To fulfil the entire lifecycle of cyberfabric, we would need to support a new command.
deploy should be in charge of creating for now:
- Basic helm templates: Chart + deployment + configmap(based on the config specified) + ingress(if required) + service + secrets(if required)
- Dockerfile (simple one)
For now that command should be very restricted, while we are developing features and collecting feedback.
-c/--config as we currently use.
--docker should build an image based on:
- Generate the manifest and main file with our internal logic for
build.
- Use multi-stage build in the Dockerfile, so first step you upload the
Cargo.toml, Cargo.lock, src/main.rs, .cargo/config.toml from .cyberfabric/<NAME> and workspace from the project: Cargo.toml + modules/* or other related members from that workspace.
- Second step will be to take that binary, put it in a place like
/srv/{EXECUTABLE_NAME:cyberfabric} with an ENTRYPOINT to run it. Also make sure to load the configuration file pointed with -c to /srv/config.yml.
I expect one PR with the previous, and another PR with --helm in order to generate that helm templates based on the configuration(although for now let's keep it basic).
To fulfil the entire lifecycle of cyberfabric, we would need to support a new command.
deployshould be in charge of creating for now:For now that command should be very restricted, while we are developing features and collecting feedback.
-c/--configas we currently use.--dockershould build an image based on:build.Cargo.toml,Cargo.lock,src/main.rs,.cargo/config.tomlfrom.cyberfabric/<NAME>and workspace from the project:Cargo.toml+modules/*or other related members from that workspace./srv/{EXECUTABLE_NAME:cyberfabric}with an ENTRYPOINT to run it. Also make sure to load the configuration file pointed with-cto/srv/config.yml.I expect one PR with the previous, and another PR with
--helmin order to generate that helm templates based on the configuration(although for now let's keep it basic).