12. Plugins guide

12.1. Plugins Configuration

A plugin has its own configuration in config.ini file stored in the root directory of the plugin.

This config.ini file is created with its requisite configuration during the bootstrap_plugin.py create command.

You can change some of the configuration parameters and/or add configuration specific to your plugin.

Every time you change the configuration file, you have to stop and start the Metwork module to reload the configuration, by entering the commands :

  • either .. code-block:: bash

    {MFMODULE_RUNTIME_HOME}.stop {MFMODULE_RUNTIME_HOME}.start

  • or (as root user) .. code-block:: bash

    service metwork restart {MFMODULE_RUNTIME_HOME}

with {MFMODULE_RUNTIME_HOME} the Metwork module you are working with, e.g. mfdata.

The configuration parameters are described and explained in the ${MFMODULE_RUNTIME_HOME}/config/config.ini file.

12.2. Plugins Commands

This section sums up all the commands you will need to build, install a plugin.

Use -h or --help option of the command to review all available options of the command.

12.2.1. plugins.list

List installed plugins.

plugins.list [options]

Note: if you have a version named dev_link, the plugin is installed as a symbolic link to a source directory. It’s a kind of “development mode”.

12.2.2. plugins.install

Install a given plugin from a .plugin file.

plugins.install {plugin file path} [options]

12.2.3. plugins.uninstall

Uninstall a given plugin name (not file). The name corresponds to the first column of plugins.list output.

plugins.uninstall {plugin name} [options]

12.2.4. plugins.info

Get more infomation about a plugin.

plugins.info {plugin_name}

12.2.5. plugin_env

If you have a virtualenv in your plugin or if you want to simulate the exact environment your code will run in the context of your plugin, you can use the plugin_env interactive command.

If you are developping, the best way is to go in the root directory of your plugin and use plugin_env without argument.

If you have a “normal” installed plugin, you can use plugin_env {PLUGIN_NAME}.

When you are inside a plugin environment, you will find some extra environment variables:

env | grep "^${MFMODULE}_" | grep CURRENT
Output:

MFDATA_CURRENT_PLUGIN_NAME=my_plugin
MFDATA_CURRENT_PLUGIN_DIR=/home/mfdata/var/plugins/my_plugin
MFDATA_CURRENT_PLUGIN_LABEL=plugin_my_plugin@mfdata

And of course, paths of your plugin “layer” (kind of virtualenv generalization) are loaded before system paths.

12.2.6. plugin_wrapper

This command follows the same idea than plugin_env but in a non-interactive way. Use it for example for crontabs or if you want to execute a command in a given plugin environment without changing anything to the current shell.

plugin_wrapper {PLUGIN_NAME} {COMMAND} [{COMMAND_ARG1}] [{COMMAND_ARG2}] [...]

Note: the plugin {PLUGIN_NAME} must be installed.

12.3. Make Commands

12.3.1. make develop

Install your plugin in “development mode” (you don’t need to release and install each time you make a modification), the plugin installed through a symlink.

make develop

12.3.2. make release

Release your plugin as a .plugin file.

make release

12.3.3. make clean

Clean (before release)

make clean

12.3.4. make superclean

Hard-clean, can be useful when you have error with your virtualenv

make superclean

12.3.5. make

Build your virtualenv from sources

make

12.4. Python virtualenv

When developing Python applications, it’s standard practice to have a requirements.txt file.

This file can be used in different ways, and typically takes one of these two forms:

  • Simple requirements: A list of top-level dependencies a plugin has, often without versions specified.

  • Exact requirements: A complete list of all dependencies a plugin has, each with exact versions specified.

12.4.1. The ‘Simple’ requirements

A list of top-level dependencies a plugin has, often without versions specified.

The requirements.txt looks like:

requests[security]
flask
gunicorn==19.4.5

When a requirements.txt file like above is used to deploy in production environment, unexpected consequences can occur. Effectively, because versions haven’t been pinned, running pip install will give you different results today than it will tomorrow.

This is rather a bad way. As different versions of sub-dependencies are released, the result of a fresh pip install -r requirements.txt will result in different packages being installed, and potentially, your application failing for unknown and hidden reasons.

12.4.2. The ‘Exact’ Requirements

A complete list of all dependencies a plugin has, each with exact package versions specified.

The requirements.txt looks like:

cffi==1.5.2
cryptography==1.2.2
enum34==1.1.2
Flask==0.10.1
gunicorn==19.4.5
idna==2.0
ipaddress==1.0.16
itsdangerous==0.24
Jinja2==2.8
MarkupSafe==0.23
ndg-httpsclient==0.4.0
pyasn1==0.1.9
pycparser==2.14
pyOpenSSL==0.15.1
requests==2.9.1
six==1.10.0
Werkzeug==0.11.4

This is a best-practice for deploying applications, and ensures an explicit runtime environment with deterministic builds.

All dependencies, including sub-dependencies, are listed, each with an exact version specified.

While the this method for requirements.txt is best practice, it is a bit cumbersome. Namely, if you want to upgrade some or all of the packages, it’s not so easy to do.

12.4.3. The ‘Best’ Requirements

The best an simple way is to have two requirements file instead of having one:

  • requirements-to-freeze.txt

  • requirements.txt

The requirements-to-freeze.txt uses ‘simple’ requirements’, and is used to specify the top-level dependencies, and any explicit versions you need to specify.

the requirements.txt uses ‘exact’ requirements, and contains the output of a pip freeze after pip install requirements-to-freeze.txt has been run.

Working with Metwork plugin, you just need to fill the requirements-to-freeze.txt file. Then, when building/releasing the plugin, the requirements.txt will be generated.

Depending on which python version (python2 or python3) you are working with, the requirements.txt file will be named, respectively requirements2.txt and requirements3.txt

12.4.4. Python3 requirements-to-freeze.txt

If you are working with python3, an empty requirements-to-freeze.txt is created in the python3_virtualenv_sources plugin directory (during the bootstrap_plugin.py create command)

Add your python3 dependencies inside this file (with or without package version)

12.4.5. Python2 requirements-to-freeze.txt

If you are working with python2, an empty requirements-to-freeze.txt is created in the python2_virtualenv_sources plugin directory (during the bootstrap_plugin.py create command)

Add your python2 dependencies inside this file (with or without package version)