Prediction of Process Performance

Requirements:

1. This project requires both linux and MacOS environments to collect data respectively.
2. Python version >= 3.0; use pip3; kernel version >= 5.6
3. Linux enrionment dependencies:

   $ sudo apt-get install build-essential

4. Python dependencies:

   $ pip3 install -r REQUIREMENTS.txt

How to install proclog kernel module to collect linux process data:

1. cd into kernel_modules folder
2. Compile the proclog kernel module by following command.

   $ make

3. Install the kernel module by following command. Then a file /proc/log_file can be written to by kernel module.

   $ make exec

4. You can verify this by running the following code, which lists all of the modules currently in the kernel. Among them, you should see proclog. Note that the kernel replaces dashes in your module’s filename with underscores when it loads it. If you want to remove it, you can run the following command.

   $ lsmod

   $ make stop

5. You can verify the virtual file by running the following code, which lists all of process tasks in virtual file.

   $ cat /proc/log_file

6. In the source code, we can check some logging to let it be known our module loaded okay by running the following command. If the operation gets denied, set the restriction to 0.

   $ dmesg 

   $ sudo sysctl kernel.dmesg_restrict=0

How to export linux process data from /proc/log_file into actual file:

1. cd into data_collecting folder
2. Compile the index.c by following command.

   $ make

3. Execute the following command to collect data.

   $ ./index 

4. A csv file will be generated.

How to use python functions to collect mac processes data:

1. cd into data_collecting folder
2. Run the following command.

   $ sudo python3 index.py

3. A csv file will be generated.

How to run python functions to predict processes performance:

1. cd into ml_py folder
2. Run the following command.

   $ python3 index.py

3. The results will be shown in terminal.