Just started with containers and kubernetes and can see clearly at work when dealing with customer's servers how using containers would prevent or solve a lot of issues ranging from updates breaking some applications while update is needed for others, scaling, redundancy etc. We provide dedicated servers and unfortunately our customers are mostly using bare metal servers or esxi and a few hyper-V servers, xenservers and not containers.

Setting now some of my stuff to use containers/pods but would be interested about code deployment with containers and how it is done at your work place , what works and what doesn't. I think there are tools to automate in github, gitlab and other 3rd party tools. Would appreciate if you could explain the deployment process at your workplace from how code is packaged , versioned, tools used, version control in general, where everything is stored, rollbacks, is rolling update used etc.. Basically the whole process for developing for applications running in containers and how code is managed and deployed.

Dont know much about that so it would really help to understand how the whole process works in production.

Also if you know any good forums for container related discussions where you can talk more about concepts and best practises, rather than needing to have a specific problem like with stackoverflow, would be interested to hear.

Not at liberty to talk about specifics but we run our own Kubernetes clusters across a multiple public clouds and some of our tools include Github, Jenkins, Gitlab runners, Spinnaker, Consul for service discovery, Vault for secrets management, etc.

I think this talk by Kelsey Hightower really drives home use cases for K8s. Worth a listen IMO.

If you haven't already you may want to get a grasp on Docker first, then dive into K8s. Container.training is a good resource for both Docker and K8s. Kubernetes in Action is a really good book, so I'd recommend checking that out. I use k3d for for experimenting and learning various concepts - I like it better than Minikube because it's lighter on resources and you can deploy multinode clusters. Out of the box it uses SQLite rather than etcd but you can change that if you'd like. Spinnaker offers a short book explaining what it is and why you'd want to use it. You can play around with Vault and Consul using Vaultron.

Thanks for the recommendations. I have seen a lot good reviews regarding kubernetes in action and will likely read it at some point. I have played around with Dockers and planning to learn more about it as well. Also watched the YouTube video and it had some good points.

Will need to see options for quick testing and experimenting(k3/minikube) but now I run everything on my esxi host and will likely to use vagrant to spin new cluster for testing things whenever I need.

Regarding turning new code to a container image does Spinnaker let you do that and what other options turning code or rpm packages to updated container image you could recommend? This is the bit I am not quite sure yet how to do it or what would be good options for my personal Websites and other applications and on the other hand options used in enterprise environments. Obviously combined with version control with git.

Spinnaker is a pretty versatile and can manage various parts of the pipeline, but all the magic happens after the image is created. Gotta build a container image first. Usually the flow is to write code, check your feature branch into version control, run tests/peer review, then merge into master, which pushes a notification via webhook to your CI tool (CircleCI, etc.) to trigger a build using the latest version of your code and build the image then push to a container registry (Docker Hub, etc.) from which Kubernetes, your IAAS, etc. will pull.

For a personal site or some other trivial application you could skip the third-party CI/CD tools and just use the automated builds feature on Docker Hub. I wouldn't fool with Spinnaker yet - you'll have enough of a time getting a grap on Docker and Kubernetes, no need to add another layer of complexity on top

Thanks a lot for explaining the process and for the recommendations.