Setup

Kubernetes

Opctl can be deployed to Kubernetes in order to run ops across an arbitrary number of nodes.

Example deployment yaml:

apiVersion: apps/v1
kind: Deployment
metadata:
name: opctl-in-kubernetes
spec:
replicas: 1
template:
    spec:
    containers:
        - name: opctl
        image: ghcr.io/opctl/opctl:0.1.58-dind
        ports:
            # expose to other containers
            - name: http
            containerPort: 42224
            protocol: TCP
        securityContext:
            privileged: true

Docker

Opctl can be run in docker containers using either DinD (Docker in Docker) or DooD (Docker out of Docker).

DinD

The -dind variant uses DinD (Docker in Docker) as the container runtime and requires a --privileged flag.

For example, to run github.com/opspec-pkgs/uuid.v4.generate#1.1.0 with DinD:

docker run \
    --privileged \
    ghcr.io/opctl/opctl:0.1.58-dind \
    opctl run github.com/opspec-pkgs/uuid.v4.generate#1.1.0

DooD

The -dood variant uses DooD (Docker out of Docker) as the container runtime and requires:

• -v /var/run/docker.sock:/var/run/docker.sock to mount the socket of the external docker daemon.
• -v opctl_data_dir:/root/opctl to mount an external directory as opctl's data dir.

For example, to run github.com/opspec-pkgs/uuid.v4.generate#1.1.0 with DooD:

docker run \
    -v /var/run/docker.sock:/var/run/docker.sock \
    -v ~/opctl:/root/opctl \
    ghcr.io/opctl/opctl:0.1.58-dood \
    opctl run github.com/opspec-pkgs/uuid.v4.generate#1.1.0

Github

Opctl can be used from Github Actions to run ops.

Example workflow yaml:

# .github/workflows/your_workflow.yml
name: your_workflow_name

on: push

jobs:
build:
    runs-on: ubuntu-latest
    steps:
    - uses: actions/checkout@v1

    - name: Install Opctl
    run: curl -L https://github.com/opctl/opctl/releases/latest/download/opctl-linux-amd64.tgz | sudo tar -xzv -C /usr/local/bin
    
    - name: Build
    run: opctl run build

Gitlab

Opctl can be used from Gitlab CI/CD to run ops.

Example .gitlab-ci.yml:

image: ghcr.io/opctl/opctl:0.1.58-dind
stages:
- build
- deploy
build:
stage: build
script:
    # passes args to opctl from gitlab variables
    - export gitlabUsername="$CI_REGISTRY_USER"
    - export gitlabSecret="$CI_REGISTRY_PASSWORD"
    - opctl run build
deploy:
stage: deploy
only:
    - main
script:
    - opctl run deploy

VSCode

Intellisense for opspec can be enabled in VSCode.

1. install vscode-yaml plugin

2. add to your user or workspace settings

   "yaml.schemas": {
     "https://raw.githubusercontent.com/opctl/opctl/main/opspec/opfile/jsonschema.json": "/op.yml"
   }

3. edit or create an op.yml w/ your fancy intellisense.

Linux

Opctl can be installed and run natively on linux.

Dependencies

Opctl currently supports multiple container runtimes. Dependencies vary based on whichever you choose:

container runtime	dependencies
docker	docker
qemu (experimental)	lima

Installation

from terminal, run:

curl -L https://github.com/opctl/opctl/releases/latest/download/opctl-linux-amd64.tgz | sudo tar -xzv -C /usr/local/bin

Upgrading

from terminal, run:

# update to latest release from https://github.com/opctl/opctl/releases
opctl self-update

OSX

Opctl can be installed and run natively on OSX.

Dependencies

Opctl currently supports multiple container runtimes. Dependencies vary based on whichever you choose:

container runtime	dependencies
docker	docker
qemu (experimental)	lima

Installation

M1

from terminal, run:

curl -L https://github.com/opctl/opctl/releases/latest/download/opctl-darwin-arm64.tgz | sudo tar -xzv -C /usr/local/bin

Intel

from terminal, run:

curl -L https://github.com/opctl/opctl/releases/latest/download/opctl-darwin-amd64.tgz | sudo tar -xzv -C /usr/local/bin

Upgrading

from terminal, run:

# update to latest release from https://github.com/opctl/opctl/releases
opctl self-update

Windows

Use Windows Subsystem for Linux (WSL) and the linux install.