blog/BigData-Notes
2
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Update Hadoop-MapReduce.md

Browse Source
This commit is contained in:
heibaiying 2019-03-14 14:07:50 +08:00 committed by GitHub
parent cf17652c35
commit b647004213
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 13 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
notes/Hadoop-MapReduce.md
View File

@ -1,5 +1,6 @@
# 分布式计算框架——MapReduce
<nav>
<nav>
<a href="#一MapReduce-概述">一、MapReduce 概述</a><br/>
<a href="#二MapReduce-编程模型简述">二、MapReduce 编程模型简述</a><br/>
<a href="#三MapReduce-编程模型详述">三、MapReduce 编程模型详述</a><br/>
@ -80,7 +81,7 @@ combiner是map运算后的可选操作，其实际上是一个本地化的reduce
但并非所有场景都适合使用combiner，使用它的原则是combiner的输入不会影响到reduce计算的最终输入，例如：如果计算只是求总数，最大值，最小值可以使用combiner，但是做平均值计算使用combiner的话，最终的reduce计算结果就会出错。
<div align="center"> <img width="600px" src="https://github.com/heibaiying/BigData-Notes/blob/master/pictures/mapreduce-combiner.png"/> </div>
<div align="center"> <img src="https://github.com/heibaiying/BigData-Notes/blob/master/pictures/mapreduce-combiner.png"/> </div>
### 3.3 partitioner