3 ms·
MapReduce and Parallel DBMSs: Friends or Foes? [2010]
- _delirium 16y agoThe MapReduce folks have a paper in the same issue, giving their take on things (and objecting to a previously published benchmark paper): http://cacm.acm.org/magazines/2010/1/55744-mapreduce-a-flexible-data-processing-tool/fulltext http://cacm.acm.org/magazines/2010/1/55744-mapreduce-a-flexi... Earlier HN discussion of their paper is here: http://news.ycombinator.com/item?id=1019134 http://news.ycombinator.com/item?id=1019134. The paper linked in this post ("friends or foes?") was paywalled at the time, though, so wasn't included in the discussion.
- daniel02216 16y agoI'm surprised Stonebraker is still around deliberately misinterpreting how MapReduce works in order to sell more of his expensive database software. Some of his blog entries about MR vs databases a few years ago are laughably wrong: http://databasecolumn.vertica.com/database-innovation/mapreduce-a-major-step-backwards/ http://databasecolumn.vertica.com/database-innovation/mapred... and a rebuttal: http://typicalprogrammer.com/?p=16 http://typicalprogrammer.com/?p=16