为了方便 MapReduce 直接訪问关系型数据库（Mysql,Oracle）。Hadoop提供了DBInputFormat和DBOutputFormat两个类。通过DBInputFormat类把数据库表数据读入到HDFS，依据DBOutputFormat类把MapReduce产生的结果集导入到数据库表中。
执行MapReduce时候报错：java.io.IOException: com.mysql.jdbc.Driver，通常是因为程序找不到mysql驱动包。解决方法是让每一个tasktracker执行MapReduce程序时都能够找到该驱动包。
加入包有两种方式：
（1）在每一个节点下的${HADOOP_HOME}/lib下加入该包。重新启动集群，通常是比較原始的方法。
（2）a)把包传到集群上： hadoop fs -put mysql-connector-java-5.1.0- bin.jar /hdfsPath/
b)在mr程序提交job前，加入语句：DistributedCache.addFileToClassPath(new Path(“/hdfsPath/mysql- connector-java-5.1.0-bin.jar”),conf);
mysql数据库存储到hadoop hdfs
mysql表创建和数据初始化
DROP TABLE IF EXISTS `wu_testhadoop`;
CREATE TABLE `wu_testhadoop` (
`id` int(11) NOT NULL AUTO_INCREMENT,
`title` varchar(255) DEFAULT NULL,
`content` varchar(255) DEFAULT NULL,
PRIMARY KEY (`id`)
) ENGINE=InnoDB AUTO_INCREMENT=3 DEFAULT CHARSET=utf8; -- ----------------------------
-- Records of wu_testhadoop
-- ----------------------------
INSERT INTO `wu_testhadoop` VALUES ('1', '123', '122312');
INSERT INTO `wu_testhadoop` VALUES ('2', '123', '123456');
定义hadoop数据訪问
mysql表创建完成后，我们须要定义hadoop訪问mysql的规则。
hadoop提供了org.apache.hadoop.io.Writable接口来实现简单的高效的可序列化的协议，该类基于DataInput和DataOutput来实现相关的功能。
hadoop对数据库訪问也提供了org.apache.hadoop.mapred.lib.db.DBWritable接口，当中write方法用于对PreparedStatement对象设定值，readFields方法用于对从数据库读取出来的对象进行列的值绑定。
以上两个接口的使用例如以下（内容是从源代码得来）
writable
public class MyWritable implements Writable {
// Some data
private int counter;
private long timestamp;
public void write(DataOutput out) throws IOException {
out.writeInt(counter);
out.writeLong(timestamp);
}
public void readFields(DataInput in) throws IOException {
counter = in.readInt();
timestamp = in.readLong();
}
public static MyWritable read(DataInput in) throws IOException {
MyWritable w = new MyWritable();
w.readFields(in);
return w;
}
}
DBWritable
public class MyWritable implements Writable, DBWritable {
// Some data
private int counter;
private long timestamp;
//Writable#write() implementation
public void write(DataOutput out) throws IOException {
out.writeInt(counter);
out.writeLong(timestamp);
}
//Writable#readFields() implementation
public void readFields(DataInput in) throws IOException {
counter = in.readInt();
timestamp = in.readLong();
}
public void write(PreparedStatement statement) throws SQLException {
statement.setInt(1, counter);
statement.setLong(2, timestamp);
}
public void readFields(ResultSet resultSet) throws SQLException {
counter = resultSet.getInt(1);
timestamp = resultSet.getLong(2);
}
}
数据库相应的实现
package com.wyg.hadoop.mysql.bean; import java.io.DataInput;
import java.io.DataOutput;
import java.io.IOException;
import java.sql.PreparedStatement;
import java.sql.ResultSet;
import java.sql.SQLException; import org.apache.hadoop.io.Text;
import org.apache.hadoop.io.Writable;
import org.apache.hadoop.mapred.lib.db.DBWritable; public class DBRecord implements Writable, DBWritable{
private int id;
private String title;
private String content;
public int getId() {
return id;
} public void setId(int id) {
this.id = id;
} public String getTitle() {
return title;
} public void setTitle(String title) {
this.title = title;
} public String getContent() {
return content;
} public void setContent(String content) {
this.content = content;
} @Override
public void readFields(ResultSet set) throws SQLException {
this.id = set.getInt("id");
this.title = set.getString("title");
this.content = set.getString("content");
} @Override
public void write(PreparedStatement pst) throws SQLException {
pst.setInt(1, id);
pst.setString(2, title);
pst.setString(3, content);
} @Override
public void readFields(DataInput in) throws IOException {
this.id = in.readInt();
this.title = Text.readString(in);
this.content = Text.readString(in);
} @Override
public void write(DataOutput out) throws IOException {
out.writeInt(this.id);
Text.writeString(out, this.title);
Text.writeString(out, this.content);
} @Override
public String toString() {
return this.id + " " + this.title + " " + this.content;
}
}
实现Map/Reduce
package com.wyg.hadoop.mysql.mapper; import java.io.IOException; import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapred.MapReduceBase;
import org.apache.hadoop.mapred.Mapper;
import org.apache.hadoop.mapred.OutputCollector;
import org.apache.hadoop.mapred.Reporter; import com.wyg.hadoop.mysql.bean.DBRecord; @SuppressWarnings("deprecation")
public class DBRecordMapper extends MapReduceBase implements Mapper<LongWritable, DBRecord, LongWritable, Text>{ @Override
public void map(LongWritable key, DBRecord value,
OutputCollector<LongWritable, Text> collector, Reporter reporter)
throws IOException {
collector.collect(new LongWritable(value.getId()), new Text(value.toString()));
} }
測试hadoop连接mysql并将数据存储到hdfs
package com.wyg.hadoop.mysql.db;
import java.io.IOException; import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapred.FileOutputFormat;
import org.apache.hadoop.mapred.JobClient;
import org.apache.hadoop.mapred.JobConf;
import org.apache.hadoop.mapred.lib.IdentityReducer;
import org.apache.hadoop.mapred.lib.db.DBConfiguration;
import org.apache.hadoop.mapred.lib.db.DBInputFormat; import com.wyg.hadoop.mysql.bean.DBRecord;
import com.wyg.hadoop.mysql.mapper.DBRecordMapper; public class DBAccess {
public static void main(String[] args) throws IOException {
JobConf conf = new JobConf(DBAccess.class);
conf.setOutputKeyClass(LongWritable.class);
conf.setOutputValueClass(Text.class);
conf.setInputFormat(DBInputFormat.class);
Path path = new Path("hdfs://192.168.44.129:9000/user/root/dbout");
FileOutputFormat.setOutputPath(conf, path);
DBConfiguration.configureDB(conf,"com.mysql.jdbc.Driver", "jdbc:mysql://你的ip:3306/数据库名","username","password");
String [] fields = {"id", "title", "content"};
DBInputFormat.setInput(conf, DBRecord.class, "wu_testhadoop",
null, "id", fields);
conf.setMapperClass(DBRecordMapper.class);
conf.setReducerClass(IdentityReducer.class);
JobClient.runJob(conf);
}
}
运行程序，结果例如以下：
15/08/11 16:46:18 INFO jvm.JvmMetrics: Initializing JVM Metrics with processName=JobTracker, sessionId=
15/08/11 16:46:18 WARN mapred.JobClient: Use GenericOptionsParser for parsing the arguments. Applications should implement Tool for the same.
15/08/11 16:46:18 WARN mapred.JobClient: No job jar file set. User classes may not be found. See JobConf(Class) or JobConf#setJar(String).
15/08/11 16:46:19 INFO mapred.JobClient: Running job: job_local_0001
15/08/11 16:46:19 INFO mapred.MapTask: numReduceTasks: 1
15/08/11 16:46:19 INFO mapred.MapTask: io.sort.mb = 100
15/08/11 16:46:19 INFO mapred.MapTask: data buffer = 79691776/99614720
15/08/11 16:46:19 INFO mapred.MapTask: record buffer = 262144/327680
15/08/11 16:46:19 INFO mapred.MapTask: Starting flush of map output
15/08/11 16:46:19 INFO mapred.MapTask: Finished spill 0
15/08/11 16:46:19 INFO mapred.TaskRunner: Task:attempt_local_0001_m_000000_0 is done. And is in the process of commiting
15/08/11 16:46:19 INFO mapred.LocalJobRunner:
15/08/11 16:46:19 INFO mapred.TaskRunner: Task 'attempt_local_0001_m_000000_0' done.
15/08/11 16:46:19 INFO mapred.LocalJobRunner:
15/08/11 16:46:19 INFO mapred.Merger: Merging 1 sorted segments
15/08/11 16:46:19 INFO mapred.Merger: Down to the last merge-pass, with 1 segments left of total size: 48 bytes
15/08/11 16:46:19 INFO mapred.LocalJobRunner:
15/08/11 16:46:19 INFO mapred.TaskRunner: Task:attempt_local_0001_r_000000_0 is done. And is in the process of commiting
15/08/11 16:46:19 INFO mapred.LocalJobRunner:
15/08/11 16:46:19 INFO mapred.TaskRunner: Task attempt_local_0001_r_000000_0 is allowed to commit now
15/08/11 16:46:19 INFO mapred.FileOutputCommitter: Saved output of task 'attempt_local_0001_r_000000_0' to hdfs://192.168.44.129:9000/user/root/dbout
15/08/11 16:46:19 INFO mapred.LocalJobRunner: reduce > reduce
15/08/11 16:46:19 INFO mapred.TaskRunner: Task 'attempt_local_0001_r_000000_0' done.
15/08/11 16:46:20 INFO mapred.JobClient: map 100% reduce 100%
15/08/11 16:46:20 INFO mapred.JobClient: Job complete: job_local_0001
15/08/11 16:46:20 INFO mapred.JobClient: Counters: 14
15/08/11 16:46:20 INFO mapred.JobClient: FileSystemCounters
15/08/11 16:46:20 INFO mapred.JobClient: FILE_BYTES_READ=34606
15/08/11 16:46:20 INFO mapred.JobClient: FILE_BYTES_WRITTEN=69844
15/08/11 16:46:20 INFO mapred.JobClient: HDFS_BYTES_WRITTEN=30
15/08/11 16:46:20 INFO mapred.JobClient: Map-Reduce Framework
15/08/11 16:46:20 INFO mapred.JobClient: Reduce input groups=2
15/08/11 16:46:20 INFO mapred.JobClient: Combine output records=0
15/08/11 16:46:20 INFO mapred.JobClient: Map input records=2
15/08/11 16:46:20 INFO mapred.JobClient: Reduce shuffle bytes=0
15/08/11 16:46:20 INFO mapred.JobClient: Reduce output records=2
15/08/11 16:46:20 INFO mapred.JobClient: Spilled Records=4
15/08/11 16:46:20 INFO mapred.JobClient: Map output bytes=42
15/08/11 16:46:20 INFO mapred.JobClient: Map input bytes=2
15/08/11 16:46:20 INFO mapred.JobClient: Combine input records=0
15/08/11 16:46:20 INFO mapred.JobClient: Map output records=2
15/08/11 16:46:20 INFO mapred.JobClient: Reduce input records=2
同一时候能够看到hdfs文件系统多了一个dbout的文件夹，里边的文件保存了数据库相应的数据，内容保存例如以下
1 1 123 122312
2 2 123 123456
hdfs数据导入到mysql
hdfs文件存储到mysql，也须要上边的DBRecord类作为辅助。由于数据库的操作都是通过DBInput和DBOutput来进行的；
首先须要定义map和reduce的实现（map用以对hdfs的文档进行解析，reduce解析map的输出并输出）
package com.wyg.hadoop.mysql.mapper; import java.io.IOException;
import java.io.DataInput;
import java.io.DataOutput;
import java.sql.PreparedStatement;
import java.sql.ResultSet;
import java.sql.SQLException;
import java.util.Iterator; import org.apache.hadoop.filecache.DistributedCache;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.io.Writable;
import org.apache.hadoop.mapred.JobClient;
import org.apache.hadoop.mapred.MapReduceBase;
import org.apache.hadoop.mapred.Mapper;
import org.apache.hadoop.mapred.OutputCollector;
import org.apache.hadoop.mapred.Reducer;
import org.apache.hadoop.mapred.Reporter;
import com.wyg.hadoop.mysql.bean.DBRecord; public class WriteDB {
// Map处理过程
public static class Map extends MapReduceBase implements Mapper<Object, Text, Text, DBRecord> {
private final static DBRecord one = new DBRecord(); private Text word = new Text(); @Override public void map(Object key, Text value, OutputCollector<Text, DBRecord> output, Reporter reporter) throws IOException { String line = value.toString();
String[] infos = line.split(" ");
String id = infos[0].split(" ")[1];
one.setId(new Integer(id));
one.setTitle(infos[1]);
one.setContent(infos[2]);
word.set(id);
output.collect(word, one);
} } public static class Reduce extends MapReduceBase implements
Reducer<Text, DBRecord, DBRecord, Text> {
@Override
public void reduce(Text key, Iterator<DBRecord> values,
OutputCollector<DBRecord, Text> collector, Reporter reporter)
throws IOException {
DBRecord record = values.next();
collector.collect(record, new Text());
}
}
}
測试hdfs导入数据到数据库
package com.wyg.hadoop.mysql.db; import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.IntWritable;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapred.FileInputFormat;
import org.apache.hadoop.mapred.JobClient;
import org.apache.hadoop.mapred.JobConf;
import org.apache.hadoop.mapred.TextInputFormat;
import org.apache.hadoop.mapred.lib.db.DBConfiguration;
import org.apache.hadoop.mapred.lib.db.DBInputFormat;
import org.apache.hadoop.mapred.lib.db.DBOutputFormat; import com.wyg.hadoop.mysql.bean.DBRecord;
import com.wyg.hadoop.mysql.mapper.WriteDB; public class DBInsert {
public static void main(String[] args) throws Exception { JobConf conf = new JobConf(WriteDB.class);
// 设置输入输出类型 conf.setInputFormat(TextInputFormat.class);
conf.setOutputFormat(DBOutputFormat.class); // 不加这两句，通只是，可是网上给的样例没有这两句。
//Text, DBRecord
conf.setMapOutputKeyClass(Text.class);
conf.setMapOutputValueClass(DBRecord.class);
conf.setOutputKeyClass(Text.class);
conf.setOutputValueClass(DBRecord.class);
// 设置Map和Reduce类
conf.setMapperClass(WriteDB.Map.class);
conf.setReducerClass(WriteDB.Reduce.class);
// 设置输如文件夹
FileInputFormat.setInputPaths(conf, new Path("hdfs://192.168.44.129:9000/user/root/dbout"));
// 建立数据库连接
DBConfiguration.configureDB(conf,"com.mysql.jdbc.Driver", "jdbc:mysql://数据库ip:3306/数据库名称","username","password");
String[] fields = {"id","title","content" };
DBOutputFormat.setOutput(conf, "wu_testhadoop", fields);
JobClient.runJob(conf);
} }
測试结果例如以下
15/08/11 18:10:15 INFO jvm.JvmMetrics: Initializing JVM Metrics with processName=JobTracker, sessionId=
15/08/11 18:10:15 WARN mapred.JobClient: Use GenericOptionsParser for parsing the arguments. Applications should implement Tool for the same.
15/08/11 18:10:15 WARN mapred.JobClient: No job jar file set. User classes may not be found. See JobConf(Class) or JobConf#setJar(String).
15/08/11 18:10:15 INFO mapred.FileInputFormat: Total input paths to process : 1
15/08/11 18:10:15 INFO mapred.JobClient: Running job: job_local_0001
15/08/11 18:10:15 INFO mapred.FileInputFormat: Total input paths to process : 1
15/08/11 18:10:15 INFO mapred.MapTask: numReduceTasks: 1
15/08/11 18:10:15 INFO mapred.MapTask: io.sort.mb = 100
15/08/11 18:10:15 INFO mapred.MapTask: data buffer = 79691776/99614720
15/08/11 18:10:15 INFO mapred.MapTask: record buffer = 262144/327680
15/08/11 18:10:15 INFO mapred.MapTask: Starting flush of map output
15/08/11 18:10:16 INFO mapred.MapTask: Finished spill 0
15/08/11 18:10:16 INFO mapred.TaskRunner: Task:attempt_local_0001_m_000000_0 is done. And is in the process of commiting
15/08/11 18:10:16 INFO mapred.LocalJobRunner: hdfs://192.168.44.129:9000/user/root/dbout/part-00000:0+30
15/08/11 18:10:16 INFO mapred.TaskRunner: Task 'attempt_local_0001_m_000000_0' done.
15/08/11 18:10:16 INFO mapred.LocalJobRunner:
15/08/11 18:10:16 INFO mapred.Merger: Merging 1 sorted segments
15/08/11 18:10:16 INFO mapred.Merger: Down to the last merge-pass, with 1 segments left of total size: 40 bytes
15/08/11 18:10:16 INFO mapred.LocalJobRunner:
15/08/11 18:10:16 INFO mapred.TaskRunner: Task:attempt_local_0001_r_000000_0 is done. And is in the process of commiting
15/08/11 18:10:16 INFO mapred.LocalJobRunner: reduce > reduce
15/08/11 18:10:16 INFO mapred.TaskRunner: Task 'attempt_local_0001_r_000000_0' done.
15/08/11 18:10:16 INFO mapred.JobClient: map 100% reduce 100%
15/08/11 18:10:16 INFO mapred.JobClient: Job complete: job_local_0001
15/08/11 18:10:16 INFO mapred.JobClient: Counters: 14
15/08/11 18:10:16 INFO mapred.JobClient: FileSystemCounters
15/08/11 18:10:16 INFO mapred.JobClient: FILE_BYTES_READ=34932
15/08/11 18:10:16 INFO mapred.JobClient: HDFS_BYTES_READ=60
15/08/11 18:10:16 INFO mapred.JobClient: FILE_BYTES_WRITTEN=70694
15/08/11 18:10:16 INFO mapred.JobClient: Map-Reduce Framework
15/08/11 18:10:16 INFO mapred.JobClient: Reduce input groups=2
15/08/11 18:10:16 INFO mapred.JobClient: Combine output records=0
15/08/11 18:10:16 INFO mapred.JobClient: Map input records=2
15/08/11 18:10:16 INFO mapred.JobClient: Reduce shuffle bytes=0
15/08/11 18:10:16 INFO mapred.JobClient: Reduce output records=2
15/08/11 18:10:16 INFO mapred.JobClient: Spilled Records=4
15/08/11 18:10:16 INFO mapred.JobClient: Map output bytes=34
15/08/11 18:10:16 INFO mapred.JobClient: Map input bytes=30
15/08/11 18:10:16 INFO mapred.JobClient: Combine input records=0
15/08/11 18:10:16 INFO mapred.JobClient: Map output records=2
15/08/11 18:10:16 INFO mapred.JobClient: Reduce input records=2
測试之前我对原有表进行了清空处理，能够看到运行后数据库里边加入了两条内容；
下次在运行的时候会报错，属于正常情况，原因在于我们导入数据的时候对id进行赋值了，假设忽略id。是能够一直加入的；
源代码下载地址
源代码已上传，下载地址为download.csdn.net/detail/*nggui10000/8974585