Distribute Data Process in Parallel Combine Results
← Back to MapReduce
The MapReduce framework partitions input data across nodes, applies a map function to each partition in parallel, shuffles intermediate results by key, and applies a reduce function to combine results. Originally developed at Google for web-scale data processing.