Skip to main content

Podman Installation Instructions

Looking for a GUI? You can find Podman Desktop here.

Installing on Mac & Windows

While "containers are Linux," Podman also runs on Mac and Windows, where it provides a native podman CLI and embeds a guest Linux system to launch your containers. This guest is referred to as a Podman machine and is managed with the podman machine command. Podman on Mac and Windows also listens for Docker API clients, supporting direct usage of Docker-based tools and programmatic access from your language of choice.


On Mac, each Podman machine is backed by a virtual machine. Once installed, the podman command can be run directly from the Unix shell in Terminal, where it remotely communicates with the podman service running in the Machine VM.

Download Podman Installer (Recommended)

Podman can be downloaded from the website.

We also upload the installers and other binaries on our Github release page.

Though not recommended, Podman can also be obtained through Homebrew, the package manager.

Install via Brew

Since Brew is a community-maintained package manager, we cannot guarantee stability of Brew installs of Podman. Thus, installing via Brew is not recommended.

However, if you do wish to use Brew, you must first install Homebrew. Once you have set up brew, you can use the brew install command to install Podman:

brew install podman

After installing, you need to create and start your first Podman machine:

podman machine init
podman machine start

You can then verify the installation information using:

podman info

We also provide binaries and a pkginstaller on our Github release page


On Windows, each Podman machine is backed by a virtualized Windows Subsystem for Linux (WSLv2) distribution. Once installed, the podman command can be run directly from your Windows PowerShell (or CMD) prompt, where it remotely communicates with the podman service running in the WSL environment. Alternatively, you can access Podman directly from the WSL instance if you prefer a Linux prompt and Linux tooling.

See the Podman for Windows guide for setup and usage instructions.

Installing on Linux

Linux Distributions

Arch Linux & Manjaro Linux

sudo pacman -S podman

If you have problems when running Podman in rootless mode follow the instructions here

For more information on Podman on ArchLinux click here

Alpine Linux

sudo apk add podman

For further details, please refer to the instructions on the Alpine Linux wiki.


Podman is available in the default Extras repos for CentOS 7 and in the AppStream repo for CentOS 8 and Stream.

sudo yum -y install podman


The podman package is available in the Debian 11 (Bullseye) repositories and later.

sudo apt-get -y install podman


sudo dnf -y install podman

Fedora CoreOS, Fedora Silverblue

Built-in, no need to install


sudo emerge app-containers/podman


Bitbake recipes for Podman and its dependencies are available in the meta-virtualization layer. Add the layer to your OpenEmbedded build environment and build Podman using:

bitbake podman


sudo zypper install podman

openSUSE Kubic

Built-in, no need to install

Raspberry Pi OS arm64 (beta)

Raspberry Pi OS use the standard Debian repositories, so it is fully compatible with Debian's arm64 repository. You can simply follow the steps for Debian to install Podman.


Subscribe, then enable Extras channel and install Podman.

sudo subscription-manager repos --enable=rhel-7-server-extras-rpms
sudo yum -y install podman


Podman is included in the container-tools module, along with Buildah and Skopeo.

sudo yum module enable -y container-tools:rhel8
sudo yum module install -y container-tools:rhel8

The container-tools:rhel8 is the fast application stream, containing most recent rolling versions of the tools. Use the container-tools:2.0 stream for stable versions of Podman 1.6. The command yum module list container-tools shows the available streams.


The podman package is available in the official repositories for Ubuntu 20.10 and newer.

# Ubuntu 20.10 and newer
sudo apt-get update
sudo apt-get -y install podman

Linux Mint

Follow the steps for Ubuntu (or Debian if you use LMDE).

Replace $(lsb_release -rs) with $(grep DISTRIB_RELEASE= /etc/upstream-release/lsb-release | cut -d "=" -f 2) for Ubuntu steps.

Installing development versions of Podman


You can test the very latest Podman in Fedora's updates-testing repository before it goes out to all Fedora users.

sudo dnf update --refresh --enablerepo=updates-testing podman

If you use a newer Podman package from Fedora's updates-testing, we would appreciate your +1 feedback in Bodhi, Fedora's update management system.

Installing bleeding-edge versions of Podman

If you like danger and are interested in testing the latest unreleased bits of Podman on Fedora, CentOS and RHEL, we have a Copr repository.

CAUTION: This repository contains rpm builds generated using the main branch of upstream container tools repositories, and simply CANNOT be recommended for any production use.

RHEL8 / CentOS 8 Stream users would first need to disable the container-tools module. All other users can skip this step.

sudo dnf module disable container-tools -y

Enable the Copr and install podman.

sudo dnf copr enable rhcontainerbot/podman-next -y
sudo dnf install podman

Installing on FreeBSD 14.0

[!WARNING] The FreeBSD port of the Podman container engine is experimental and should be used for evaluation and testing purposes only.

You can install Podman on FreeBSD using pkg:

pkg install podman

There's also a podman-suite meta package that will pull additional packages for you (buildah, skopeo).

Initial configuration

To properly support Podman's container restart policy, conmon needs fdescfs(5) to be mounted on /dev/fd.

If /dev/fd is not already mounted:

mount -t fdescfs fdesc /dev/fd

To make it permanent, add the following line to /etc/fstab:

fdesc   /dev/fd         fdescfs         rw      0       0

To start Podman after reboot:

service podman enable

Container networking relies on NAT to allow container network packets out to the host's network. This requires a PF firewall to perform the translation. A simple example is included - to use it:

cp /usr/local/etc/containers/pf.conf.sample /etc/pf.conf

Edit /etc/pf.conf and set v4egress_if, v6egress_if variables to your network interface(s)s

Enable and start pf:

service pf enable
service pf start

The sample PF configuration includes support for port redirections. These are implemented as redirect rules in anchors nested under cni-rdr.

Support for redirecting connections from the container host to services running inside a container is included for FreeBSD 13.3 and later. To enable this, first load the pf kernel module and enable PF support for these redirections using sysctl:

echo 'pf_load="YES"' >> /boot/loader.conf
kldload pf
echo '' >> /etc/sysctl.conf.local
service pf restart

Redirect rules will work if the destination address is localhost (e.g. or ::1) - to enable this, the following line must be included in your /etc/pf.conf:

nat-anchor "cni-rdr/*"

if upgrading from an older version, this needs to be added to /etc/pf.conf.

For example if host port 1234 is redirected to an http service running in a container, you could connect to it using:

fetch -o- http://$(hostname):1234


fetch -o- http://localhost:1234

Container images and related state is stored in /var/db/containers. It is recommended to use ZFS for this:

zfs create -o mountpoint=/var/db/containers zroot/containers

If your system cannot use ZFS, change storage.conf to use the vfs storage driver:

sed -I .bak -e 's/driver = "zfs"/driver = "vfs"/' /usr/local/etc/containers/storage.conf

After following these steps you should be able to run native images:

podman run --rm
Linux Emulation

It is possible to run many Linux container images using FreeBSD's Linux emulation:

sudo sysrc linux_enable=YES
sudo service linux start
sudo podman run --rm --os=linux alpine cat /etc/os-release | head -1
NAME="Alpine Linux"

Building from Source

Build and Run Dependencies


On Fedora:

# Install build dependencies
sudo dnf -y builddep rpm/podman.spec

# Install runtime dependencies
sudo dnf -y install catatonit conmon containers-common-extra

On all RHEL and CentOS Stream, first install dnf-builddep:

sudo dnf -y install 'dnf-command(builddep)'

Install build dependencies:

# CentOS Stream 8
sudo dnf -y builddep rpm/podman.spec --enablerepo=powertools

# CentOS Stream 9
sudo dnf -y builddep rpm/podman.spec --enablerepo=crb

# RHEL (8 and newer)
sudo dnf -y builddep rpm/podman.spec --enablerepo=codeready-builder-for-rhel-$(rpm --eval %{?rhel})-$(uname -m)-rpms

Install runtime dependencies:

sudo dnf -y install \
conmon \
containers-common \
crun \
iptables \
netavark \
nftables \

Debian, Ubuntu, and related distributions:

sudo apt-get install \
btrfs-progs \
crun \
git \
golang-go \
go-md2man \
iptables \
libassuan-dev \
libbtrfs-dev \
libc6-dev \
libdevmapper-dev \
libglib2.0-dev \
libgpgme-dev \
libgpg-error-dev \
libprotobuf-dev \
libprotobuf-c-dev \
libseccomp-dev \
libselinux1-dev \
libsystemd-dev \
netavark \
pkg-config \

The netavark package may not be available on older Debian / Ubuntu versions. Install the containernetworking-plugins package instead.

On openSUSE Leap 15.x and Tumbleweed:

sudo zypper -n in libseccomp-devel libgpgme-devel

On Manjaro (and maybe other Linux distributions):

Make sure that the Linux kernel supports user namespaces:

> zgrep CONFIG_USER_NS /proc/config.gz

If not, please update the kernel. For Manjaro Linux the instructions can be found here:

After that enable user namespaces:

sudo sysctl kernel.unprivileged_userns_clone=1

To enable the user namespaces permanently:

echo 'kernel.unprivileged_userns_clone=1' > /etc/sysctl.d/userns.conf

Building missing dependencies

If any dependencies cannot be installed or are not sufficiently current, they have to be built from source. This will mainly affect Debian, Ubuntu, and related distributions, or RHEL where no subscription is active (e.g. Cloud VMs).


Be careful to double-check that the version of golang is new enough (i.e. go version), as of January 2022 version is 1.16.x or higher is required. The current minimum required version can always be found in the go.mod file. If needed, golang kits are available at Alternatively, go can be built from source as follows (it's helpful to leave the system-go installed, to avoid having to bootstrap go:

export GOPATH=~/go
git clone $GOPATH
cd src
export PATH=$GOPATH/bin:$PATH


The latest version of conmon is expected to be installed on the system. Conmon is used to monitor OCI Runtimes. To build from source, use the following:

git clone
cd conmon
export GOCACHE="$(mktemp -d)"
sudo make podman

crun / runc

The latest version of at least one container runtime is expected to be installed on the system. crun or runc are some of the possibilities, and one is picked up as the default runtime by Podman (crun has priority over runc). Supported versions of crun or runc are available for example on Ubuntu 22.04. runc version 1.0.0-rc4 is the minimal requirement, which is available since Ubuntu 18.04.

To double-check, runc --version should produce at least spec: 1.0.1, otherwise build your own:

git clone $GOPATH/src/
cd $GOPATH/src/
make BUILDTAGS="selinux seccomp"
sudo cp runc /usr/bin/runc

Add configuration

sudo mkdir -p /etc/containers
sudo curl -L -o /etc/containers/registries.conf
sudo curl -L -o /etc/containers/policy.json

Optional packages

Fedora, CentOS, RHEL, and related distributions:

(no optional packages)

Debian, Ubuntu, and related distributions:

apt-get install -y \

Get Source Code

First, ensure that the go version that is found first on the $PATH is 1.16.x or higher. Instruction above will help you compile newer version of Go if needed. Then we can build Podman:

git clone
cd podman
make BUILDTAGS="selinux seccomp" PREFIX=/usr
sudo make install PREFIX=/usr

Build Tags

Otherwise, if you do not want to build Podman with seccomp or selinux support you can add BUILDTAGS="" when running make.

sudo make install

Podman supports optional build tags for compiling support of various features. To add build tags to the make option the BUILDTAGS variable must be set, for example:

make BUILDTAGS='seccomp apparmor'
Build TagFeatureDependency
apparmorapparmor supportlibapparmor
cniCNI networking
exclude_graphdriver_btrfsexclude btrfslibbtrfs
exclude_graphdriver_devicemapperexclude device-mapperlibdm
libdm_no_deferred_removeexclude deferred removal in libdmlibdm
seccompsyscall filteringlibseccomp
selinuxselinux process and mount labeling
systemdjournald logginglibsystemd

Note that Podman does not officially support device-mapper. Thus, the exclude_graphdriver_devicemapper tag is mandatory.

Vendoring - Dependency Management

This project is using go modules for dependency management. If the CI is complaining about a pull request leaving behind an unclean state, it is very likely right about it. After changing dependencies, make sure to run make vendor to synchronize the code with the go module and repopulate the ./vendor directory.


An Ansible Role is also available to automate the installation of the above statically linked binary on its supported OS:

sudo su -
mkdir -p ~/.ansible/roles
cd ~/.ansible/roles
git clone podman
cd ~/.ansible/roles/podman
pip3 install --upgrade --ignore-installed --requirement requirements.txt
molecule converge
molecule verify

Configuration files


Man Page: registries.conf.5


registries.conf is the configuration file which specifies which container registries should be consulted when completing image names which do not include a registry or domain portion.

Example from the Fedora containers-common package

$ cat /etc/containers/registries.conf
# For more information on this configuration file, see containers-registries.conf(5).
# We recommend always using fully qualified image names including the registry
# server (full dns name), namespace, image name, and tag
# (e.g., Pulling by digest (i.e.,
# further eliminates the ambiguity of tags.
# When using short names, there is always an inherent risk that the image being
# pulled could be spoofed. For example, a user wants to pull an image named
# `foobar` from a registry and expects it to come from If
# is not first in the search list, an attacker could place a
# different `foobar` image at a registry earlier in the search list. The user
# would accidentally pull and run the attacker's image and code rather than the
# intended content. We recommend only adding registries which are completely
# trusted (i.e., registries which don't allow unknown or anonymous users to
# create accounts with arbitrary names). This will prevent an image from being
# spoofed, squatted or otherwise made insecure. If it is necessary to use one
# of these registries, it should be added at the end of the list.
# # An array of host[:port] registries to try when pulling an unqualified image, in order.
unqualified-search-registries = ["", "", ""]
# [[registry]]
# # The "prefix" field is used to choose the relevant [[registry]] TOML table;
# # (only) the TOML table with the longest match for the input image name
# # (taking into account namespace/repo/tag/digest separators) is used.
# #
# # If the prefix field is missing, it defaults to be the same as the "location" field.
# prefix = ""
# # If true, unencrypted HTTP as well as TLS connections with untrusted
# # certificates are allowed.
# insecure = false
# # If true, pulling images with matching names is forbidden.
# blocked = false
# # The physical location of the "prefix"-rooted namespace.
# #
# # By default, this equal to "prefix" (in which case "prefix" can be omitted
# # and the [[registry]] TOML table can only specify "location").
# #
# # Example: Given
# # prefix = ""
# # location = ""
# # requests for the image will actually work with the
# # image.
# location = ""
# # (Possibly-partial) mirrors for the "prefix"-rooted namespace.
# #
# # The mirrors are attempted in the specified order; the first one that can be
# # contacted and contains the image will be used (and if none of the mirrors contains the image,
# # the primary location specified by the "registry.location" field, or using the unmodified
# # user-specified reference, is tried last).
# #
# # Each TOML table in the "mirror" array can contain the following fields, with the same semantics
# # as if specified in the [[registry]] TOML table directly:
# # - location
# # - insecure
# [[registry.mirror]]
# location = "example-mirror-0.local/mirror-for-foo"
# [[registry.mirror]]
# location = "example-mirror-1.local/mirrors/foo"
# insecure = true
# # Given the above, a pull of will try:
# # 1. example-mirror-0.local/mirror-for-foo/image:latest
# # 2. example-mirror-1.local/mirrors/foo/image:latest
# # 3.
# # in order, and use the first one that exists.
# short-name-mode="enforcing"



/usr/share/containers/mounts.conf and optionally /etc/containers/mounts.conf

The mounts.conf files specify volume mount directories that are automatically mounted inside containers when executing the podman run or podman build commands. Container process can then use this content. The volume mount content does not get committed to the final image.

Usually these directories are used for passing secrets or credentials required by the package software to access remote package repositories.

For example, a mounts.conf with the line "/usr/share/rhel/secrets:/run/secrets", the content of /usr/share/rhel/secrets directory is mounted on /run/secrets inside the container. This mountpoint allows Red Hat Enterprise Linux subscriptions from the host to be used within the container.

Note this is not a volume mount. The content of the volumes is copied into container storage, not bind mounted directly from the host.

Example from the Fedora containers-common package:

cat /usr/share/containers/mounts.conf



seccomp.json contains the whitelist of seccomp rules to be allowed inside of containers. This file is usually provided by the containers-common package.

The link above takes you to the seccomp.json



Man Page: policy.json.5

Example from the Fedora containers-common package:

cat /etc/containers/policy.json
"default": [
"type": "insecureAcceptAnything"
"": [{"type":"insecureAcceptAnything"}]