-
Notifications
You must be signed in to change notification settings - Fork 57
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Improve deployment procedure #2
Comments
helm charts please :) there is already a chart currently. I believe that your approach is superior. It would be better to change it. |
For the record, i'm probably going for helm charts but i'd like to support multiple deployment strategies with regards to CA management; right now it's done outside of the cluster. That assumes a lot about proper CA management but at least enables people who know what they're doing to do it correctly :). What chart did you find? |
I'm using terraform for infrastructure as code and have recently created a personal VPN project using some of your code and some terraform code all deploying to GCP/GKE: https://github.com/zambien/terraform-gcp-gke-openvpn Please feel free to use any of this if you like. I'd welcome any feedback, issues, or PRs. |
That looks like a fair amount of work! I haven't (seriously) worked with terraform yet but i'll make an honest effort to review what you've done here in the next week. |
Cool, thanks. High level, it does the following:
Looking back at it you are right.. it was more work than I intended! :) I've been using Terraform for a few months now and really like it. My team has been creating immutable stacks for various technologies mostly around AWS. We have templates for ECS, EC2 clusters, security groups, etc. Unfortunately I can't share most of that because it was done for the company I work for on an internal repo. This was my first foray into GCP and Kubernetes so the end result is a little rough.. more of a learning/prototyping exercise than anything else. Since the motivation was to create a personal VPN with very low cloud cost (I failed) I'm going to try again with just a plain old GCE f1-micro instance and a docker machine deployment. That repo is a WIP and here: |
Trying to get the ticket back on topic: I will need to define a clear strategy for dealing with this problem before i'll cut a 1.0 release. I'd like to implement a full-fledged kubernetes Operator function that will do automatic key rotation & management all server(cluster)-side out of UX concerns and not being able to trust human operators to do regular key rotation, but at the same time leave the option of managing your own PKI. That last part shouldn't be too hard as it is what we have now. That first option leaves a huge amount of trust to the cluster and considering the deployment scenarios that trust is not always warranted and should be explicitly kept out of the cluster. |
That sounds like a separate project possibly but I may not be understanding the scope of what needs to be solved. I've written infrastructure as code similar to what I linked earlier for HashiCorp's Vault. I use that for my internal PKI among other things. Am I on the right track here or off base? |
I'm currently not happy with the scripts used for deployment but i'm not sure what to do about it yet.
The text was updated successfully, but these errors were encountered: