Skip to content

damphyr/rplex

Repository files navigation

rplex

Build Status

Dead simple asynchronous data based job demultiplexing in Ruby.

Which is a very long winded way of saying that we sent a bunch of data and it lands in N queues where it waits for some process to pick them up.

No persistence, no databases, no fancy management API, no hassle when installing.

Why?

Because anything involving redis, postgres etc. is just a pain to install on Windows and we just wanted to parallelize a bunch of build tasks.

How do you use it?

In a CI system to reduce our build times by parallelizing test execution:

The main build finishes and posts as data the revision and the location of the build artifacts to rplex. rplex queues the data for every worker that has appeared until that point. Each worker polls the rplex queue, and then uses the information to grab the build and run a bunch of tests.

Is that all?

Queues are created automatically when a worker tries to get data from the queue, data can be posted to a subset of the workers if necessary and that, as they say, is it.

Really?

Well, it used to be all of it. Now, you can also limit the size of the queues and remove queues using the client (and the API)

##Example Start the rplex service.

Use Rplex::Client to post data:

Rplex::Client.new("name","http://rplex.host:7777/job").new_job(job_id,data)

Use Rplex::Processor to check the queue and do something with any data present

Rplex::Processor.new(Rplex::Client.new("name","http://rplex.host:7777/job"),5).run!{|job_data| p job_data}

REST

Not really, but pretty damn close. Check rplex/server.rb for the gruesome details, here's a whirlwind tour

  • GET / --> Version Info
  • GET /job/worker --> pop the worker queue
  • POST /job --> push a job in all or some of the queues present
  • GET /backlog --> how full is each queue?
  • GET /configuration --> how full can each queue get? 0 means whatever
  • GET /configuration/worker --> get worker's current configuration
  • POST /configuration --> change a queue's limit
  • DELETE /configuration/worker --> remove the worker queue
  • POST /reset --> empty all or some of the queues

TODO

Get some metrics collection done, write a persistent backend alternative for the queues.

LICENSE:

(The MIT License)

Copyright (c) 2011-2012 Vassilis Rizopoulos

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the 'Software'), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

About

Simple demultiplexer for build job parallelization

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages