https://github.com/eladkehat/couch-rqs

A Relaxed Queue Service
https://github.com/eladkehat/couch-rqs

Last synced: 27 days ago
JSON representation

A Relaxed Queue Service

• Host: GitHub
• URL: https://github.com/eladkehat/couch-rqs
• Owner: eladkehat
• License: mit
• Created: 2011-05-21T07:48:37.000Z (over 13 years ago)
• Default Branch: master
• Last Pushed: 2011-11-19T06:42:20.000Z (about 13 years ago)
• Last Synced: 2023-03-12T01:26:06.424Z (almost 2 years ago)
• Language: Java
• Homepage:
• Size: 231 KB
• Stars: 11
• Watchers: 3
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.markdown
  • License: LICENSE

Awesome Lists containing this project

README

        
Relaxed Queue Service

=====================

Couch-RQS (Relaxed Queue Service) is an open source queue system that enables

applications to submit and receive messages through a web interface.

It is based on CouchDB, a robust, fast and easy to use document-oriented

database - http://couchdb.apache.org/

The queue interface is provided by Java code (CouchDB itself is written in

Erlang).

Inspiration

-----------

The inspiration for Couch-RQS was [Amazon SQS][0] (Simple Queue Service).

Like SQS, RQS lets components of your application communicate through a queue

buffer that is accessible through HTTP. However, RQS trades-off SQS' high

availability guarantee for real queue functionality (see the wiki pages for

a detailed comparison).

Also, RQS is completely free and runs on your own infrastructure - completely

under your control.

Key Features

------------

* The queue is accessible as a web service. Different components in your

application can access it from multiple locations - just run a CouchDB instance

on a machine that is accessible to the rest of your applications.

* Guaranteed message delivery: dequeued messages remain in the queue, hidden,

until deleted by the acquiring process. If that process crashes before it

finishes processing the message, the message will appear again after a preset

timeout period, to be consumed by another process.

* Supports FIFO/LIFO functionality. Messages are guaranteed to be consumed in

sequence.

* Very large message payload size (up to 4GB, limited by available RAM on the

server - as per CouchDB limitations).

* Messages are guaranteed to arrive only once, and are guaranteed to arrive if

they're available. No need to code around these SQS limitations.

More Goodness

-------------

* Relies on the very robust and generally wonderful CouchDB.

* It is simple to implement a hot backup (or multiple backups) of your queue

data across the network using CouchDB's replication mechanism.

* Simple Java API.

* Unlike SQS, it's open source and completely free.

Limitations

-----------

* Couch-RQS cannot run safely in a distributed / replicated environment.

Check out the wiki pages to learn more

[0]: http://aws.amazon.com/sqs/