According to experience, the effect of increasing this value is kind of slight, and the suggestion is to increase this value only if the map output is very large in your job.

根据个人经验，增加这个值的效果不太明显，建议只在map输出非常大的情况下增加这个值。

youdao

Then MapReduce shuffles the map output by key and does another parallel computation on the redistributed map output, writing the results to the file system in the Reduce phase of the computation.

然后MapReduce按照各个键改组映射输出，对重新分配的映射输出执行另一项并行计算，在计算的归纳阶段将结果写入到文件系统中。

youdao

Then MapReduce shuffles the map output by key and does another parallel computation on the redistributed map output, writing the results to the file system in the Reduce phase of the computation.

然后MapReduce按照各个键改组映射输出，对重新分配的映射输出执行另一项并行计算，在计算的归纳阶段将结果写入到文件系统中。

youdao