Skip to content

andreendo/noderacer

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

35 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

NodeRacer

A dynamic race detector with guided execution for Node.js applications.

This page contains the source code of NodeRacer and instructions to replicate the experiments, reported in the paper "NodeRacer: Event Race Detection for Node.js Applications" accepted in IEEE International Conference on Software Testing, Verification and Validation (ICST 2020).

Table of Contents

Install

  • System configuration:

    • Dell Latitude 5590 Intel I7 1.9GHz 16GB RAM
    • Ubuntu 18.04 LTS (kernel 4.15.0-64-generic)
  • Third-party software:

    • Install Node.js and npm (version used LTS 10.16.3)
    • Install git, Redis and MongoDb (used by some benchmarks), R and cloc (collect and process stats of the paper)
      sudo apt install git redis-server mongodb r-base cloc
  • Instructions to install NodeRacer:

    • In the terminal, run the following commands to clone the repository, install dependencies, and make command noderacer available.
      git clone https://github.com/noderacer/noderacer.git
      cd noderacer
      npm install
      npm link
    • To run the micro-benchmark tests (it may take a couple of minutes). See the micro-benchmark in here.
      npm test

Usage and tool's modules

If correctly installed, you can run NodeRacer in the terminal.

noderacer --help

This will show the options available to configure NodeRacer. The entry point for the CLI is coded in lib/cli/cli.js.

  • Observation phase
    • This phase can be run with noderacer log <command>, where <command> starts the sample run used; further configurations can be added using -c <config_file>, see here for a config file example.
    • Example:
      noderacer log node tests/micro-benchmark/paper-examples/archive-like-test.js  
    • This command creates a log file in folder log/<ID_FOR_COMMAND>/noderacer?.log.json.
    • See this module code here.
  • Happens-before identification phase
    • This phase can be run with noderacer hb <logfile>, where <logfile> is the path to the log file generated in the previous phase.
    • Example:
      noderacer hb log/<ID_FOR_COMMAND>/noderacer1.log.json -i
    • This command creates files and images related to the identified happens-before relations in folder log/<ID_FOR_COMMAND>.
    • See this module code here; in particular, the implementation of the nine rules used to identify the happens-before relations can be seen here.
  • Guided execution phase
    • This phase can be run with noderacer control -s <strategy> -r <runs> -h <hbfile> <command>, where <strategy> defines how guided execution is performed (parameter random is the strategy described in the paper), <runs> specifies how many repetitions (runs), hbfile is the path to file noderacer?.hb.json (generated in the previous phase), and <command> starts the sample run used (as in observation phase).
    • Example:
      noderacer control -s random -r 10 -h log/<ID_FOR_COMMAND>/noderacer1.hb.json node tests/micro-benchmark/paper-examples/archive-like-test.js
    • This command creates folder log/<ID_FOR_COMMAND>/exec? with log files that track the execution and decisions of each guided run.
    • See this module code here; in particular, the implementation of the guided execution described in the paper can be seen here.
  • Reports and diagnosis
    • All artefacts generated by NodeRacer (JSON files and images) can be found in the log folder determined (i.e., folder log/<ID_FOR_COMMAND>). A preliminary version of NodeRacer's web report can be seen by running noderacer report <pathtologfolder>.
    • The diagnosis mode is implemented as a different strategy for the guided execution phase (parameter 1-postpone-history). To use it, run noderacer control -s 1-postpone-history -r <runs> -h <hbfile> <command>.

Experimental package

  • Download and unzip the benchmark applications from here.

    tar -xvf noderacer-benchmarks.tar.gz
    • This zip file also includes a compiled version of the Node.fz tool, obtained from here.
  • Set up the location of the benchmarks' folder

    • Edit file tests/experiments/benchmark_config.js (see it here)
    • Update it to the folder where noderacer-benchmarks.tar.gz was extracted. This configuration file is used by the experiments' scripts.
  • RQ1 (comparison with Node.fz)

    • To collect results related to the comparisons, run the following commands. Specific details and configurations are described as comments in the scripts.
      node tests/experiments/known-bugs/start-noderacer.js 
      node tests/experiments/known-bugs/start-nodefz.js
    • See the scripts here.
    • The raw data collected and used in the paper and the R script can be seen here.
  • RQ2 (no happens-before used - NR-naive)

    • To collect results related to NR-naive, run the following command. Specific details and configurations are described as comments in the script.
      node tests/experiments/naive/start-naive.js   
    • See the scripts here.
    • The raw data collected and used in the paper can be seen here.
  • RQ3 (open issues)

    • The full description for each issue can be read here.
    • For each open issue analyzed, we provided a script that runs NodeRacer. Specific details and configurations to reproduce the analysis of each issue are described as comments in these scripts. See the scripts here.
    • We used a scrapper to collect GitHub data. The scrapper code can be seen here.
  • RQ4 (existing test suites)

    • To run all the 159 tests that have callback interleavings, see the script here.
      node tests/experiments/exploratory/all-tests/start-run-all.js  
    • For each test that failed at least once in 100 runs, we conducted an investigation and provided a script that runs NodeRacer. Specific details and configurations to reproduce the analysis of each failing test are described as comments in the scripts; see the scripts here. For example, the 1st test of mongo-express that failed can be run with:
      node tests/experiments/exploratory/failing-tests/mongo-express-1.js  
    • To replicate the steps performed to filter the tests used in the experiment, see here.

    License

Copyright 2019 casa.au.dk

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

Releases

No releases published

Packages

No packages published