Getting started

A Ruby gem is available for installation with RubyGems. To install it globally,

$ gem install qm

API client

A client for Ruby is under development. Basic low-level functions are now available in the stable release, but no effort has been made to leverage any of Ruby’s language-level concurrency features. It is possible to use the current functions in a Ruby program to submit jobs to be executed by volunteer compute nodes, but it is not yet convenient.

API server

The “teaching version” of QMachine has now been merged into the Ruby gem, and a basic web server is also provided to enable the use of web browsers as compute nodes if so desired.

A QM server can be launched by a Ruby program as shown in the following example, which includes default configuration values and commented database connection strings:

require 'qm'

    avar_ttl:           86400,      # expire avars after _ seconds
    enable_api_server:  false,
    enable_cors:        false,
    enable_web_server:  false,
    gc_interval:        60,         # evict unused avars every _ seconds
    hostname:           '',
    max_body_size:      65536,      # 64 * 1024 = 64 KB
    persistent_storage: {
      # mongo:          'mongodb://localhost:27017/test'
      # postgres:       'postgres://localhost:5432/' + ENV['USER']
      # redis:          'redis://'
      # sqlite:         'qm.db'
    port:               8177,
    public_folder:      'public',
    trafficlog_storage: {
      # mongo:          'mongodb://localhost:27017/test'
      # postgres:       'postgres://localhost:5432/' + ENV['USER']
    worker_procs:       1

The Ruby version of the API server has less flexibility than the original Node.js version does. There are now four choices to persist storage for the message-passing interface, but MongoDB is strongly recommended. Experimental support for PostgreSQL, Redis, and SQLite is available, but that support may be removed in the future. Currently, MongoDB is the only supported database for logging traffic data.