Last week I gave a talk about processing data with the MapReduce framework using Gearman::XS and the gearmand server. Check out the pdf and demo code.
The demo code processes a small corpus of financial filings and is modelled after Jeffrey Dean's and Sanjay Ghemawat's original paper on MapReduce, entitled MapReduce: Simplified Data Processing on Large Clusters.
Also, here are the slides on sideshare: