http://philippeadjiman.com/blog/2010/01/07/hadoop-tutorial-series-issue-3-counters-in-action/ Just Another Blog About Geek Stuff, by Philippe Adjiman Tue, 23 Aug 2011 18:43:29 -0400 http://wordpress.org/?v=2.8.4 hourly 1 http://philippeadjiman.com/blog/2010/01/07/hadoop-tutorial-series-issue-3-counters-in-action/comment-page-1/#comment-486 Iterative algorithms in Hadoop « Kenkyuu Fri, 10 Jun 2011 19:18:56 +0000 http://philippeadjiman.com/blog/?p=818#comment-486 [...] The general idea for iterative algorithms in MapReduce is to chain multiple jobs together, using the output of the last one as the input of the next one. An important consideration is that, given the usual size of the data, the termination condition must be computed within the MapReduce program. The standard MapReduce model does not offer simple elegant ways to do this, but Hadoop has some added features that simplify this task: Counters. [...] [...] The general idea for iterative algorithms in MapReduce is to chain multiple jobs together, using the output of the last one as the input of the next one. An important consideration is that, given the usual size of the data, the termination condition must be computed within the MapReduce program. The standard MapReduce model does not offer simple elegant ways to do this, but Hadoop has some added features that simplify this task: Counters. [...]

]]>