Skip to content

veg/datamonkey-js-server

Repository files navigation

Datamonkey JS Server

Datamonkey JS Server works in conjunction with datamonkey-js. When the daemon is running, it acts as the job dispatcher and provisioner backend for jobs submitted to datamonkey-js.

Dependencies

Other versions might work, but not well tested.

  • node - 8.x.x
  • redis - 3.x.x
  • PBS server - 4.x.x
  • tn93 - 1.x.x

Dependencies are mostly handled by the package.json file and yarn or npm. Outside of the JavaScript dependencies, the daemon expects that a redis instance is running, and that there is a scheduler available that works with the command qsub. Redis stores transient data information such as qsub job status and standard output information.

Installation

Installation should be fairly straightforward as long as the dependencies are met.

  • git clone https://github.com/veg/datamonkey-js-server.git
  • cp config.json.tpl config.json
  • cp pm2.config.js.tpl pm2.config.js

Please make appropriate edits to config.json and pm2.config.js.

Configuration overview

 - port - The port that daemon will be hosted on
 - redis_host - The host that the redis is residing on
 - redis_port - The port that redis is listening on
 - loglevel - The log level, see the [syslog protocol](https://tools.ietf.org/html/rfc5424) for more information
 - tn93dist - The path to the tn93 binary installed on the system. (HIV-TRACE is the only method that uses it)
 - qsub_queue - Which qsub queue you wish to use
 - qsub_avx_queue - To be deprecated, specify queue that has avx nodes. (HyPhy > 2.3.0 depends on avx system calls being available)
  • make install
  • npm install
  • pm2 start pm2.config.js

Tests

Tests can be run using mocha. These tests are especially useful in ensuring that your system is configured correctly and that you are able to dispatch jobs for all analyses. Please see the test/ directory