https://github.com/abailly/raptr

Raft distributed consensus over HTTP
https://github.com/abailly/raptr

Last synced: 11 months ago
JSON representation

Raft distributed consensus over HTTP

• Host: GitHub
• URL: https://github.com/abailly/raptr
• Owner: abailly
• License: bsd-3-clause
• Created: 2015-12-26T08:19:31.000Z (over 10 years ago)
• Default Branch: master
• Last Pushed: 2020-02-07T07:12:00.000Z (over 6 years ago)
• Last Synced: 2025-07-26T10:59:03.707Z (11 months ago)
• Language: Haskell
• Size: 57.6 KB
• Stars: 7
• Watchers: 4
• Forks: 2
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
Raptr

=====

> raptr (n.m.): old norse for raft

Raptr is a burgeonning implementation of a [Raft](https://ramcloud.stanford.edu/wiki/download/attachments/11370504/raft.pdf) cluster, based on

[Kontiki](https://github.com/NicolasT/kontiki),

that aims at providing an out-of-the-box library for building linearizable distributed data storage systems communicating over plain

HTTP(S), in Haskell. 

[![Build Status](https://travis-ci.org/capital-match/raptr.png?branch=master)](https://travis-ci.org/capital-match/raptr)

# Usage

TBD

# Requirements

* Raptr should be easily embeddable within a host system and impose minimal requirements in terms of infrastructure and dependency,

  hence the use of plain HTTP as transport protocol and opaque `ByteString`s as "commands" to be linearized by the system,

* Raptr should be safe and correct, e.g. live up to the promises of *linearizability* and *resilience* of underlying protocol:

    * Linearizable means that concurrent writes appear to take effect instantaneously at some point between start and end of write

      request by a client,

    * Resilient means *committed* data must survive irrecoverable crash of a minority of servers in a cluster,

# Contributing

Pull requests are welcomed.