Hadoop Tutorial (vii) sorting in shuffle

1.map write to buffer time, pre-order (for the back of the fast row)

2.spill, two times, Fast platoon.

3. Again according to Partioner sort, each partioner in accordance with key sort

4. All spill files will be merged into an index file and a data file before the map task is completed

5. When all the map outputs are copied, the Reduce task enters the sorting phase (more appropriately the merge phase, since the sort is done on the map side), which merges all the map outputs, which can be repeated several times to complete

There are about 5 sort of

See more highlights of this column: http://www.bianceng.cnhttp://www.bianceng.cn/webkf/tools/