podman-generate-systemd - Man Page

A Kubernetes YAML can be executed in systemd via the podman-kube@.service systemd template.  The template's argument is the path to the YAML file.  Given a workload.yaml file in the home directory, it can be executed as follows:

$ escaped=$(systemd-escape ~/workload.yaml)
$ systemctl --user start podman-kube@$escaped.service
$ systemctl --user is-active podman-kube@$escaped.service
active

Add the systemd unit after (After=) option, that ordering dependencies between the list of dependencies and this service. This option may be specified more than once.

User-defined dependencies are appended to the generated unit file, but any existing options needed or defined by default (e.g., online.target) are not removed or overridden.

Set the systemd unit name prefix for containers. The default is container.

Set environment variables to the systemd unit files.

If an environment variable is specified without a value, Podman checks the host environment for a value and sets the variable only if it is set on the host. As a special case, if an environment variable ending in * is specified without a value, Podman searches the host environment for variables starting with the prefix and adds those variables to the systemd unit files.

Generate files instead of printing to stdout.  The generated files are named {container,pod}-{ID,name}.service and are placed in the current working directory.

Note: On a system with SELinux enabled, the generated files inherits contexts from the current working directory. Depending on the SELinux setup, changes to the generated files using restorecon, chcon, or semanage may be required to allow systemd to access these files. Alternatively, use the -Z option when running mv or cp.

Print the created units in the specified format (json). If --files is specified, the paths to the created files are printed instead of the unit content.

Use the name of the container for the start, stop, and description in the unit file

This option yields unit files that do not expect containers and pods to exist.  Instead, new containers and pods are created based on their configuration files.  The unit files are created best effort and may need further editing; please review the generated files carefully before using them in production.

Note that --new only works on containers and pods created directly via Podman (i.e., podman [container] {create,run} or podman pod create).  It does not work on containers or pods created via the REST API or via podman kube play.  For podman kube play, use the podman-kube@.service systemd template instead.

Do not generate the header including meta data such as the Podman version and the timestamp.

Set the systemd unit name prefix for pods. The default is pod.

Set the systemd unit requires (Requires=) option. Similar to wants, but declares a stronger requirement dependency.

Set the systemd restart policy.  The restart-policy must be one of: "no", "on-success", "on-failure", "on-abnormal", "on-watchdog", "on-abort", or "always".  The default policy is on-failure unless the container was created with a custom restart policy.

Note that generating a unit without --new on a container with a custom restart policy can lead to issues on shutdown; systemd attempts to stop the unit while Podman tries to restart it.  Creating the container without --restart and using the --restart-policy option when generating the unit file is recommended.

Set the systemd service restartsec value. Configures the time to sleep before restarting a service (as configured with restart-policy). Takes a value in seconds.

Set the systemd unit name separator between the name/id of a container/pod and the prefix. The default is -.

Override the default start timeout for the container with the given value in seconds.

Override the default stop timeout for the container with the given value in seconds.

Add template specifiers to run multiple services from the systemd unit file.

Note that if --new was not set to true, it is set to true by default. However, the command fails if --new is set to false explicitly.

Add the systemd unit wants (Wants=) option, that this service is (weak) dependent on. This option may be specified more than once. This option does not influence the order in which services are started or stopped.

User-defined dependencies are appended to the generated unit file, but any existing options needed or defined by default (e.g., online.target) are not removed or overridden.

Generate a systemd unit file for a container running nginx with an always restart policy and 1-second timeout to stdout. Note that the RequiresMountsFor option in the Unit section ensures that the container storage for both the GraphRoot and the RunRoot are mounted prior to starting the service. For systems with container storage on disks like iSCSI or other remote block protocols, this ensures that Podman is not executed prior to any necessary storage operations coming online.

$ podman create --name nginx nginx:latest
$ podman generate systemd --restart-policy=always -t 1 nginx
# container-de1e3223b1b888bc02d0962dd6cb5855eb00734061013ffdd3479d225abacdc6.service
# autogenerated by Podman 1.8.0
# Wed Mar 09 09:46:45 CEST 2020

[Unit]
Description=Podman container-de1e3223b1b888bc02d0962dd6cb5855eb00734061013ffdd3479d225abacdc6.service
Documentation=man:podman-generate-systemd(1)
Wants=network-online.target
After=network-online.target
RequiresMountsFor=/var/run/container/storage

[Service]
Restart=always
ExecStart=/usr/bin/podman start de1e3223b1b888bc02d0962dd6cb5855eb00734061013ffdd3479d225abacdc6
ExecStop=/usr/bin/podman stop   
        -t 1 de1e3223b1b888bc02d0962dd6cb5855eb00734061013ffdd3479d225abacdc6
KillMode=none
Type=forking
PIDFile=/run/user/1000/overlay-containers/de1e3223b1b888bc02d0962dd6cb5855eb00734061013ffdd3479d225abacdc6/userdata/conmon.pid

[Install]
WantedBy=default.target

The --new flag generates systemd unit files that create and remove containers at service start and stop commands (see ExecStartPre and ExecStopPost service actions). Such unit files are not tied to a single machine and can easily be shared and used on other machines.

$ sudo podman generate systemd --new --files --name bb310a0780ae
# container-busy_moser.service
# autogenerated by Podman 1.8.3
# Fri Apr  3 09:40:47 EDT 2020

[Unit]
Description=Podman container-busy_moser.service
Documentation=man:podman-generate-systemd(1)
Wants=network-online.target
After=network-online.target
RequiresMountsFor=/var/run/container/storage

[Service]
Environment=PODMAN_SYSTEMD_UNIT=%n
Restart=on-failure
ExecStartPre=/bin/rm -f %t/%n-pid %t/%n-cid
ExecStart=/usr/local/bin/podman run   
        --conmon-pidfile %t/%n-pid   
	--cidfile %t/%n-cid   
	--cgroups=no-conmon   
	-d   
	-dit alpine
ExecStop=/usr/local/bin/podman stop   
        --ignore   
        --cidfile %t/%n-cid -t 10
ExecStopPost=/usr/local/bin/podman rm   
        --ignore   
        -f   
	--cidfile %t/%n-cid
PIDFile=%t/%n-pid
KillMode=none
Type=forking

[Install]
WantedBy=default.target

Note systemctl must only be used on the pod unit and not used to start or stop containers individually. The containers are managed by the pod service along with the internal infra-container.

Use systemctl status or journalctl to examine container or pod unit files.

$ podman pod create --name systemd-pod
$ podman create --pod systemd-pod alpine top
$ podman create --pod systemd-pod alpine top
$ podman generate systemd --files --name systemd-pod
/home/user/pod-systemd-pod.service
/home/user/container-amazing_chandrasekhar.service
/home/user/container-jolly_shtern.service
$ cat pod-systemd-pod.service
# pod-systemd-pod.service
# autogenerated by Podman 1.8.0
# Wed Mar 09 09:52:37 CEST 2020

[Unit]
Description=Podman pod-systemd-pod.service
Documentation=man:podman-generate-systemd(1)
Requires=container-amazing_chandrasekhar.service container-jolly_shtern.service
Before=container-amazing_chandrasekhar.service container-jolly_shtern.service
Wants=network-online.target
After=network-online.target
RequiresMountsFor=/var/run/container/storage

[Service]
Restart=on-failure
ExecStart=/usr/bin/podman start 77a818221650-infra
ExecStop=/usr/bin/podman stop   
        -t 10 77a818221650-infra
KillMode=none
Type=forking
PIDFile=/run/user/1000/overlay-containers/ccfd5c71a088768774ca7bd05888d55cc287698dde06f475c8b02f696a25adcd/userdata/conmon.pid

[Install]
WantedBy=default.target

Podman-generated unit files include an [Install] section, which carries installation information for the unit. It is used by the enable and disable commands of systemctl(1) during installation.

Once the systemd unit file is generated, install it to /etc/systemd/system to be run by the root user or to $HOME/.config/systemd/user for installing it as a non-root user. Enable the copied unit file or files using systemctl enable.

Note: Copying unit files to /etc/systemd/system and enabling it marks the unit file to be automatically started at boot. And similarly, copying a unit file to $HOME/.config/systemd/user and enabling it marks the unit file to be automatically started on user login.

# Generated systemd files.
$ podman pod create --name systemd-pod
$ podman create --pod systemd-pod alpine top
$ podman generate systemd --files --name systemd-pod

# Copy all the generated files.

$ sudo cp pod-systemd-pod.service container-great_payne.service /etc/systemd/system
$ systemctl enable pod-systemd-pod.service
Created symlink /etc/systemd/system/multi-user.target.wants/pod-systemd-pod.service → /etc/systemd/system/pod-systemd-pod.service.
Created symlink /etc/systemd/system/default.target.wants/pod-systemd-pod.service → /etc/systemd/system/pod-systemd-pod.service.
$ systemctl is-enabled pod-systemd-pod.service
enabled

To run the user services placed in $HOME/.config/systemd/user on first login of that user, enable the service with --user flag.

$ systemctl --user enable <.service>

The systemd user instance is killed after the last session for the user is closed. The systemd user instance can be started at boot and kept running even after the user logs out by enabling lingering using

$ loginctl enable-linger <username>

Create and enable systemd unit files for a pod using the above examples as reference and use systemctl to perform operations.

Since systemctl defaults to using the root user, all the changes using the systemctl can be seen by appending sudo to the podman cli commands. To perform systemctl actions as a non-root user use the --user flag when interacting with systemctl.

Note: If the previously created containers or pods are using shared resources, such as ports, make sure to remove them before starting the generated systemd units.

$ systemctl --user start pod-systemd-pod.service
$ podman pod ps
POD ID         NAME          STATUS    CREATED          # OF CONTAINERS   INFRA ID
0815c7b8e7f5   systemd-pod   Running   29 minutes ago   2                 6c5d116f4bbe
$ sudo podman ps # 0 Number of pods on root.
CONTAINER ID  IMAGE  COMMAND  CREATED  STATUS  PORTS  NAMES
$ systemctl stop pod-systemd-pod.service
$ podman pod ps
POD ID         NAME          STATUS   CREATED          # OF CONTAINERS   INFRA ID
272d2813c798   systemd-pod   Exited   29 minutes ago   2                 6c5d116f4bbe

Create a simple alpine container and generate the systemd unit file with --new flag. Enable the service and control operations using the systemctl commands.

Note: When starting the container using systemctl start rather than altering the already running container it spins up a "new" container with similar configuration.

# Enable the service.

$ sudo podman ps -a
CONTAINER ID  IMAGE                            COMMAND  CREATED        STATUS     PORTS  NAMES
bb310a0780ae  docker.io/library/alpine:latest  /bin/sh  2 minutes ago  Created           busy_moser
$ sudo systemctl start container-busy_moser.service
$ sudo podman ps -a
CONTAINER ID  IMAGE                            COMMAND  CREATED        STATUS        PORTS      NAMES
772df2f8cf3b  docker.io/library/alpine:latest  /bin/sh  1 second ago   Up 1 second              distracted_albattani
bb310a0780ae  docker.io/library/alpine:latest  /bin/sh  3 minutes ago  Created                  busy_moser