fluxy

Work with your code, from laptop to prod.

User stories

I want to update a specific image in a specific service, and deploy it.

fluxctl release --service=S --update-image=I

I want to deploy specific service/image pairs from some other source of truth, e.g. a dev environment to a prod environment.

# Some script wrapping this command:
fluxctl release --service=S --update-image=I

I want to deploy the latest images for a given service.

fluxctl release --service=S --update-all-images

I want to release a specific image to all services that are using that image, except some services that I have manually excluded somehow.

fluxctl release --all --update-image=I

I want to deploy the latest images for all services on the platform, except some services that I have manually excluded somehow.

fluxctl release --all --update-all-images

I want to deploy a service with no change of image, just taking the latest resource definition file. This may be known as a config change deployment.

fluxctl release --service=S

I want to automatically deploy the latest images for a set of opt-in services.

fluxctl automate --service=S

I want to show all recognized services and their status.

fluxctl list-services

I want to find out what images are available for a service.

fluxctl list-images --service=S

I want to inspect the history of actions taken with Fluxy, both per-service and overall.

fluxctl history [--service=S]

Installing

For the minute you will have to build or use the container image quay.io/weaveworks/fluxy. The directory deploy/ has example Kubernetes configuration and instructions for using it.

Developing

Build

Ensure the repository is checked out into $GOPATH/src/github.com/weaveworks/fluxy. Then, from the root,

$ gvt restore
# .. time passes ..
$ make

This makes Docker images, and installs binaries to $GOPATH/bin.

Test

$ go test ./...

Dependency management

We use gvt to manage vendored dependencies. Note that we do not check in the vendor folder.

To get all the dependencies put in the vendor/ folder, use

$ go get -u github.com/FiloSottile/gvt # install gvt if you don't have it
$ gvt restore

To add dependencies, use

$ gvt fetch <dependency>

gvt does not discover dependencies for you, but it will add them recursively; so, it should be sufficient to just add packages you import.

Contribution

Flux follows a typical PR workflow. All contributions should be made as PRs that satisfy the guidelines below.

Guidelines

• All code must abide Go Code Review Comments
• Names should abide What's in a name
• Code must build on both Linux and Darwin, via plain go build
• Code should have appropriate test coverage, invoked via plain go test

In addition, several mechanical checks are enforced. See the lint script for details.