yacfg 0.8.3

Template based configuration generator

yacfg - YAML Configurator

This tool can generate a set of configuration files mainly needed for AMQ Broker, but it is not limited to only generating files for one product.

It has a user facing Command Line Tool for quick and easy command line usage. Furthermore, it is possible to use its API in your python code.

Getting started

• python 3.5+ or python2.7
• current requirements from setup.py (runtime requirements only)
• python virtualenv recommended (install via system package manager or pip install --user virtualenv)

for contributors:

• requirements from requirements.txt (there are Dev and QA requirements as well)

From git

git clone [e-mail]:rh-messaging-qe/yacfg.git
python -m virtualenv -p python3 venv3
source venv3/bin/activate
./setup.py install
yacfg --help

From PyPI

python -m virtualenv -p python3 venv3
source venv3/bin/activate
pip install yacfg
yacfg --help

User (CLI) guide

yacfg --help

yacfg --list-profiles
yacfg --list-templates

# perform a generation of a default profile
yacfg --profile artemis/2.5.0/default.yaml.jinja2
# also save result to [OUTDIR] directory
yacfg --profile [PROFILE] --output [OUTDIR]

Customization

Quickest way to customize data is to use hot-variables, basically variables that the profile itself provides for tuning. Next step is to write (modify) custom profile with completely custom values. If that does not satisfy your needs, then a custom template might be required.

Profile tuning

Simply export tuning values from profile you want to tune and change those you need to change. Then supply the custom tuning file(s) when generating the profile.

yacfg --profile [PROFILE] --export-tuning my_values.yaml
vim my_values.yaml
yacfg --profile [PROFILE] --tune my_values.yaml

# multiple tuning files can be overlaid
# they are updated in sequence, only values present are overwritten
yacfg --profile [PROFILE] --tune my_values.yaml --tune machine_specific.yaml \
       --tune logging_debug.yaml --output [OUTDIR]

Custom profiles

Write your own, or simply export an existing profile and modify that.

You can export dynamic version with includes of some modules, that would still work. Either you can use imports from package, or your own local files.

Or you can export completely rendered profile file without any includes or variables and modify that as you like.

# export profile with dynamic includes still active jinja2 files
yacfg --profile [PROFILE] --new-profile my_new_profile.yaml.jinja2
# export completely generated profile without any jinja2 fields, plain yaml
yacfg --profile [PROFILE] --new-profile-static my_new_profile.yaml
vim my_new_profile.yaml
yacfg --profile my_new_profile.yaml

Profile is just another jinja2 file that enables customization of profile data -- that is tuning. Becuase of that we recommend to keep the extension .yaml.jinja2 unless it is static profile withou any jinja2 capabilities, in that case it could be named .yaml. That way we can run yaml lint against static profiles and verify that they are correct.

All profiles have to be used to generate files without any tuning. That means, if they are tune-able, they have to contain all default values in _defaults section. That section is also used for tuning, so any variable in there will be exported as tuning.

Custom templates

The last resort is to export a template and modify that. But remember a template, or more correctly a template set is a directory containing a set of main templates that subsequently generate_via_tuning_files a new file.

Of course feel free to write your own templates. Especially when you need to generate_via_tuning_files files for something that is not packaged.

Just remember for a template set to be identified the directory must contain a file named '_template' and then main templates ending with '.jinja2'.

yacfg --template [TEMPLATE] --new-template my_new_template
vim my_new_template/[MAIN_TEMPLATES].jinja2
yacfg --template my_new_template --profile [PROFILE]

API guide

Direct use of API is to use generate() nearly the same as the CLI. With option to use tuning values directly.

Tuning data will be overlaid in order of appearance, using python dict.update(), so values that will appear later will overwrite previous values. We recommend that tuning values are always flat, because update is not recursive. The same applies for data from tuning files as well as the directly provided data.

Data application order:

• profile defaults
• data from tuning files (in order of appearance) tuning_files_list
• data provided directly (in order of appearance) tuning_data_list

import yacfg

# generating only broker.xml config using default values from profile,
# no tuning, writing output to a target path
yacfg.generate(
    profile='artemis/2.5.0/default.yaml.jinja2',
    output_filter=['broker.xml'],
    output_path='/opt/artemis-2.5.0-i0/etc/',
)

# using both files and direct values, and writing generated configs to
# a target directory
yacfg.generate(
    profile='artemis/2.5.0/default.yaml.jinja2',
    tuning_files_list=[
        'my_values.yaml',
        'machine_specific.yaml',
        'logging_debug.yaml'
    ],
    tuning_data_list=[
        {'name': 'custom name', 'config': 'option_a'},
        {'address': '10.0.0.1'},
        {'LOG_LEVEL': 'debug'},
    ],
    output_path='/opt/artemis-2.5.0-i0/etc/',
)

# just get generated data for further processing, using just tuning files
data = yacfg.generate(
    profile='artemis/2.5.0/default.yaml.jinja2',
    tuning_files_list=[
        'my_values.yaml',
        'machine_specific.yaml',
        'logging_debug.yaml'
    ],
)
print(data['broker.xml'])

Batch configurations

In case you have multiple services to configure in your environment, that you probably will have at some point, there is a tool for that as well. The tool is called yacfg-batch. It has only yaml input and it uses yacfg to generate configurations as you are already used to.

Input yaml file defines all services you need to generate, what profiles to use, and what tuning to provide to yacfg. It allows you to configure defaults and common for services.

Batch profile file

As said it is YAML. It has two special sections: _default and _common. As the name suggests, _default values are used when values are not defined per specific section. Where _common is added to the values of all sections. The important thing here is that _default has lower priority than _common and that has lower priority than specific section values.

Every section has 4 values: profile, template, tuning_files, and tuning. As the name suggests, profile defines what generation profile to select, and it directly correlates with yacfg's --profile. template defines what generation template to use (overrides one in the profile if defined), and it directly correlates with --template from yacfg. tuning_files option is a list of tuning files to use, when combining defaults, commons, and specific values, tuning_files list is concatenated. Finally tuning is a map of specific tuning values, correlates with --opt of yacfg. When combining defaults, commons, and specifics, it will be updated over using python dict.update() and it will work only on first level, so it is recommended to use flat values for tuning only.

Example:

_default:
    profile: artemis/2.5.0/default.yaml.jinja2
    tuning_files:
      - defaults/broker_default.yaml

_common:
    tuning_files:
      - common/security.yaml
      - common/logging.yaml
    tuning_values:
      LOG_LEVEL_ALL: INFO

brokerA/opt/artemis/etc:
    pass: true

brokerB/opt/artemis/etc:
    profile: artemis/2.5.0/AIOBasic.yaml.jinja2
    tuning_files:
      - brokerB/queues.yaml

---

_default:
    profile: amq_broker/7.2.0/default.yaml.jinja2
    tuning_files:
      - defaults/amq_broker_default.yaml

brokerC/opt/amq/etc:
    tuning:
      LOG_LEVEL_ALL: DEBUG

As you can see, yacfg-batch supports multiple sections, in single batch profile file, that allows you to generate multiple groups using separated _default and _common sections for that.

executing batch

When you have defined all tuning files you need, and in the root of this batch configuration you have your batch profile file, you can now simply run yacfg-batch:

yacfg-batch --input [batch_profile_file] --output [output_path]

You can use multiple input files and all of those will be generated consecutively. In the output path, new subdirectories will be created for every item you configure (every section), section key will be used for that subdirectory. If the section name resembles a path, whole path will be created. For example for brokerA/opt/artemis/etc the configuration will be generated into [output_path]/brokerA/opt/artemis/etc/.

Documentation

Formatted documentation can be viewed at rh-messaging-qe.github.io/yacfg/.

Contributing

If you find a bug or room for improvement, submit either a ticket or PR.

Contributors

Alphabetically ordered

• Michal Tóth [e-mail]
• Otavio Piske [e-mail]
• Sean Davey [e-mail]
• Zdenek Kraus [e-mail] (maintainer)

License

Copyright 2018-2020 Red Hat Inc.

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

Acknowledgments

• jinja2 -- awesome templating engine
• yaml -- very convenient user readable format
• learn_yaml -- great YAML cheat sheet
• pyyaml -- python YAML parser
• jq -- great tool for working with structured data (JSON)
• yq -- YAML variant of jq
• github templates examples -- Nice set of ISSUE_TEMPLATE.md and PULL_REQUESTS_TEMPLATE.md examples
• contributing example -- example/template of CONTRIBUTING.md
• Fedora Project code-of-conduct -- the inspiration for CODE_OF_CONDUCT.md