The following web application is designed to store and monitor processes usage on Linux systems.
The Process Usage API collects the following pieces of metadata :
id: the process IDuser: the user running the processcpu: CPU load used by the processmemory: RAM memory used by the processcommand: name of the command used to run the processtime: time elapsed since the beginning of the process
To make it easier for you to collect those pieces of data, you can use the top Linux command and filter out the metadata described above.
In order to configure the Process Usage API, you shoud first install a Python version superior or equal to 3.11.
To install the project dependencies, you can refer to the requirements.txtand run :
pip install requirements.txt
Before deploying the Process Usage API, you should create a table in your Postgres database. To do so, you can use the src/init.sql file.
If you use Docker, you can add the following line in your docker-compose.yml file :
services:
database:
image: postgres:16.4
volumes:
- "./init.sql:/docker-entrypoint-initdb.d/init.sql"
The Process Usage API was built using FastAPI. Please refer to its official documentation (https://fastapi.tiangolo.com/) to find additional informations on how to deploy such applications.
In order to serve the API, you can use a Python webserver such as Uvicorn. You'll find its official documentation here : https://www.uvicorn.org/.
If you want to serve the API with Uvicorn, you can run the following command :
uvicorn main:app --host 0.0.0.0 --port <port_number>
The Process Usage API requires an access to Postgres database (version 16.4) to run properly.
In order to allow the Process Usage API to connect to a Postgres database, you should provide the following environment variables :
POSTGRES_USER: the database userPOSTGRES_PASSWORD: the database user's passwordPOSTGRES_HOST: the host name of the database machinePOSTGRES_DB: the name of the database