Writing Hadoop MapReduce program using PHP and Shell_PHP Tutorial

Enables any executable program that supports standard IO (stdin, stdout) to become a hadoop mapper or reducer. For example:

Copy the code The code is as follows:

hadoop jar hadoop-streaming.jar -input SOME_INPUT_DIR_OR_FILE -output SOME_OUTPUT_DIR -mapper / bin/cat -reducer /usr/bin/wc

In this example, the cat and wc tools that come with Unix/Linux are used as mapper/reducer. Isn’t it amazing?

If you are used to using some dynamic languages, use dynamic languages ​​to write mapreduce. It is no different from previous programming. Hadoop is just a framework to run it. Let me demonstrate how to use PHP to implement mapreduce of Word Counter.

1. Find the Streaming jar

There is no hadoop-streaming.jar in the Hadoop root directory. Because streaming is a contrib, you have to find it under the contrib. Taking hadoop-0.20.2 as an example, it is here:

Copy code The code is as follows:

$HADOOP_HOME/contrib/streaming/hadoop-0.20.2-streaming.jar

2. Write Mapper

Create a new wc_mapper.php and write the following code:

Copy code The code is as follows:

#!/usr/bin/php
$in = fopen(“php://stdin”, “r”);
$results = array();
while ( $line = fgets($in, 4096) )
{
$words = preg_split('/W/', $line, 0, PREG_SPLIT_NO_EMPTY);
foreach ($words as $word)
$results[] = $word;
}
fclose ($in);
foreach ($results as $key => $value)
{
print “$valuet1n”;
}

The general meaning of this code is: find the words in each line of input text and output it in the form of "
hello 1
world 1"
.

It’s basically no different from the PHP I wrote before, right? There are two things that may make you feel a little strange:

PHP as an executable program

The "#!/usr/bin/php" in the first line tells Linux to use the program /usr/bin/php as the interpreter for the following code. People who have written Linux shells should be familiar with this writing method. The first line of every shell script is like this: #!/bin/bash, #!/usr/bin/python

With this line, after saving the file, you can directly execute wc_mapper.php as cat and grep commands like this: ./wc_mapper.php

Use stdin to receive input

PHP supports multiple methods of passing in parameters. The most familiar ones should be to get the parameters passed through the Web from the $_GET, $_POST super global variables, and the second is to get the parameters passed from $_SERVER['argv'] Parameters passed in from the command line. Here, the standard input stdin

is used.

The effect of its use is:

Enter ./wc_mapper.php in the linux console

wc_mapper.php runs, and the console enters the state of waiting for user keyboard input

User enters text via keyboard

The user presses Ctrl + D to terminate the input, wc_mapper.php starts executing the real business logic and outputs the execution results

So where is stdout? Print itself is already stdout, which is no different from when we wrote web programs and CLI scripts before.

3. Write Reducer

Create a new wc_reducer.php and write the following code:

Copy the code The code is as follows:

#!/usr /bin/php
$in = fopen(“php://stdin”, “r”);
$results = array();
while ( $line = fgets($in, 4096) )
{
list($key, $value) = preg_split(“/t/”, trim($line), 2);
$results[$key] += $value;
}
fclose($in);
ksort($results);
foreach ($results as $key => $value)
{
print “$keyt$valuen”;
}

The main idea of ​​this code is to count how many times each word appears and output it in the form of "
hello 2
world 1"
.

4. Use Hadoop to run

Upload the sample text to be counted

Copy the code The code is as follows:

hadoop fs - put *.TXT /tmp/input

Execute PHP mapreduce program in Streaming mode

Copy code The code is as follows:

hadoop jar hadoop-0.20.2-streaming.jar -input /tmp/input -output /tmp /output -mapper absolute path to wc_mapper.php -reducer absolute path to wc_reducer.php

Note:

The input and output directories are paths on HDFS

The mapper and reducer are paths on the local machine. Be sure to write absolute paths, do not write relative paths, otherwise Hadoop will report an error saying that the mapreduce program cannot be found.

View results

Copy code The code is as follows:

hadoop fs -cat /tmp/output/part -00000

5. Shell version of Hadoop MapReduce program

Copy code The code is as follows:

#!/bin/bash -

# Load configuration file
source './config.sh'

# Process command line parameters
while getopts "d:" arg
do
case $arg in
d)
date=$OPTARG

?)
                                                                                                                                                                                                                              been have – echo "unkonw argument"

# The default processing date is yesterday
default_date=`date -v-1d +%Y-%m-%d`

# Final processing date. If the date format is incorrect, exit execution

date=${date:-${default_date}}
if ! [[ "$date" =~ [12][0- 9]{3}-(0[1-9]|1[12])-(0[1-9]|[12][0-9]|3[01]) ]]

then

echo "invalid date(yyyy-mm-dd): $date"
exit 1
fi

# Files to be processed
log_files=$(${hadoop_home}bin/hadoop fs -ls ${log_file_dir_in_hdfs} | awk '{print $8}' | grep $date)

# If the number of files to be processed is zero, exit execution

log_files_amount=$(($(echo $log_files | wc -l) + 0))
if [ $log_files_amount -lt 1 ]

then

echo "no log files found"
exit 0
fi

# Input file list
for f in $log_files
do

input_files_list="${input_files_list} $f"

done

function map_reduce () {
if ${hadoop_home}bin/hadoop jar ${streaming_jar_path} -input${input_files_list} -output ${mapreduce_output_dir}${date}/${1}/ -mapper "$ {mapper} ${1}" -reducer "${reducer}" -file "${mapper}"
then

echo "streaming job done!"

else
exit 1
fi
}

# Loop through each bucket
for bucket in ${bucket_list[@]}
do

map_reduce $bucket

done

http://www.bkjia.com/PHPjc/754798.html

www.bkjia.com

truehttp: //www.bkjia.com/PHPjc/754798.htmlTechArticle enables any executable program that supports standard IO (stdin, stdout) to become a hadoop mapper or reducer. For example: Copy the code The code is as follows: hadoop jar hadoop-streaming.jar -input...