Introduction to Parallel Computing (CMSC416/CMSC818X)

Extra Credit Assignment 7: Study of Performance Analysis Workflows

Due: Monday December 13, 2021 @ 11:59 PM Eastern Time

In this assignment you will start with the performance data you collected as part of Assignment 2, and analyze it using Hatchet after making two small changes:

Install Hatchet manually:

You can install Hatchet on deepthought2 by following these steps:


        module load python/3.7.7 (only works on deepthought2)
        pip install textx --user
        git clone https://github.com/hpcgroup/hatchet
        
Then add the directory where you installed hatchet to your PYTHONPATH and then do:

        export PYTHONPATH=$PYTHONPATH:<path>
        cd hatchet/
        ./install.sh 
        

Enable logging in each of your Python scripts:

You can enable logging in your Assignment 2 Python notebooks or scripts using:


        gf = ht.GraphFrame.from_hpctoolkit(dirname, logging=True)
        
This will generate a log file for each Python file that you used for the tasks in Assignment 2.

There are two ways to get extra credit for this assignment:

  1. If you consent to sharing the logs with us for research, use this Google form to record your consent and upload the logs.
  2. If you do not consent to sharing the logs with us for research, use this Google form to answer some simple questions about the analysis.