Build Parameters

Environment vs User Parameters

atomic-reactor requires various parameters to build container images. These range from the Koji Hub URL to git commit. We can categorize them into environment and user parameters. The main difference between them is how they are reused. Environment parameters are shared by different build requests, while user parameters are unique to each build request.

Environment parameters are used for configuring usage of external services such as Koji, ODCS, SMTP, etc. They are also used for controlling some aspects of the container images built, for example, distribution scope, vendor, authoritative-registry, etc.

User parameters contain the unique information for a user’s build request: git repository, git branch, Koji target, etc.

Reactor Configuration

The environment configuration is supplied as a ConfigMap like:

apiVersion: v1
kind: ConfigMap
data:
    "config.yaml": <encoded yaml>

The schema definition config.json in atomic-reactor contains a description for each property.

Example of REACTOR_CONFIG:

version: 1

remote_hosts:
    slots_dir: path/foo
    pools:
        x86_64:
            osbs-remote-hosts-1-x86_64:
                enabled: true
                auth: /secret-path
                username: podman-user
                slots: 1
                socket_path: /run/user/2022/podman/podman.sock
            osbs-remote-hosts-2-x86_64:
                enabled: false
                auth: /secret-path
                username: podman-user
                slots: 2
                socket_path: /run/user/2022/podman/podman.sock
        ppc64le:
            osbs-remote-hosts-1-ppc64le:
                enabled: true
                auth: /secret-path
                username: podman-user
                slots: 3
                socket_path: /run/user/2022/podman/podman.sock

koji:
    hub_url: https://koji.example.com/hub
    root_url: https://koji.example.com/root
    auth:
        ssl_certs_dir: /var/run/secrets/atomic-reactor/kojisecret
    use_fast_upload: false

odcs:
    api_url: https://odcs.example.com/api/1
    auth:
        ssl_certs_dir: /var/run/secrets/atomic-reactor/odcssecret
    signing_intents:
    - keys: ['R123', 'R234']
      name: release
    - keys: ['B123', 'B234', 'R123', 'R234']
      name: beta
    - keys: []
      name: unsigned
    default_signing_intent: release

smtp:
    host: smtp.example.com
    from_address: osbs@example.com
    error_addresses:
    - support@example.com
    domain: example.com
    send_to_submitter: True
    send_to_pkg_owner: True

arrangement_version: 6

artifacts_allowed_domains:
- download.example.com/released
- download.example.com/candidates

image_labels:
    vendor: "Spam, Inc."
    authoritative-source-url: registry.public.example.com
    distribution-scope: public

image_equal_labels:
- [description, io.k8s.description]

openshift:
    url: https://openshift.example.com
    auth:
        enable: True

group_manifests: False

platform_descriptors:
- platform: x86_64
  architecture: amd64

content_versions:
- v2

# Output registry (built images are pushed here).
registry:
    url: https://container-registry.example.com/v2

# Default source registry (base images are pulled from here)
source_registry:
    url: https://registry.private.example.com

# Additional source registries
pull_registries:
- url: https://registry.public.example.com

registries_cfg_path: /var/run/secrets/atomic-reactor/v2-registry-dockercfg

sources_command: "fedpkg sources"

required_secrets:
- kojisecret
- odcssecret
- v2-registry-dockercfg
- client-config-secret

skip_koji_check_for_base_image: False

build_env_vars:
- name: HTTP_PROXY
  value: "http://proxy.example.com"
- name: HTTPS_PROXY
  value: "https://proxy.example.com"
- name: NO_PROXY
  value: localhost,127.0.0.1

• registries_cfg_path: A directory path where holds a docker configuration file for registry authentication. Either .dockercfg or .dockerconfigjson is supported. A typical configuration

User Parameters

TBD