咱们一般写mapreduce是通过java和streaming来写的,身为pythoner的我,

java不会,没办法就用streaming来写mapreduce日志分析。 这里要介绍一个

模块,是基于streaming搞的东西。

mrjob 可以让用 Python 来编写 MapReduce 运算,并在多个不同平台上运行,你可以:

  • 使用纯 Python 编写多步的 MapReduce 作业

  • 在本机上进行测试

  • 在 Hadoop 集群上运行

pip 的安装方法:

pip install mrjob

我测试的脚本

#coding:utf-8from mrjob.job import MRJobimport re#xiaorui.cc#WORD_RE = re.compile(r"[\w']+")WORD_RE = re.compile(r"\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}")class MRWordFreqCount(MRJob):    def mapper(self, word, line):        for word in WORD_RE.findall(line):            yield word.lower(), 1    def combiner(self, word, counts):        yield word, sum(counts)    def reducer(self, word, counts):        yield word, sum(counts)if __name__ == '__main__':    MRWordFreqCount.run()

用法算简单:

python i.py -r inline input1 input2 input3 > out 命令可以将处理多个文件的结果输出到out文件里面。

本地模拟hadoop运行:python 1.py -r local <input> output

这个会把结果输出到output里面,这个output必须写。

hadoop集群上运行:python 1.py -r hadoop <input> output

执行脚本 ~

[root@kspc ~]# python mo.py -r local  <10.7.17.7-dnsquery.log.1> outputno configs found; falling back on auto-configurationno configs found; falling back on auto-configurationcreating tmp directory /tmp/mo.root.20131224.040935.241241reading from STDINwriting to /tmp/mo.root.20131224.040935.241241/step-0-mapper_part-00000> /usr/bin/python mo.py --step-num=0 --mapper /tmp/mo.root.20131224.040935.241241/input_part-00000 | sort | /usr/bin/python mo.py --step-num=0 --combiner > /tmp/mo.root.20131224.040935.241241/step-0-mapper_part-00000writing to /tmp/mo.root.20131224.040935.241241/step-0-mapper_part-00001> /usr/bin/python mo.py --step-num=0 --mapper /tmp/mo.root.20131224.040935.241241/input_part-00001 | sort | /usr/bin/python mo.py --step-num=0 --combiner > /tmp/mo.root.20131224.040935.241241/step-0-mapper_part-00001Counters from step 1:  (no counters found)writing to /tmp/mo.root.20131224.040935.241241/step-0-mapper-sorted> sort /tmp/mo.root.20131224.040935.241241/step-0-mapper_part-00000 /tmp/mo.root.20131224.040935.241241/step-0-mapper_part-00001writing to /tmp/mo.root.20131224.040935.241241/step-0-reducer_part-00000> /usr/bin/python mo.py --step-num=0 --reducer /tmp/mo.root.20131224.040935.241241/input_part-00000 > /tmp/mo.root.20131224.040935.241241/step-0-reducer_part-00000writing to /tmp/mo.root.20131224.040935.241241/step-0-reducer_part-00001> /usr/bin/python mo.py --step-num=0 --reducer /tmp/mo.root.20131224.040935.241241/input_part-00001 > /tmp/mo.root.20131224.040935.241241/step-0-reducer_part-00001Counters from step 1:  (no counters found)Moving /tmp/mo.root.20131224.040935.241241/step-0-reducer_part-00000 -> /tmp/mo.root.20131224.040935.241241/output/part-00000Moving /tmp/mo.root.20131224.040935.241241/step-0-reducer_part-00001 -> /tmp/mo.root.20131224.040935.241241/output/part-00001Streaming final output from /tmp/mo.root.20131224.040935.241241/outputremoving tmp directory /tmp/mo.root.20131224.040935.241241

执行的时候,资源的占用情况。

发现一个很奇妙的东西,mrjob居然调用shell下的sort来排序。。。。

为了更好的理解mrjob的用法,再来个例子。

from mrjob.job import MRJob#from xiaorui.ccclass MRWordFrequencyCount(MRJob):#把东西拼凑起来    def mapper(self, _, line):        yield "chars", len(line)        yield "words", len(line.split())        yield "lines", 1#总结kv    def reducer(self, key, values):        yield key, sum(values)if __name__ == '__main__':    MRWordFrequencyCount.run()

看下结果:

下面是官网给的一些个用法:

我们可以看到他是支持hdfs和s3存储的 !

Running your job different ways

The most basic way to run your job is on the command line:

$ python my_job.py input.txt

By default, output will be written to stdout.

You can pass input via stdin, but be aware that mrjob will just dump it to a file first:

$ python my_job.py < input.txt

You can pass multiple input files, mixed with stdin (using the - character):

$ python my_job.py input1.txt input2.txt - < input3.txt

By default, mrjob will run your job in a single Python process. This provides the friendliest debugging experience, but it’s not exactly distributed computing!

You change the way the job is run with the -r/--runner option. You can use -rinline (the default), -rlocal, -rhadoop, or -remr.

To run your job in multiple subprocesses with a few Hadoop features simulated, use -rlocal.

To run it on your Hadoop cluster, use -rhadoop.

If you have Elastic MapReduce configured (see ), you can run it there with -remr.

Your input files can come from HDFS if you’re using Hadoop, or S3 if you’re using EMR:

$ python my_job.py -r emr s3://my-inputs/input.txt$ python my_job.py -r hadoop hdfs://my_home/input.txt