SYSTEMD-VMSPAWN(1) systemd-vmspawn SYSTEMD-VMSPAWN(1)

systemd-vmspawn - Spawn an OS in a virtual machine

systemd-vmspawn [OPTIONS...] [ARGS...]

systemd-vmspawn may be used to start a virtual machine from an OS image. In many ways it is similar to systemd-nspawn(1), but launches a full virtual machine instead of using namespaces.

File descriptors for /dev/kvm and /dev/vhost-vsock can be passed to systemd-vmspawn via systemd's native socket passing interface (see sd_listen_fds(3) for details about the precise protocol used and the order in which the file descriptors are passed), these file descriptors must be passed with the names "kvm" and "vhost-vsock" respectively.

Note: on Ubuntu/Debian derivatives systemd-vmspawn requires the user to be in the "kvm" group to use the VSOCK options.

The excess arguments are passed as extra kernel command line arguments using SMBIOS.

The following options are understood:

-q, --quiet

Turns off any status output by the tool itself. When this switch is used, the only output from vmspawn will be the console output of the Virtual Machine OS itself.

Added in version 256.

--system, --user

Specify whether to interact with the user manager or the system manager and whether to register with the user machined instance or the system machined instance. If unspecified, the system manager and machined instance will be used when running as root, otherwise the user manager and machined instance will be used.

Added in version 260.

-D, --directory=

Directory to use as file system root for the virtual machine.

One of either --directory= or --image= must be specified. If neither are specified --directory=. is assumed.

Note: If mounting a non-root owned directory you may require --private-users= to map into the user's subuid namespace. An example of how to use /etc/subuid for this is given later.

Added in version 256.

-x, --ephemeral

If specified, the VM is run with a temporary snapshot of its file system that is removed immediately when the VM terminates. Only works with --image= currently. Note that --ephemeral will not work with --extra-drive=.

Added in version 260.

-i, --image=

Root file system disk image (or device node) for the virtual machine.

Added in version 255.

--image-format=FORMAT

Specifies the format of the disk image passed to --image=. Takes either "raw" or "qcow2". Defaults to "raw". Note that "qcow2" is only supported for regular files, not block devices.

Added in version 260.

--image-disk-type=TYPE

Specifies the disk type to use for the root disk passed to --image=. Extra drives added via --extra-drive= inherit this disk type unless overridden with an explicit disk type prefix. Takes one of "virtio-blk", "virtio-scsi", "nvme", or "scsi-cd". Defaults to "virtio-blk". When "scsi-cd" is specified, the disk is attached as a read-only CD-ROM drive.

Added in version 261.

--discard-disk=BOOL

Controls whether qemu processes discard requests from the VM. This prevents long running VMs from using more disk space than required. This is enabled by default.

Added in version 256.

--grow-image=BYTES, -G BYTES

Grows the image file specified by --image= to the specified size in bytes if it is smaller. Executes no operation if no image file is used or the image file is already as large or larger than requested. The specified size accepts the usual K, M, G suffixes (to the base of 1024). Specified values are rounded up to multiples of 4096.

If --ephemeral is specified, the original image file is left untouched and the requested size is applied to the ephemeral copy instead. In that case qcow2 images are supported too.

Added in version 258.

--cpus=CPUS

The number of CPUs to start the virtual machine with. Defaults to 1.

Added in version 255.

--ram=BYTES[:MAXBYTES[:SLOTS]]

The amount of memory to start the virtual machine with. Defaults to 2G. If a maximum size is specified after a colon, memory hotplug is enabled with the given upper limit. The number of hotplug slots can optionally be specified after a second colon and defaults to 1.

Added in version 255.

--kvm=BOOL

Controls whether to enable KVM acceleration. If not specified or set to auto, KVM support will be detected automatically. If true, KVM is insisted on. If false, disable KVM.

Added in version 255.

--vsock=BOOL

Controls whether to allocate a VSOCK socket for guest. If not specified or set to auto, VSOCK networking support will be detected automatically. If true, VSOCK is insisted on. If false, VSOCK networking is disabled.

Added in version 255.

--vsock-cid=CID

Sets the specific CID to use for the guest. Valid CIDs are in the range 3 to 4294967294 (0xFFFF_FFFE). CIDs outside of this range are reserved. By default, vmspawn will attempt to derive a CID for the guest derived from the machine name, falling back to a random CID if this CID is taken.

Added in version 255.

--tpm=BOOL

Controls whether to serve a TPM device for guest, via swtpm(8). If not specified or set to auto, vmspawn will detect the presence of swtpm binary automatically. If yes, swtpm support is insisted on. If no, TPM is disabled.

Added in version 256.

--tpm-state=PATH|auto|off

Configures where to place TPM state, in case TPM support is enabled (see --tpm= above). This takes an absolute file system path to a directory to persistently place the state in. If the directory is missing it is created as needed. If set to the special string "auto" a persistent path is automatically derived from the VM image path or directory path, with the ".tpmstate" suffix appended. If set to the special string "off" the TPM state is only maintained transiently and flushed out when the VM shuts down. This mode is not suitable for VMs which lock disk encryption keys to the TPM, as these keys will be lost on every reboot. Defaults to "auto".

If --ephemeral is specified, "auto" behaves like "off".

Added in version 258.

--efi-nvram-template=PATH

Takes an absolute path, or a relative path beginning with ./. Specifies the path to an EFI NVRAM template file to copy and use as the initial EFI variable NVRAM state. If not specified, the default NVRAM template from the firmware definition is copied and used. Cannot be used together with --coco=, as confidential computing firmware is stateless.

Added in version 261.

--efi-nvram-state=PATH|auto|off

Configures where to place the EFI variable NVRAM state. This takes an absolute file system path to a regular file to persistently place the state in. If the file is missing it is created as needed. If set to the special string "auto" a persistent path is automatically derived from the VM image path or directory path, with the ".efinvramstate" suffix appended. If set to the special string "off" the EFI variable NVRAM state is only maintained transiently and flushed out when the VM shuts down. Defaults to "auto". An explicit path cannot be used together with --coco=, as confidential computing firmware is stateless.

If --ephemeral is specified, "auto" behaves like "off".

Added in version 261.

--secure-boot=BOOL

Configure whether to search for firmware which supports Secure Boot. Takes a boolean or "auto". Setting this to yes is equivalent to --firmware-features=secure-boot and setting this to no is equivalent to --firmware-features=~secure-boot. Setting this to "auto" removes "secure-boot" from both the included and excluded feature lists.

With --coco=, setting this to "yes" additionally requires the "enrolled-keys" firmware feature, overriding its default exclusion: confidential computing firmware is stateless, so keys cannot be enrolled at runtime and Secure Boot is only operative with keys baked into the image at build time, which in turn makes the firmware refuse to boot unsigned images from the very first boot. The default exclusion of "enrolled-keys" hence keeps unsigned images bootable.

Added in version 255.

--firmware=PATH

Selects which firmware to use in the VM. Takes one of "auto", "uefi", "bios", "none", an absolute path, or a relative path beginning with ./. Defaults to "auto", which selects UEFI firmware unless --linux= specifies a non-PE kernel image, in which case "none" is selected. "uefi" loads OVMF firmware (use a path to a JSON firmware definition file to select a specific one). "bios" skips OVMF loading and lets QEMU use its built-in BIOS (e.g. SeaBIOS on x86). "none" disables firmware loading entirely and requires --linux= to be specified for direct kernel boot. Booting a UKI requires "uefi". If the special string "list" is specified, all discovered firmware definition files are listed. If the special string "describe" is specified, the UEFI firmware that would be selected (taking --firmware-features= and --coco= into account) is printed and the program exits. If an empty string is specified, the option is reset to its default.

Added in version 256.

--firmware-features=FEATURE[,FEATURE...]

Takes a comma-delimited list of firmware feature strings. This option may be specified multiple times, in which case the feature lists are combined. When specified, only firmware definitions that have all the required features will be considered during automatic firmware discovery. Features prefixed with "~" are excluded: firmware that has such a feature will be skipped. If a feature appears in both the included and excluded lists, inclusion takes priority. By default, firmware with the "enrolled-keys" feature is excluded. If an empty string is passed, both the included and excluded feature lists are reset. If the special string "list" is specified, lists all available firmware features.

Added in version 261.

--coco=

Caveat: This feature is experimental, and is likely to be changed (or removed in its current form) in a future version of systemd.

Configures whether to run the guest as a confidential VM. Takes one of "no", "sev-snp" or "tdx". Defaults to "no".

"sev-snp" enables AMD SEV-SNP. This requires KVM on an x86_64 host with SNP-capable hardware and firmware. A suitable SNP-built OVMF firmware is picked automatically from the installed QEMU firmware descriptors, by requiring the "amd-sev-snp" firmware feature; use --firmware= with a path to a firmware descriptor file to select a specific one. Secure Boot is effectively off by default (see --secure-boot=). Direct kernel boot via --linux= is required so that the kernel, initrd and command line are hashed into the launch measurement ("kernel-hashes=on"); booting the kernel off the disk image via the firmware would leave it outside the measurement. Credentials passed via --set-credential= or --load-credential= are bundled into a cpio archive appended to the initrd (mirroring what systemd-stub does for ESP credentials), so they enter the launch measurement via "kernel-hashes=on"; the SMBIOS and fw_cfg channels normally used to deliver credentials are not used because they are unmeasured and would be discarded by PID1 in confidential guests. This channel is measured but not confidential with respect to the host or VMM: the initrd (and thus the credentials it carries) is supplied to QEMU as plaintext and only its hash enters the launch measurement, which guarantees integrity but does not keep the credentials secret from the host. This requires the guest to run a sufficiently recent version of systemd (supporting /.extra/system_credentials/). A vTPM, if attached via --tpm=, must be treated as untrusted by the guest.

"tdx" enables Intel TDX. This requires KVM on an x86_64 host with TDX-capable hardware and a TDX-enabled host kernel. As with "sev-snp", a TDX-built OVMF (TDVF) firmware is picked automatically from the installed QEMU firmware descriptors, by requiring the "intel-tdx" firmware feature; use --firmware= with a path to a firmware descriptor file to select a specific one. The CPU model is fixed to "host". Firmware is measured into MRTD when the TD is built. Secure Boot cannot be enrolled at runtime (there is no writable NVRAM); its state is fixed by the selected TDVF image and is part of the measured firmware (see --secure-boot=). When booting a UKI, the whole UKI PE is measured into RTMR 1, and the loaded sections are measured individually by systemd-stub into RTMR 2. For direct linux boot, firmware measures the kernel PE into RTMR 1, and the Linux EFI stub measures initrd and command line into RTMR 2. Credentials passed via --set-credential= or --load-credential= are delivered through SMBIOS Type 11 OEM strings, which the firmware (TDVF) measures into RTMR 0. This channel is measured but not confidential with respect to the host or VMM, since the host assembles the SMBIOS table. A vTPM, if attached via --tpm=, must be treated as untrusted by the guest. To obtain TD Quotes for remote attestation, the guest is wired to the host's local TDX Quote Generation Service automatically: the unix socket /run/tdx-qgs/qgs.socket is used if it exists, otherwise vsock port 4050 on the host (cid 2). If the QGS is listening on neither channel, the guest's quote requests will fail. Added in version 262.

Added in version 261.

-n, --network-tap

Create a TAP device to network with the virtual machine.

Note: root privileges are required to use TAP networking. Additionally, systemd-networkd(8) must be running and correctly set up on the host to provision the host interface. The relevant ".network" file can be found at /usr/lib/systemd/network/80-vm-vt.network.

Added in version 255.

--network-user-mode

Use user mode networking.

Added in version 255.

--linux=PATH

Set the linux kernel image to use for direct kernel boot. If a directory type image is used and --linux= was omitted, vmspawn will search for boot loader entries according to the UAPI.1 Boot Loader Specification[1] assuming XBOOTLDR to be located at /boot and ESP to be /efi respectively. If no kernel was installed into the image then the image will fail to boot.

Added in version 256.

--initrd=PATH

Set the initrd to use for direct kernel boot. If the --linux= supplied is a UAPI.1 Boot Loader Specification[1] Type #2 entry, then this argument is not required. If no initrd was installed into the image then the image will fail to boot.

--initrd= can be specified multiple times and vmspawn will merge them together.

Added in version 256.

--smbios11=STRING, -s STRING

Passes the specified string into the VM as an SMBIOS Type #11 vendor string. This is useful to parameterize the invoked VM in various ways. See smbios-type-11(7) for details.

Added in version 258.

--notify-ready=

Configures support for notifications from the VM's init process to systemd-vmspawn. If true, systemd-vmspawn will consider the machine as ready only when it has received a "READY=1" message from the init process in the VM. If false, systemd-vmspawn will consider the machine as ready immediately after creation. In either case, systemd-vmspawn sends its own readiness notification to its manager after the spawned VM is ready. For more details about notifications see sd_notify(3).

Defaults to true. (Note that this is unlike the option of the same name to systemd-nspawn(1) that defaults to false.)

Added in version 258.

-M, --machine=

Sets the machine name for this virtual machine. This name may be used to identify this virtual machine during its runtime (for example in tools like machinectl(1) and similar).

Added in version 255.

--uuid=

Set the specified UUID for the virtual machine. The init system will initialize /etc/machine-id from this if this file is not set yet. Note that this option takes effect only if /etc/machine-id in the virtual machine is unpopulated.

Added in version 256.

-S, --slice=

Make the VM part of the specified slice, instead of the default machine.slice. This applies only if the machine is run in its own scope unit, i.e. if --keep-unit is not used.

Added in version 258.

--property=

Set a unit property on the scope unit to register for the machine. This applies only if the machine is run in its own scope unit, i.e. if --keep-unit is not used. Takes unit property assignments in the same format as systemctl set-property. This is useful to set memory limits and similar for the VM.

Added in version 258.

--register=

Controls whether the virtual machine is registered with systemd-machined(8). Takes a boolean argument or "auto", and defaults to "auto". This ensures that the virtual machine is accessible via machinectl(1). When set to "auto", registration is attempted but failures are ignored.

Added in version 256.

--private-users=UID_SHIFT[:UID_RANGE]

Controls user namespacing under --directory=. If enabled, virtiofsd(1) is instructed to map user and group ids (UIDs and GIDs). This involves mapping the private UIDs/GIDs used in the virtual machine (starting with the virtual machine's root user 0 and up) to a range of UIDs/GIDs on the host that are not used for other purposes (usually in the range beyond the host's UID/GID 65536).

If one or two colon-separated numbers are specified, user namespacing is turned on. UID_SHIFT specifies the first host UID/GID to map, UID_RANGE is optional and specifies number of host UIDs/GIDs to assign to the virtual machine. If UID_RANGE is omitted, 65536 UIDs/GIDs are assigned.

When user namespaces are used, the GID range assigned to each virtual machine is always chosen identical to the UID range.

Added in version 256.

--bind=PATH, --bind-ro=PATH

Mount a directory from the host into the virtual machine. Takes one of: a path argument — in which case the specified path will be mounted from the host to the same path in the virtual machine, or a colon-separated pair of paths — in which case the first specified path is the source in the host, and the second path is the destination in the virtual machine. If the source path is not absolute, it is resolved relative to the current working directory. The --bind-ro= option creates read-only bind mounts. Backslash escapes are interpreted, so "\:" may be used to embed colons in either path. This option may be specified multiple times for creating multiple independent bind mount points.

Added in version 256.

--extra-drive=[FORMAT:][DISKTYPE:]PATH

Takes a disk image or block device on the host and supplies it to the virtual machine as another drive. Optionally, the image format and/or disk type can be specified by prefixing the path with their values separated by colons. The format and disk type prefixes can appear in any order. The format defaults to "raw" and the disk type defaults to the value of --image-disk-type= (which itself defaults to "virtio-blk"). Note that "qcow2" is only supported for regular files, not block devices.

Added in version 256.

--bind-volume=PROVIDER:VOLUME[:CONFIG][:K=V,...]

Acquire a storage volume from a storagectl(1) provider and attach it to the virtual machine. PROVIDER is the provider name (typically "block" or "fs"). VOLUME is the volume name passed to the provider's Acquire() method. CONFIG selects the guest device type and takes one of "virtio-blk", "virtio-scsi", "nvme", or "scsi-cd". If empty or omitted, defaults to "virtio-blk".

The trailing comma-separated K=V list passes parameters to io.systemd.StorageProvider.Acquire(): template=, create= (one of "any", "new", "open"), read-only= (or ro=; takes a boolean or "auto"), size= / create-size= (size for created volumes), request-as= (one of "blk", "reg", "dir"; "dir" is rejected by vmspawn).

Each attached volume is identified by the name "PROVIDER:VOLUME". Volumes attached at startup via this option cannot be detached at runtime via machinectl unbind-volume; only volumes added at runtime via machinectl bind-volume are removable.

The provider is looked up under /run/systemd/io.systemd.StorageProvider/ for system mode (or $XDG_RUNTIME_DIR/systemd/io.systemd.StorageProvider/ for user mode), matching the runtime scope chosen via --user / --system.

Added in version 261.

--bind-user=

Binds the home directory of the specified user on the host into the virtual machine. Takes the name of an existing user on the host as argument. May be used multiple times to bind multiple users into the virtual machine. This does two things:
1.The user's home directory is made available from the host into /run/vmhost/home/ using virtiofs. virtiofsd id translation to map the host user's UID/GID to its assigned UID/GID in the virtual machine.
2.JSON user and group records are generated in that describes the mapped user which are passed into the virtual machine using "userdb.transient.*" credentials. They contain a minimized representation of the host's user record, adjusted to the UID/GID and home directory path assigned to the user in the virtual machine. The nss-systemd(8) glibc NSS module will pick up these records from there and make them available in the virtual machine's user/group databases.

The combination of the two operations above ensures that it is possible to log into the virtual machine using the same account information as on the host. The user is only mapped transiently, while the virtual machine is running, and the mapping itself does not result in persistent changes to the virtual machine (except maybe for log messages generated at login time, and similar). Note that in particular the UID/GID assignment in the virtual machine is not made persistently. If the user is mapped transiently, it is best to not allow the user to make persistent changes to the virtual machine. If the user leaves files or directories owned by the user, and those UIDs/GIDs are reused during later virtual machine invocations (possibly with a different --bind-user= mapping), those files and directories will be accessible to the "new" user.

The user/group record mapping only works if the virtual machine contains systemd 258 or newer, with nss-systemd properly configured in nsswitch.conf. See nss-systemd(8) for details.

Note that the user record propagated from the host into the virtual machine will contain the UNIX password hash of the user, so that seamless logins in the virtual machine are possible. If the virtual machine is less trusted than the host it is hence important to use a strong UNIX password hash function (e.g. yescrypt or similar, with the "$y$" hash prefix).

Added in version 259.

--bind-user-shell=

When used with --bind-user=, includes the specified shell in the user records of users bound into the virtual machine. Takes either a boolean or an absolute path.
•If false (the default), no shell is passed in the user records for users bound into the virtual machine. This causes bound users to the use the virtual machine's default shell.
•If true, the shells specified by the host user records are included in the user records of all users bound into the virtual machine.
•If passed an absolute path, sets that path as the shell for user records of all users bound into the virtual machine.

Note: This will not check whether the specified shells exist in the virtual machine.

This operation is only supported in combination with --bind-user=.

Added in version 259.

--bind-user-group=NAME

When used with --bind-user=, includes the specified group as an auxiliary group in the user records of users bound into the virtual machine. Takes a group name.

Note: This will not check whether the specified groups exist in the virtual machine.

This operation is only supported in combination with --bind-user=.

Added in version 259.

--forward-journal=FILE|DIR

Forward the virtual machine's journal to the host. systemd-journal-remote(8) is currently used to receive the guest VM's forwarded journal entries. This option determines where this journal is saved on the host and has the same semantics as -o/--output described in systemd-journal-remote(8).

Added in version 256.

--forward-journal-max-use=BYTES, --forward-journal-keep-free=BYTES, --forward-journal-max-file-size=BYTES, --forward-journal-max-files=N

These options configure the corresponding settings of systemd-journal-remote(8) when forwarding journal entries from the VM. See journal-remote.conf(5) for the descriptions of these settings.

Added in version 261.

--pass-ssh-key=BOOL

By default, an SSH key is generated to allow systemd-vmspawn to open a D-Bus connection to the VM's systemd bus. Setting this to "no" will disable SSH key generation.

The generated keys are ephemeral. That is they are valid only for the current invocation of systemd-vmspawn, and are typically not persisted.

Added in version 256.

--ssh-key-type=TYPE

Configures the type of SSH key to generate, see ssh-keygen(1) for more information.

By default, "ed25519" keys are generated, however "rsa" keys may also be useful if the VM has a particularly old version of sshd(8).

Added in version 256.

--console=MODE

Configures how to set up the console of the VM. Takes one of "interactive", "read-only", "native", "gui", "headless". Defaults to "interactive". "interactive" provides an interactive terminal interface to the VM. "read-only" is similar, but is strictly read-only, i.e. does not accept any input from the user. "native" also provides a TTY-based interface, but uses qemu native implementation (which means the qemu monitor is available). "gui" shows the qemu graphical UI. "headless" runs the VM without any console, which is useful for automated or scripted usage.

Added in version 256.

--console-transport=TRANSPORT

Configures the transport to use for the VM console. Takes one of "virtio" or "serial". Defaults to "virtio". "virtio" uses a virtio-serial device, which appears as /dev/hvc0 in the VM. "serial" uses a regular serial port, which appears as /dev/ttyS0 (or /dev/ttyAMA0 on ARM) in the VM. This option only has an effect in --console=interactive, --console=read-only, and --console=native modes.

Added in version 261.

--background=COLOR

Change the terminal background color to the specified ANSI color as long as the VM runs. The color specified should be an ANSI X3.64 SGR background color, i.e. strings such as "40", "41", ..., "47", "48;2;...", "48;5;...". See ANSI Escape Code (Wikipedia)[2] for details. Assign an empty string to disable any coloring. This only has an effect in --console=interactive and --console=read-only modes.

Added in version 256.

--load-credential=ID:PATH, --set-credential=ID:VALUE

Pass a credential to the virtual machine. These two options correspond to the LoadCredential= and SetCredential= settings in unit files. See systemd.exec(5) for details about these concepts, as well as the syntax of the option's arguments.

In order to embed binary data into the credential data for --set-credential=, use C-style escaping (i.e. "\n" to embed a newline, or "\x00" to embed a NUL byte). Note that the invoking shell might already apply unescaping once, hence this might require double escaping!

Credentials are preferably passed to the VM via SMBIOS Type 11 strings or QEMU fw_cfg files. If neither mechanism is available, credentials are passed on the kernel command line using systemd.set_credential_binary= which is not a confidential channel. Do not use this for passing secrets to the VM in that case.

Under --coco=sev-snp, SMBIOS and fw_cfg are not covered by the SNP launch measurement and are discarded by PID1 in confidential guests. Credentials are therefore packaged into a cpio archive containing .extra/system_credentials/ID.cred entries and appended to the initrd that QEMU loads, so they enter the launch measurement via "kernel-hashes=on". PID1 imports them from the initramfs at boot. As with the kernel command line, this is a measured but not a confidential channel: QEMU receives the initrd (and thus the embedded credentials) as plaintext from the host and only its hash is covered by the launch measurement, so a modified initrd produces a different launch measurement that a relying party can detect via remote attestation, but the credentials are not hidden from the host or VMM. This requires the guest to run a sufficiently recent version of systemd (supporting /.extra/system_credentials/).

Added in version 255.

--no-pager

Do not pipe output into a pager.

-h, --help

Print a short help text and exit.

--version

Print a short version string and exit.

--no-ask-password

Do not query the user for authentication for privileged operations.

$SYSTEMD_LOG_LEVEL

The maximum log level of emitted messages (messages with a higher log level, i.e. less important ones, will be suppressed). Takes a comma-separated list of values. A value may be either one of (in order of decreasing importance) emerg, alert, crit, err, warning, notice, info, debug, or an integer in the range 0...7. See syslog(3) for more information. Each value may optionally be prefixed with one of console, syslog, kmsg or journal followed by a colon to set the maximum log level for that specific log target (e.g. SYSTEMD_LOG_LEVEL=debug,console:info specifies to log at debug level except when logging to the console which should be at info level). Note that the global maximum log level takes priority over any per target maximum log levels.

$SYSTEMD_LOG_COLOR

A boolean. If true, messages written to the tty will be colored according to priority.

This setting is only useful when messages are written directly to the terminal, because journalctl(1) and other tools that display logs will color messages based on the log level on their own.

$SYSTEMD_LOG_TIME

A boolean. If true, console log messages will be prefixed with a timestamp.

This setting is only useful when messages are written directly to the terminal or a file, because journalctl(1) and other tools that display logs will attach timestamps based on the entry metadata on their own.

$SYSTEMD_LOG_LOCATION

A boolean. If true, messages will be prefixed with a filename and line number in the source code where the message originates.

Note that the log location is often attached as metadata to journal entries anyway. Including it directly in the message text can nevertheless be convenient when debugging programs.

$SYSTEMD_LOG_TID

A boolean. If true, messages will be prefixed with the current numerical thread ID (TID).

Note that the this information is attached as metadata to journal entries anyway. Including it directly in the message text can nevertheless be convenient when debugging programs.

$SYSTEMD_LOG_TARGET

The destination for log messages. One of console (log to the attached tty), console-prefixed (log to the attached tty but with prefixes encoding the log level and "facility", see syslog(3), kmsg (log to the kernel circular log buffer), journal (log to the journal), journal-or-kmsg (log to the journal if available, and to kmsg otherwise), auto (determine the appropriate log target automatically, the default), null (disable log output).

$SYSTEMD_LOG_RATELIMIT_KMSG

Whether to ratelimit kmsg or not. Takes a boolean. Defaults to "true". If disabled, systemd will not ratelimit messages written to kmsg.

$SYSTEMD_PAGER, $PAGER

Pager to use when --no-pager is not given. $SYSTEMD_PAGER is used if set; otherwise $PAGER is used. If neither $SYSTEMD_PAGER nor $PAGER are set, a set of well-known pager implementations is tried in turn, including less(1) and more(1), until one is found. If no pager implementation is discovered, no pager is invoked. Setting those environment variables to an empty string or the value "cat" is equivalent to passing --no-pager.

Note: if $SYSTEMD_PAGERSECURE is not set, $SYSTEMD_PAGER and $PAGER can only be used to disable the pager (with "cat" or ""), and are otherwise ignored.

$SYSTEMD_LESS

Override the options passed to less (by default "FRSXMK").

Users might want to change two options in particular:

K

This option instructs the pager to exit immediately when Ctrl+C is pressed. To allow less to handle Ctrl+C itself to switch back to the pager command prompt, unset this option.

If the value of $SYSTEMD_LESS does not include "K", and the pager that is invoked is less, Ctrl+C will be ignored by the executable, and needs to be handled by the pager.

X

This option instructs the pager to not send termcap initialization and deinitialization strings to the terminal. It is set by default to allow command output to remain visible in the terminal even after the pager exits. Nevertheless, this prevents some pager functionality from working, in particular paged output cannot be scrolled with the mouse.

Note that setting the regular $LESS environment variable has no effect for less invocations by systemd tools.

See less(1) for more discussion.

$SYSTEMD_LESSCHARSET

Override the charset passed to less (by default "utf-8", if the invoking terminal is determined to be UTF-8 compatible).

Note that setting the regular $LESSCHARSET environment variable has no effect for less invocations by systemd tools.

$SYSTEMD_PAGERSECURE

Common pager commands like less(1), in addition to "paging", i.e. scrolling through the output, support opening of or writing to other files and running arbitrary shell commands. When commands are invoked with elevated privileges, for example under sudo(8) or pkexec(1), the pager becomes a security boundary. Care must be taken that only programs with strictly limited functionality are used as pagers, and unintended interactive features like opening or creation of new files or starting of subprocesses are not allowed. "Secure mode" for the pager may be enabled as described below, if the pager supports that (most pagers are not written in a way that takes this into consideration). It is recommended to either explicitly enable "secure mode" or to completely disable the pager using --no-pager or PAGER=cat when allowing untrusted users to execute commands with elevated privileges.

This option takes a boolean argument. When set to true, the "secure mode" of the pager is enabled. In "secure mode", LESSSECURE=1 will be set when invoking the pager, which instructs the pager to disable commands that open or create new files or start new subprocesses. Currently only less(1) is known to understand this variable and implement "secure mode".

When set to false, no limitation is placed on the pager. Setting SYSTEMD_PAGERSECURE=0 or not removing it from the inherited environment may allow the user to invoke arbitrary commands.

When $SYSTEMD_PAGERSECURE is not set, systemd tools attempt to automatically figure out if "secure mode" should be enabled and whether the pager supports it. "Secure mode" is enabled if the effective UID is not the same as the owner of the login session, see geteuid(2) and sd_pid_get_owner_uid(3), or when running under sudo(8) or similar tools ($SUDO_UID is set [3]). In those cases, SYSTEMD_PAGERSECURE=1 will be set and pagers which are not known to implement "secure mode" will not be used at all. Note that this autodetection only covers the most common mechanisms to elevate privileges and is intended as convenience. It is recommended to explicitly set $SYSTEMD_PAGERSECURE or disable the pager.

Note that if the $SYSTEMD_PAGER or $PAGER variables are to be honoured, other than to disable the pager, $SYSTEMD_PAGERSECURE must be set too.

$SYSTEMD_COLORS

Takes a boolean argument, or a special value. By default (unset), systemd and related utilities will use colors in their output if possible. If $COLORTERM is set to "truecolor" or "24bit", 24-bit colors will be enabled, 256 colors otherwise, unless $NO_COLOR or $TERM indicates colors are disabled.

true

Same as unset, except that $NO_COLOR is ignored.

false

The output will be monochrome.

"16", "256", "24bit"

Always use the base 16 ANSI colors, 256 colors, or 24 bit color, respectively.

"auto-16", "auto-256", "auto-24bit"

Use the given quantity of colours, subject to $TERM, and what the console is connected to.

$SYSTEMD_URLIFY

The value must be a boolean. Controls whether clickable links should be generated in the output for terminal emulators supporting this. This can be specified to override the decision that systemd makes based on $TERM and other conditions.

Example 1. Run an Arch Linux VM image generated by mkosi

$ mkosi -d arch -p systemd -p linux --autologin -o image.raw -f build
$ systemd-vmspawn --image=image.raw

Example 2. Import and run a Fedora 44 Cloud image using importctl

$ curl -L \
       -O https://download.fedoraproject.org/pub/fedora/linux/releases/44/Cloud/x86_64/images/Fedora-Cloud-Base-Generic-44-1.7.x86_64.qcow2 \
       -O https://download.fedoraproject.org/pub/fedora/linux/releases/44/Cloud/x86_64/images/Fedora-Cloud-44-1.7-x86_64-CHECKSUM \
       -O https://fedoraproject.org/fedora.gpg
$ gpgv --keyring ./fedora.gpg Fedora-Cloud-44-1.7-x86_64-CHECKSUM
$ sha256sum -c Fedora-Cloud-44-1.7-x86_64-CHECKSUM
# importctl import-raw -m Fedora-Cloud-Base-Generic-44-1.7.x86_64.qcow2 fedora-44-cloud
# systemd-vmspawn -M fedora-44-cloud

Example 3. Build and run systemd's system image and forward the VM's journal to a local file

$ mkosi build
$ systemd-vmspawn \
    -D mkosi.output/system \
    --private-users $(grep $(whoami) /etc/subuid | cut -d: -f2) \
    --linux mkosi.output/system.efi \
    --forward-journal=vm.journal \
    enforcing=0

Note: this example also uses a kernel command line argument to ensure SELinux is not started in enforcing mode.

Example 4. SSH into a running VM using systemd-ssh-proxy

$ mkosi build
$ my_vsock_cid=3735928559
$ systemd-vmspawn \
    -D mkosi.output/system \
    --private-users $(grep $(whoami) /etc/subuid | cut -d: -f2) \
    --linux mkosi.output/system.efi \
    --vsock-cid $my_vsock_cid \
    enforcing=0
$ ssh root@vsock/$my_vsock_cid -i /run/user/$UID/systemd/vmspawn/machine-*-system-ed25519

If an error occurred the value errno is propagated to the return code. If EXIT_STATUS is supplied by the running image that is returned. Otherwise, EXIT_SUCCESS is returned.

systemd(1), mkosi(1), machinectl(1), importctl(1), UAPI.1 Boot Loader Specification[1]

1.
UAPI.1 Boot Loader Specification
https://uapi-group.org/specifications/specs/boot_loader_specification
2.
ANSI Escape Code (Wikipedia)
https://en.wikipedia.org/wiki/ANSI_escape_code#SGR_(Select_Graphic_Rendition)_parameters
3.
It is recommended for other tools to set and check $SUDO_UID as appropriate, treating it is a common interface.
systemd 262