Bugmark Simulation Platform

Description

The Bugmark Simulation Platform is a system for research purposes. The system builds on the Bugmark Exchange and adapts Test_Bench. The system architecture consists of the Bugmark Exchange, an Issue Tracker, several Scripts for automated behavior, and the User Interface.

Goal

The Bugmark Simulation Platform is developed for research experiments. The objective of the research is to investigate how developers conceptualize the use of Bugmark market mechanisms. Specifically, we are interested in how signals about the Market and Open Source Project Health are related and compete. For more information on the setup, please read the documentation on how to run an experiment.

Documentation

• How to Install and Configure
• How to Run an Experiment
• Where to Find Everything: Software Design
  • Structure of Repository
  • Use Case Diagram
  • Activity Diagrams

How to Install

1. Installation of Bugmark Exchange

This covers the config and install of software components for Bugmark Exchange.

Required Skills

To be successful, we recommend good skills with the following:

• Git
• Linux command line
• Web development
• PostgreSQL
• Ruby on Rails
• Slim template engine
• Tmux (nice to have)
• InfluxDB
• Grafana

0) Host Machine

We require Ubuntu v16.04 for your host machine. Newer versions have known issues.

Your host machine can exist in a few different forms:

1. a desktop Ubuntu v16.04 system
2. a Virtual Machine running locally (using Vagrant)
3. a Virtual Machine running in the data center

Of the three options, the best and simplest is 3), running in the data center. We like Linode - you can allocate a cheap node for development that will cost $20/month.

WARNING: if you choose to install on a local system (option 1), this configuration process will install many packages and will make changes to your user configuration, including:

• adding items to your .bashrc, modifying your path
• adding your UserID to sudoers

In this case, it is usually best to use a dedicated user-id.

0.a) Development VM Configuration

Follow these steps to set up a working development environment running on an Ubuntu Virtual machine.

Let's get started:

1. Install VirtualBox and Vagrant on your host machine (Linux, Win, Mac OK)

2. Download the dev-machine Vagrantfile wget raw.githubusercontent.com/Bugmark-Simulator/exchange/master/Vagrantfile

3. From the directory with the Vagrantfile, run vagrant up to create a virtual machine.

4. Login to your virtual machine using vagrant ssh

1.a) Cloning the Bugmark Exchange

NOTE: Make sure you are not working as root user. If necessary, create a new user on Ubuntu. E.g. sudo adduser bugmarkstudy; sudo adduser bugmarkstudy sudo

• to switch to new user, either logout and back in, or run su - bugmarkstudy

1. Clone the bugmark exchange repository mkdir ~/src; cd ~/src; git clone https://github.com/Bugmark-Simulator/exchange.git

2. Open the repository directory cd exchange

1.b) Development Environment Provisioning

On the host machine:

1. Checkout the dev branch git checkout -b dev origin/dev (TODO: TEST whether it works when in master -- all of our changes are in master)

2. Install ansible script/dev/provision/install_ansible

3. Install ansible roles script/dev/provision/install_roles

4. Provision the dev machine script/dev/provision/localhost

• If Node.js or NPM fail, make sure they are installed npm -v
  • If not, add the NodeSource APT repository curl -sL https://deb.nodesource.com/setup_6.x | sudo -E bash -
  • Then install Node.js sudo apt-get install -y nodejs
• If tasks 'influxdb : setup admin user' and 'influxdb : create database' fail, then
  • Check that influxdb is installed and running systemctl status influxdb
  • If it is not running, start it sudo systemctl start influxdb
  • Re-run the provision script in step 4
• If any other task fails, try re-running the script, sometimes that help
• If an error comes up about sudo requiring a password, run sudo ls before retrying

5. Check database status: systemctl status postgresql

6. Set server timezone to UTC within sudo dpkg-reconfigure tzdata

7. Also, set the TZ variable for bash vim ~/.bashrc and add at the bottom of file export TZ=UTC

• If vim thorws an error on startup, run command :PlugInstall

8. Start a new shell: bash (required to load your new user configuration)

9. Checkout the master branch git checkout master

1.c) Application Bootstrap

Follow these steps to bootstrap the app in your development environment.

1. Install ruby gems gem install bundler; bundle install

2. Install NPM components: yarn install

3. Create databases bundle exec rails db:create

4. Run migrations bundle exec rails db:migrate

5. Start the tmux session tmux

   A cheat-sheet for tmux navigation is in ~/.tmux.conf.

1.d) Host Web Access

1. Get the host IP address ifconfig

2. On your local machine, add the VM IP address to /etc/hosts

3. We will refer to your VM IP address as <hostname>

2.) Installation of Simulation Platform

This covers the installation of simulation platform to access Bugmark Exchange and run the simulation on top of Exchange.

1. On the server, go to src directory: cd ~/src

2. Git clone the Bugmark Simulator: git clone https://github.com/Bugmark-Simulator/bugmark_simulation.git

3. Create .env setting file cd bugmark_simulation/exercise/simulation; cp .env-default .env

4. Check which directory the .env links to cat .env and change if desired vim .env, then create that directory cd ~; mkdir trial; mkdir trial/simulation; mkdir trial/simulation/.trial_data

5. Create settings cd ~/trial/simulation; mkdir settings; cp ~/src/bugmark_simulation/sample_settings/* ./settings

6. Clean database and setup the default admin user ~/src/bugmark_simulation/exercise/simulation/script/reset_scr

7. Go to the simulation application folder cd ~/src/bugmark_simulation/exercise/simulation/webapp/

8. Start the simulation platform ./run

9. On your local machine, browse to http://<hostname>:4567

Your platform is ready to go. The default admin account is

• user: [email protected]
• password: bugmark

3.) Setting up Grafana

(Note: Make sure grafana is in version 5.0+)

1. Enable Grafana for users without requiring signing by editing file /etc/grafana/grafana.ini and setting:

    [auth.anonymous]
    enabled = true
    org_name = Main Org.
    org_role = Viewer

2. Test if Grafana is running systemctl status grafana-server and if not, start Grafana sudo systemctl start grafana-server, otherwise restart to load new settings sudo systemctl restart grafana-server

3. Grafana should be available from a browser http://<hostname>:3030 with username admin and password admin

4. In Grafana, add a data source for the influxDB connection

• name: influx
• type: influxDB
• URL: http://localhost:8086
• Access: proxy
• Basic Auth (yes)
• user: admin
• pasword: admin
• database: bugm_stats
• user: admin
• password: admin

5. Only after you have data in InfluxDB does it make sense to setup a panels because Grafana will only allow you to setup visualizations for existing data.

6. Create PostgreSQL user for grafana. (How to connect to PostgreSQL is described in the helpful commands section below.)

• In tmux/bash: Open PostgreSQL command:
  1. Logon as user postgres sudo -u postgres -i
  2. Start psql
• CREATE USER grafana WITH PASSWORD 'grafana';
• GRANT SELECT ON ALL TABLES IN SCHEMA public TO grafana;

7. In Grafana, add new data source: PostgreSQL connection

• Name: PSQL
• Type: PostgreSQL
• Host: localhost:5432
• Database: bugmark_development
• User: grafana
• Password: grafana

8. If Grafana is showing an error that it cannot access the table "trackers", give permissions

sudo -u postgres -i

# then as the postgres user...
psql -d bugmark_development
GRANT SELECT ON trackers to grafana

9. To setup the panels that display the graphs, go to http://<ip-address-of-server>:3030/dashboard/import and either select "Upload .json File" or copy past the content of grafana-panels.json - Note: Might have to login first at http://<ip-address-of-server>:3030/login using username: admin and password: admin - Note: This Grafana dashboard uses Variables. The variables need only be updated if you want to view data from the dashboard.

10. By default, URLs to view dashboards should not have changed. If graphs do not show up in the simulation, try trouble shooting with the AppHelpers function update_graphs.

Resetting Bugmark Exchange and all the database

Run script reset_scr to reset exchange and database ~/src/bugmark_simulation/exercise/simulation/script/reset_scr

Alternatively, to destroy the database and setup from scratch:

1. Go to the exchange cd ~\src\exchange
2. Destroy Database and reconstruct it bundle exec rails db:drop db:create db:migrate

• If this fails, try restarting the PostgreSQL server sudo systemctl stop postgresql; sudo systemctl start postgresql

3. Run script reset_scr to reset exchange ~/src/bugmark_simulation/exercise/simulation/script/reset_scr

Running the experiment

If InfluxDB or Grafana are not running:

• sudo systemctl start influxdb
• sudo systemctl start grafana-server

Following are the steps to setup the experiment

1. Run the script to clean the bugmark excange ~/src/bugmark_simulation/exercise/simulation/script/reset_scr
2. Run the script to create users ./bugmark_simulation/exercise/simulation/script/user_gen_scr
3. Go to the simulation application folder cd ~/src/bugmark_simulation/exercise/simulation/webapp/
4. Start the simulation platform ./run
5. On your local machine, browse to http://<hostname>:4567
6. Login to using the admin account. Default:

• username: [email protected]
• password: bugmark

From the menu, go to "RESEARCHER CONTROL PANEL".

1. Configure bots by clicking on "Setup Bot". Configuration instructions are provided in the application.
2. The button "Open Users Admin" displays usernames and passwords for participants. This page allows you as the admin to login as any user for trouble shooting and testing.
3. Start the Simulation by clicking on the yellow button "Simulation is NOT running"

Simulate user behavior

To test the visualizations, data analysis, or simply see how the system behaves with many users, run a simulation script that executes simple user actions with some randomization.

1. Install Bugmark Exchange, Simulation Software, Setup Grafana, and configure the software according to the install instructions.
2. To clean the database, run ~/src/bugmark_simulation/exercise/simulation/script/reset_scr
3. Login to the RESEARCHER CONTROL PANEL using the admin account.
4. The system should by default be setup with four funders. Next step is to create users.
5. To simulate the behavior of funders and workers who create offers, do work, and form contracts, run ~/src/bugmark_simulation/exercise/simulation/script/simulate_worker_funder

Helpful Commands

Open PostgreSQL database in console

1. Logon as user postgres sudo -u postgres -i
2. Start psql
3. List databases \l
4. Switch to bugmark_development database \c bugmark_development
5. List tables \dt
6. View definition of a table (e.g. views) \d+ views

Roadmap

• Software is developed in Summer 2018.
• Pilot studies are run in December 2018.
• Experiment is executed in January 2019.
• Data analysis is planned for Spring of 2019.

How to Contribute

Report an Issue

Please file an issue in the issue tracker.

Contribute Code

Please fork the repository, commit changes to your fork, and create a pull request. Please describe in the pull request what changes were made and why -- reference all issue that the change is motivated by. Create one pull request for each fixed bug or feature. Use a Feature Branch Workflow.

How to Become a Repository Maintainer

We do not plan on adding repository maintainers.

Contributors

• Vinod Ahuja (Repository Maintainer)
• Georg Link (Repository Maintainer)

License

Mozilla Public License Version 2.0

We chose this license because the Bugmark Exchange uses it.

Please add the license at the top of every source code file, like this: SPDX-License-Identifier: MPL-2.0