Hi, Can someone describe how MapReduce is implemented when hadoop's job is processed? How does shuffling work? How is Reduce node chosen?
Is it possible to speedup job scheduling by changing standard settings? I saw in Debug log that some transactions are initiated is it possible to turn it off? Regards -- View this message in context: http://apache-ignite-users.70518.x6.nabble.com/ignite-as-Hadoop-s-Map-Reduce-engine-how-it-works-tp10551.html Sent from the Apache Ignite Users mailing list archive at Nabble.com.
