3diy
/
moonraker
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
moonraker/docs/installation.md

18 KiB
Raw Blame History

Installation

This document provides a guide on how to install Moonraker on a Raspberry Pi running Raspian/Rasperry Pi OS. Other SBCs and/or linux distributions may work, however they may need a custom install script. Moonraker requires Python 3.7 or greater, verify that your distribution's Python 3 packages meet this requirement.

Installing Klipper

Klipper should be installed prior to installing Moonraker. Please see Klipper's Documention for details. After installing Klipper you should make sure to add Moonraker's configuration requirements.

Klipper Configuration Requirements

Moonraker depends on the following Klippy extras for full functionality:

  • [virtual_sdcard]
  • [pause_resume]
  • [display_status]

If you have a [filament_switch_sensor] configured then [pause_resume] will automatically be loaded. Likewise, if you have a [display] configured then [display_status] will be automatically loaded. If your configuration is missing one or both, you can simply add the bare sections to printer.cfg:

[pause_resume]

[display_status]

[virtual_sdcard]
path: ~/moonraker_data/gcodes

Enabling the Unix Socket

After Klipper is installed it may be necessary to modify its defaults file in order to enable the Unix Domain Socket. Begin by opening the file in your editor of choice, for example:

sudo nano /etc/default/klipper

You should see a file that looks something like the following:

# Configuration for /etc/init.d/klipper

KLIPPY_USER=pi

KLIPPY_EXEC=/home/pi/klippy-env/bin/python

KLIPPY_ARGS="/home/pi/klipper/klippy/klippy.py /home/pi/printer.cfg -l /tmp/klippy.log"

Add -a /tmp/klippy_uds to KLIPPY_ARGS:

# Configuration for /etc/init.d/klipper

KLIPPY_USER=pi

KLIPPY_EXEC=/home/pi/klippy-env/bin/python

KLIPPY_ARGS="/home/pi/klipper/klippy/klippy.py /home/pi/printer.cfg -l /tmp/klippy.log -a /tmp/klippy_uds"

!!! note Your installation of Klipper may use systemd instead of the default LSB script. In this case, you need to modify the klipper.service file.

You may also want to take this opportunity to configure printer.cfg and klippy.log so they are located in Moonraker's data_path, for example:

# Configuration for /etc/init.d/klipper

KLIPPY_USER=pi

KLIPPY_EXEC=/home/pi/klippy-env/bin/python

KLIPPY_ARGS="/home/pi/klipper/klippy/klippy.py /home/pi/moonraker_data/config/printer.cfg -l /home/pi/moonraker_data/logs/klippy.log -a /tmp/klippy_uds"

Moonraker's install script will create the data folder, however you may wish to create it now and move printer.cfg to the correct location, ie:

mkdir ~/moonraker_data
mkdir ~/moonraker_data/logs
mkdir ~/moonraker_data/config
mv printer.cfg ~/moonraker_data/config

Installing Moonraker

Begin by cloning the git respository:

cd ~
git clone https://github.com/Arksine/moonraker.git

The install script will attempt to create a basic configuration if moonraker.conf does not exist at the expected location, however if you prefer to have Moonraker start with a robust configuration you may create it now. By default the configuration file should be located at $HOME/moonraker_data/config/moonraker.conf, however the location of the data path may be configured using the script's command line options. The sample moonraker.conf may be used as a starting point, full details can be found in the confguration documentation.

For a default installation run the following commands:

cd ~/moonraker/scripts
./install-moonraker.sh

The install script has a few command line options that may be useful, particularly for those upgrading:

  • -f: Force an overwrite of Moonraker's systemd script. By default the the systemd script will not be modified if it exists.
  • -a <alias>: The alias used for this instance of Moonraker. Moonraker uses this option when determining the file names for the configuration file, log file, secrets file, etc. This option is also used to name Moonraker's systemd service unit and enviroment file. If omitted this defaults to moonraker.
  • -d <path to data folder>: Specifies the path to Moonraker's data folder. This folder organizes files and directories used by moonraker. See the Data Folder Structure section for details. If omitted this defaults to $HOME/<alias>_data, ie: /home/pi/moonraker_data.
  • -c <path to configuration file> Specifies the path to Moonraker's configuation file. By default the configuration is expected at <data_folder>/config/<alias>.conf. ie: /home/pi/moonraker_data/config/moonraker.conf.
  • -l <path to log file> Specifies the path to Moonraker's log file. By default Moonraker logs to <data_folder>/logs/<alias>.log. ie: /home/pi/moonraker_data/logs/moonraker.log.
  • -z: Disables systemctl commands during install (ie: daemon-reload, restart). This is useful for installations that occur outside of a standard environment where systemd is not running.
  • -x: Skips installation of polkit rules. This may be necessary to install Moonraker on systems that do not have policykit installed.

Additionally, installation may be customized with the following environment variables:

  • MOONRAKER_VENV
  • MOONRAKER_REBUILD_ENV
  • MOONRAKER_FORCE_DEFAULTS
  • MOONRAKER_DISABLE_SYSTEMCTL
  • MOONRAKER_SKIP_POLKIT
  • MOONRAKER_CONFIG_PATH
  • MOONAKER_LOG_PATH
  • MOONRAKER_DATA_PATH

When the script completes it should start both Moonraker and Klipper. In klippy.log you should find the following entry:

webhooks client <uid>: Client info {'program': 'Moonraker', 'version': '<version>'}

Now you may install a client, such as Mainsail or Fluidd.

!!! Note Moonraker's install script no longer includes the nginx dependency. If you want to install one of the above clients on the local machine, you may want to first install nginx (sudo apt install nginx on debian/ubuntu distros).

Data Folder Structure

As mentioned previously, files and folders used by Moonraker are organized in a primary data folder. The example below illustrates the folder structure using the default data path of $HOME/moonraker_data and the default alias of moonraker:

/home/pi/moonraker_data
├── backup
│   └── 20220822T202419Z
│       ├── config
│       │   └── moonraker.conf
│       └── service
│           └── moonraker.service
├── certs
│   ├── moonraker.cert (optional)
│   └── moonraker.key (optional)
├── config
│   ├── moonraker.conf
│   └── printer.cfg
├── database
│   ├── data.mdb
│   └── lock.mdb
├── gcodes
│   ├── test_gcode_one.gcode
│   └── test_gcode_two.gcode
├── logs
│   ├── klippy.log
│   └── moonraker.log
└── moonraker.secrets (optional)

The next example illustrates how the <data file path> and <alias> command line options are used to populate the folder:

<data file path>
├── backup
│   └── <ISO8601_DATE>
│       ├── config
│       │   └── <full configuration backup>
│       └── service
│           └── <systemd service backup>
├── certs
│   ├── <alias>.cert (optional)
│   └── <alias>.key (optional)
├── config
│   ├── <alias>.conf
│   └── printer.cfg
├── database
│   ├── data.mdb
│   └── lock.mdb
├── gcodes
│   ├── test_gcode_one.gcode
│   └── test_gcode_two.gcode
├── logs
│   ├── klippy.log
│   └── <alias>.log
└── <alias>.secrets (optional)

If it is not desirible for the files and folders to exist in these specific locations it is acceptable to use symbolic links. For example, it is common for the gcode folder to be located at $HOME/gcode_files. Rather than reconfigure Klipper's virtual_sdcard it may be desirable to create a gcodes symbolic link in the data path pointing to this location.

!!! Note It is still possible to directly configure the paths to the configuration and log files if you do not wish to use the default <alias> naming scheme or if you wish for them to exist outside of the data folder.

When Moonraker attempts to update legacy installations symbolic links are used to avoid an unrecoverable error. Additionally a backup folder is created which contains the prior configuration and/or systemd service unit, ie:

/home/pi/moonraker_data
├── backup
│   └── 20220822T202419Z
│       ├── config
│       │   ├── include
│       │   │   ├── extras.conf
│       │   │   ├── power.conf
│       │   │   └── updates.conf
│       │   └── moonraker.conf
│       └── service
│           └── moonraker.service
├── certs
│   ├── moonraker.cert -> /home/pi/certs/certificate.pem
│   └── moonraker.key -> /home/pi/certs/key.pem
├── config -> /home/pi/klipper_config
├── database -> /home/pi/.moonraker_database
├── gcodes -> /home/pi/gcode_files
├── logs -> /home/pi/logs
└── moonraker.secrets -> /home/pi/moonraker_secrets.ini

!!! Warning The gcode and config paths should not contain symbolic links that result in an "overlap" of on another. Moonraker uses inotify to watch files in each of these folders and takes action when a file change is detected. The action taken depends on the "root" folder, thus it is important that they be distinct.

The systemd service file

The default installation will create /etc/systemd/system/moonraker.service. Below is a common example of service file, installed on a Raspberry Pi:

# systemd service file for moonraker
[Unit]
Description=API Server for Klipper SV1
Requires=network-online.target
After=network-online.target

[Install]
WantedBy=multi-user.target

[Service]
Type=simple
User=pi
SupplementaryGroups=moonraker-admin
RemainAfterExit=yes
WorkingDirectory=/home/pi/moonraker
EnvironmentFile=/home/pi/moonraker/moonraker.env
ExecStart=/home/pi/moonraker-env/bin/python $MOONRAKER_ARGS
Restart=always
RestartSec=10

Following are some items to take note of:

  • The Description contains a string that Moonraker uses to validate the version of the service file, (notice SV1 at the end, ie: Service Version 1).
  • The moonraker-admin supplementary group is used to grant policykit permissions.
  • The EnvironmentFile field contains Moonraker's arguments. More on this below.
  • The ExecStart field begins with the python executable, followed by by the enviroment variable MOONRAKER_ARGS. This variable is set in the environment file.

The Enivirorment File

The environment file is created in Moonraker's source directory during installation. By default the enviroment file is named moonraker.env. A default installation's enviroment file will contain the path to moonraker.py and the alias option, ie:

MOONRAKER_ARGS="/home/pi/moonraker/moonraker/moonraker.py -a moonraker"

A legacy installation converted to the updated flexible service unit might contain the following:

MOONRAKER_ARGS="/home/pi/moonraker/moonraker/moonraker.py -a moonraker -d /home/pi/moonraker_data -c /home/pi/klipper_config/moonraker.conf -l /home/pi/klipper_logs/moonraker.log"

Post installation it is simple to customize the arguments supplied to Moonraker by editing this file and restarting the service.

!!! Note The service unit and enviroment file are named based on the alias option supplied to the install script, which is moonraker by default. Supplying -a moonraker_1 to the install script will result in a service file named moonraker_1.service and an environment file named moonraker_1.env.

Command line usage

This section is intended for users that need to write their own installation script. Detailed are the command line arguments available to Moonraker:

usage: moonraker.py [-h] [-a <alias>] [-d <data path>] [-c <configfile>] [-l <logfile>] [-n]

Moonraker - Klipper API Server

options:
  -h, --help            show this help message and exit
  -a <alias>, --alias <alias>
                        Alternate name of instance
  -d <data path>, --datapath <data path>
                        Location of Moonraker Data File Path
  -c <configfile>, --configfile <configfile>
                        Location of moonraker configuration file
  -l <logfile>, --logfile <logfile>
                        log file name and location
  -n, --nologfile       disable logging to a file

The default configuration is:

  • data path: $HOME/moonraker_data
  • alias: moonraker
  • config file: $HOME/moonraker_data/config/moonraker.conf
  • log file: $HOME/moonraker_data/logs/moonraker.log
  • logging to a file is enabled

!!! Tip While the alias option may be omitted it is recommended that it always be included for new installations. This allows Moonraker to differentiate between new and legacy installations.

If is necessary to run Moonraker without logging to a file the -n option may be used, for example:

~/moonraker-env/bin/python ~/moonraker/moonraker/moonraker.py -n

In general it is not recommended to install Moonraker with file logging disabled. While moonraker will still log to stdout, all requests for support must be accompanied by moonraker.log.

These options may be changed by editing moonraker.env.

PolicyKit Permissions

Some of Moonraker's components require elevated privileges to perform actions. Previously these actions could only be run via commandline programs launched with the sudo prefix. This has significant downsides:

  • The user must be granted NOPASSWD sudo access. Raspberry Pi OS grants the Pi user this access by default, however most other distros require that this be enabled through editing visudo or adding files in /etc/sudoers.d/.
  • Some linux distributions require additional steps such as those taken in sudo_fix.sh.
  • Running CLI programs is relatively expensive. This isn't an issue for programs that are run once at startup, but is undesirable if Moonraker wants to poll information about the system.

Moonraker now supports communicating with system services via D-Bus. Operations that require elevated privileges are authrorized through PolicyKit. On startup Moonraker will check for the necessary privileges and warn users if they are not available. Warnings are presented in moonraker.log and directly to the user through some clients.

To resolve these warnings users have two options:

  1. Install the PolicyKit permissions with the set-policykit-rules.sh script, for example:
cd ~/moonraker/scripts
./set-policykit-rules.sh
sudo service moonraker restart

!!! tip If you still get warnings after installing the PolKit rules, run the install script with no options to make sure that all new dependencies are installed.

```shell
cd ~/moonraker/scripts
./install-moonraker.sh
```
  1. Configure Moonraker to use the legacy backend implementations for the machine and/or update_manager components, ie:
# Use the systemd CLI provider rather than the DBus Provider
[machine]
provider: systemd_cli

# Edit your existing [update_manager] section to disable
# PackageKit.  This will fallback to the APT CLI Package Update
# implementation.
[update_manager]
#..other update manager options
enable_packagekit: False

# Alternatively system updates can be disabled
[update_manager]
#..other update manager options
enable_system_updates: False

!!! Note Previously installed PolicyKit rules can be removed by running set-policykit-rules.sh -c

Retrieving the API Key

Some clients may require an API Key to connect to Moonraker. After the [authorization] component is first configured Moonraker will automatically generate an API Key. There are two ways in which the key may be retrieved by the user:

Retrieve the API Key via the command line (SSH):

cd ~/moonraker/scripts
./fetch-apikey.sh

Retrieve the API Key via the browser from a trusted client:

  • Navigate to http://{moonraker-host}/access/api_key, where {moonraker-host} is the host name or ip address of the desired moonraker instance.

  • The result will appear in the browser window in JSON format. Copy The API Key without the quotes.

      {"result": "8ce6ae5d354a4365812b83140ed62e4b"}

Recovering a broken repo

Currently Moonraker is deployed using git. Without going into the gritty details,git is effectively a file system, and as such is subject to file system corruption in the event of a loss of power, bad sdcard, etc. If this occurs, updates using the [update_manager] may fail. In most cases Moonraker provides an automated method to recover, however in some edge cases this is not possible and the user will need to do so manually. This requires that you ssh into your machine. The example below assumes the following:

  • You are using a Raspberry Pi
  • Moonraker and Klipper are installed at the default locations in the home directory
  • Both Moonraker and Klipper have been corrupted and need to be restored

The following commands may be used to restore Moonraker:

cd ~
rm -rf moonraker
git clone https://github.com/Arksine/moonraker.git
cd moonraker/scripts
./install-moonraker.sh
./set-policykit-rules.sh
sudo systemctl restart moonraker

And for Klipper:

cd ~
rm -rf klipper
git clone https://github.com/Klipper3d/klipper.git
sudo systemctl restart klipper