or Puppet and if you use Docker Compose for your deployments you can re-use them with Kubernetes later when you migrate

We use CircleCI because of the better value it provides in its plans. I'm sure we could have used Travis just as easily but we found CircleCI's pricing to be more reasonable. In the two years since we signed up, the service has improved. CircleCI is always innovating and iterating on their platform. We have been very satisfied.

As the maintainer of the Karate DSL open-source project - I found Travis CI very easy to integrate into the GitHub workflow and it has been steady sailing for more than 2 years now ! It works well for Java / Apache Maven projects and we were able to configure it to use the latest Oracle JDK as per our needs. Thanks to the Travis CI team for this service to the open-source community !

I use Google Cloud Build because it's my first foray into the CICD world(loving it so far), and I wanted to work with something GCP native to avoid giving permissions to other SaaS tools like CircleCI and Travis CI.

I really like it because it's free for the first 120 minutes, and it's one of the few CICD tools that enterprises are open to using since it's contained within GCP.

One of the unique things is that it has the Kaniko cache, which speeds up builds by creating intermediate layers within the docker image vs. pushing the full thing from the start. Helpful when you're installing just a few additional dependencies.

Feel free to checkout an example: Cloudbuild Example

I use Travis CI because of various reasons - 1. Cloud based system so no dedicated server required, and you do not need to administrate it. 2. Easy YAML configuration. 3. Supports Major Programming Languages. 4. Support of build matrix 6. Supports AWS, Azure, Docker, Heroku, Google Cloud, Github Pages, PyPi and lot more. 7. Slack Notifications.

You are probably looking at another hosted solution: Jenkins is a good tool but it way too work intensive to be used as just a backup solution.

I have good experience with Circle-CI, Codeship, Drone.io and Travis (as well as problematic experiences with all of them), but my go-to tool is Gitlab CI: simple, powerful and if you have problems with their limitations or pricing, you can always install runners somewhere and use Gitlab just for scheduling and management. Even if you don't host your git repository at Gitlab, you can have Gitlab pull changes automatically from wherever you repo lives.

If you are considering Jenkins I would recommend at least checking out Buildkite. The agents are self-hosted (like Jenkins) but the interface is hosted for you. It meshes up some of the things I like about hosted services (pipeline definitions in YAML, managed interface and authentication) with things I like about Jenkins (local customizable agent images, secrets only on own instances, custom agent level scripts, sizing instances to your needs).