Mapreduce has the following advantages in data processing:
FirstThis model is very easy to use, even if it is completely unavailableProgramThe same is true for programmers. It hides details of parallel computing, error Disaster Tolerance, local optimization, and load balancing. Mapreduce running developers use familiar languages for development, such as Java, C #, Python, and C ++.
SecondMapreduce can be easily used for large-scale computing needs.
For example, Google uses mapreduce to provide web search services, sorting, data mining, machine learning, and other systems.
ThirdThrough mapreduce, applications can run on a large cluster with more than 1000 nodes and provide optimized error disaster tolerance.
For more information about mapreduce recommendations, see:
Http://en.wikipedia.org/wiki/MapReduce
Http://labs.google.com/papers/mapreduce.html