September 22, 2020 Software Open Access

Mukherjee, Suvam; Deligiannis, Pantazis; Biswas, Arpita; Lal, Akash

Citation Style Language JSON Export

{
  "publisher": "Zenodo", 
  "DOI": "10.5281/zenodo.4043041", 
  "language": "eng", 
  "title": "Learning-based Controlled Concurrency Testing", 
  "issued": {
    "date-parts": [
      [
        2020, 
        9, 
        22
      ]
    ]
  }, 
  "abstract": "<p>Concurrency bugs are notoriously hard to detect and reproduce. Controlled concurrency testing (CCT) techniques aim to offer a solution, where a&nbsp;<em>scheduler</em>&nbsp;explores the space of possible interleavings of a concurrent program looking for bugs. Since the set of possible interleavings is typically very large, these schedulers employ heuristics that prioritize the search to &quot;interesting&quot;&nbsp;subspaces. However, current heuristics are typically tuned to specific bug patterns, which limits their effectiveness in practice.</p>\n\n<p>In this artifact, we present QL, a learning-based CCT framework where the likelihood of an action being selected by the scheduler is influenced by earlier explorations. We leverage the classical Q-learning algorithm to explore the space of possible interleavings, allowing the exploration to adapt to the program under test, unlike previous techniques. We have implemented and evaluated QL&nbsp;on a set of microbenchmarks, complex protocols, as well as production cloud services. In our experiments, we found QL&nbsp;to consistently outperform the state-of-the-art in CCT.</p>\n\n<p>Please refer to the README file for more details on how to run the artifact.</p>", 
  "author": [
    {
      "family": "Mukherjee, Suvam"
    }, 
    {
      "family": "Deligiannis, Pantazis"
    }, 
    {
      "family": "Biswas, Arpita"
    }, 
    {
      "family": "Lal, Akash"
    }
  ], 
  "type": "article", 
  "id": "4043041"
}