This project is written in Go and it uses go module as a dependency manager.
The tool chain I use is very simple and it does not require anything more than what Go requires.
$ go build cmd/kprofefe $ go test ./... $ go vet ./...
And so on.
This project is in continuous delivery and it uses GoReleaser with GitHub Actions.
Every time a new tag is pushed the CI runs GoReleser. It updates the changelog in the release page and it builds binaries for multiple architectures and docker images.
Docker images are available on Docker Hub.
Locally you can build a release anytime, even if you do not have a tag with:
$ goreleaser release --snapshot --rm-dist --skip-publish
The process will create binaries and Docker images.
The continuous integration is managed via GitHub Actions.
I use kubernetes-sigs/kind where I have to spin up a temporary kubernetes cluster for my test. It requires Docker.
$ kind create cluster
I can run locally
kprofefe to see how it works but it won’t be able to
actually scrape profiles because it tries to reach the pprof endpoint using POD
IPs, and they are not reachable from my local laptop.
If you would like to test the entire lifecycle you have to deploy your version of kprofefe in Kubernetes. I do not have a workflow to share about it yet.