colima README

The project's README from github.com/abiosoft/colima. As one markdown file: README.md.

colima-logo

Colima - container runtimes on macOS (and Linux) with minimal setup.

Go Integration Integration Go Report Card

Demonstration

Website & Documentation: colima.run | colima.run/docs

Features

Support for Intel and Apple Silicon macOS, and Linux

Getting Started

Installation

Colima is available on Homebrew, MacPorts, Nix and Mise. Check here for other installation options.

# Homebrew
brew install colima

# MacPorts
sudo port install colima

# Nix
nix-env -iA nixpkgs.colima

# Mise
mise use -g colima@latest

Or stay on the bleeding edge (only Homebrew)

brew install --HEAD colima

Usage

Start Colima with defaults

colima start

For more usage options

colima --help
colima start --help

Or use a config file

colima start --edit

Runtimes

On initial startup, Colima initiates with a user specified runtime that defaults to Docker.

Docker

Docker client is required for Docker runtime. Installable with brew install docker.

colima start
docker run hello-world
docker ps

You can use the docker client on macOS after colima start with no additional setup.

Containerd

colima start --runtime containerd starts and setup Containerd. You can use colima nerdctl to interact with Containerd using nerdctl.

colima start --runtime containerd
nerdctl run hello-world
nerdctl ps

It is recommended to run colima nerdctl install to install nerdctl alias script in $PATH.

Kubernetes

kubectl is required for Kubernetes. Installable with brew install kubectl.

To enable Kubernetes, start Colima with --kubernetes flag.

colima start --kubernetes
kubectl run caddy --image=caddy
kubectl get pods

Interacting with Image Registry

For Docker runtime, images built or pulled with Docker are accessible to Kubernetes.

For Containerd runtime, images built or pulled in the k8s.io namespace are accessible to Kubernetes.

Incus

Requires v0.7.0

Incus client is required for Incus runtime. Installable with brew brew install incus.

colima start --runtime incus starts and setup Incus.

colima start --runtime incus
incus launch images:alpine/edge
incus list

You can use the incus client on macOS after colima start with no additional setup.

Note: Running virtual machines on Incus is only supported on m3 or newer Apple Silicon devices.

AI Models (GPU Accelerated)

Requires v0.10.0, Apple Silicon and macOS 13+

Colima supports GPU accelerated containers for AI workloads using the krunkit VM type.

Note: To use krunkit with colima, ensure it is installed. Please follow their installation instructions

Setup and use a model.

colima start --runtime docker --vm-type krunkit
colima model run gemma3

Colima supports two model runner backends:

The default registry is the Docker AI Registry. Models can be run by name without a prefix:

colima model run gemma3
colima model run llama3.2
# HuggingFace (Docker Model Runner)
colima model run hf.co/microsoft/Phi-3-mini-4k-instruct-gguf
# Ollama (requires ramalama runner)
colima model run ollama://gemma3 --runner ramalama

See the AI Workloads documentation for more details.

Customizing the VM

The default VM created by Colima has 2 CPUs, 2GiB memory and 100GiB storage.

The VM can be customized either by passing additional flags to colima start. e.g. --cpu, --memory, --disk, --runtime. Or by editing the config file with colima start --edit.

NOTE: Disk size can be increased after the VM is created.

Customization Examples

Project Goal

To provide container runtimes on macOS with minimal setup.

What is with the name?

Colima means Containers on Lima.

Since Lima is aka Linux Machines. By transitivity, Colima can also mean Containers on Linux Machines.

The logo was contributed by Daniel Hodvogner. Check this issue for more.

Troubleshooting and FAQs

Check here for Frequently Asked Questions, or visit the online FAQ for a searchable version.

How to Contribute?

Check here for the instructions on contributing to the project.

Community

License

MIT

Sponsoring the Project

If you (or your company) are benefiting from the project and would like to support the contributors, kindly sponsor.


colima (abiosoft/colima)

Searching Database…

31.2K ·GitHub
这是干什么的?Colima 让你用最少的配置就能在 macOS(和 Linux)上使用容器运行时:它会启动一个运行 Docker、Containerd 或 Incus 的 Linux 虚拟机,全部通过一条简单的命令来管理。
有图形界面吗,还是要怎么用它?它是一个命令行工具:colima start 会启动一个 Linux 虚拟机,里面运行你选择的运行时。一切都用 colima 命令来管理,它还自带端口自动转发、卷挂载和多个实例等功能。
怎么把它跑起来?用 Homebrew、MacPorts、Nix 或 Mise 安装(比如 brew install colima),然后运行 colima start。第一次启动默认使用 Docker 运行时。
需要单独安装 Docker 或其他客户端吗?客户端不带在包里:Docker 运行时需要一个 Docker 客户端(brew install docker),Kubernetes 需要 kubectl,Incus 需要 incus 客户端。只要启动了对应的运行时,客户端无需额外配置就能直接用。
那些高级功能对硬件有要求吗?Incus 虚拟机只能在 M3 或更新的 Apple Silicon 设备上运行。其他附加功能比如 Rosetta 2 模拟,也需要 Apple Silicon 上的 macOS 13(Ventura)或更新版本。
它只能在 Apple Silicon 的 Mac 上用吗?不是——Colima 同时支持 Intel 和 Apple Silicon 的 Mac,也能在 Linux 上运行。
容器能调用 Mac 的 GPU 来跑 AI 吗?可以,从 v0.10.0 开始,在 Apple Silicon 加 macOS 13 及以上系统上,通过 krunkit 虚拟机类型即可实现。支持两种模型运行后端:Docker Model Runner(默认,支持 Docker AI Registry 和 HuggingFace)和 Ramalama(支持 HuggingFace 和 Ollama 仓库)。
能用它跑 Kubernetes 吗?可以——用 'colima start --kubernetes' 启动 Colima(需要 kubectl,可用 brew install kubectl 安装)。用 Docker 构建或拉取的镜像 Kubernetes 能自动访问;用 Containerd 时,k8s.io 命名空间里的镜像可以。
虚拟机默认分到多少 CPU 和内存?默认虚拟机是 2 个 CPU、2GiB 内存和 100GiB 存储空间。可以用 --cpu、--memory 和 --disk 参数修改,或者用 'colima start --edit' 编辑配置;磁盘大小在创建之后还可以增加。
为什么叫 Colima?Colima 的意思是 Containers on Lima(Lima 上的容器),而 Lima 是 Linux Machines(Linux 机器)的缩写。以此类推,它也可以理解为 Containers on Linux Machines(Linux 机器上的容器)。
What are similar projects?lima, containerd, xpipe, vcluster, k9s, tilt, nerdctl, minikube
Creator
abiosoft
Size
3.2 MB
Stars
31.2K
Age
5y
Commits · 90 Days