Artifact for Survey on High-Performance Serverless Computing

This repository contains the data and scripts used to support the systematic literature review on high-performance serverless computing presented in High-Performance Serverless Computing: A Systematic Literature Review on Serverless for HPC, AI, and Big Data. It is designed to ensure full transparency and reproducibility of the methodology, taxonomy, and figures presented in the paper.

Overview

The survey systematically analyzes research on high-performance serverless computing, focusing on architectural trends, workloads, optimization techniques, and benchmarking practices. This artifact includes bibliographic data, classification metadata, and scripts that can be used to regenerate all plots.

Structure

• bib/

  • search_process/ Contains BibTeX files used during the literature review process. These were gathered from three major academic databases:

    1. ACM Digital Library
    2. Elsevier ScienceDirect
    3. IEEE Xplore

    Files include:

    • search_results.bib: Raw bibliographic results retrieved using structured queries.
    • starting_set.bib: Filtered initial seed papers based on relevance.
    • final_set.bib: Final set of papers after applying inclusion/exclusion criteria and performing forward/backward snowballing.

  • final_set.ris: Final set of the 122 papers in RIS format.

  • vosviewergraph.json: Graph extracted from VOSviewer.

• fonts/
  Font files used in the plots presented in the paper.

• plots/
  Contains Python scripts used to generate all figures included in the survey, along with:

  • requirements.txt: Lists all required Python packages.
  • coauth_citation_analysis.py: Python script that provide the relevant data used in the co-author and citation analysis.

• table/
  A structured datasheet that is maintained throughout the review process. This includes metadata for each paper (e.g., title, year, venue), classification labels, and filtering decisions used to support the taxonomy and analysis.

• make_plots.sh
  A shell script that automates virtual environment setup, dependency installation, and plot generation.

Usage

To regenerate all figures from the paper:

1. Ensure you have bash and Python ≥ 3.8 installed on your system.
2. Run the script from the repository root:

  ./make_plots.sh