Mrs is a lightweight but high performance implementation of the MapReduce programming model for parallel computation. It is optimized for computationally intensive programs, especially those which are iterative, though it is applicable to a wide range of applications. It is intended to be simple to install and use. It is written in Python, supports both Python 2 and Python 3, and has no dependencies outside the standard library. The name "Mrs" derives from the practice of referring to MapReduce programs by the title "mister" (e.g., "Mr. Wordcount").
|Tags||Software Development Libraries Python Modules Scientific/Engineering Clustering/Distributed Networks|
|Operating Systems||Linux POSIX Mac OS X|
Release Notes: Prominent changes in this release include a relicensing under the Apache Software License instead of the GPL, support for data larger than the available RAM, extended documentation including a new User Guide, support for specifying custom serializers in addition to pickle, a new data-intensive example, Walk Analyzer, a maximum number of failures per task, and a variety of bugfixes.
Release Notes: This release offers significant improvements to performance, stability, and ease of use.
No changes have been submitted for this release.