OpInt Framework.
There are 3 options for the Framework installation:
- Minimal, required only for running Jupyter notebooks supplied with example applications.
- Backend service intended for running agents.
- REST service for Web deployment
Any installation scenario starts from getting the source code:
git clone https://github.com/operationalintelligence/opint-framework
Assuming that Jupyter notebook is pre installed on the target system, notebooks could be ready to run. If a notebook needs extra dependency they could be installed in a standard manned depending on what package manager you are using (conda or pip).
Create a new python3 virtual environment and activate it:
python3 -m venv ./virtenv
source ./virtenv/bin/activate
Install the global (minimal) framework requirements:
pip install -r ./opint-framework/requirements.txt
(optional) Depending on the usage scenario ones may want to install requirements for specific deployment and application intended to use, e.g:
pip install -r ./opint-framework/opint_framework/apps/workload_jobsbuster/requirements.txt
The following two lines are needed to supply paths and settings to Django:
export PYTHONPATH=$(pwd)/opint-framework
export DJANGO_SETTINGS_MODULE=opint_framework.conf.settings
At this point the agents could be executed:
python3 opint-framework/opint_framework/core/scheduler/mainlooper.py
If installation passed successfully, the following message will be displayed:
Hello World from Sample Agent
and then FrameWork exits. To activate the infinite loop mode, when agents periodically called, the following parameter in the opint-framework/opint_framework/conf/settings.py
DO_DEBUG_AGENTS = True
If the Operational Intelligence framework will be used for providing REST services, the Django dependencies should be also installed:
cd ~/projects/opint-framework/opint-framework
pip install -r requirements.txt
The following environmental variables can be set: Run the django server:
python3 opint-framework/manage.py runserver