通用MapReduce程序复制HBase表数据

  编写MR程序，让其可以适合大部分的HBase表数据导入到HBase表数据。其中包括可以设置版本数、可以设置输入表的列导入设置(选取其中某几列)、可以设置输出表的列导出设置(选取其中某几列)。

  原始表test1数据如下：

 每个row key都有两个版本的数据，这里只显示了row key为1的数据

 在hbase shell 中创建数据表：

create 'test2',{NAME => 'cf1',VERSIONS => 10}    // 保存无版本、无列导入设置、无列导出设置的数据create 'test3',{NAME => 'cf1',VERSIONS => 10}    // 保存无版本、无列导入设置、有列导出设置的数据create 'test4',{NAME => 'cf1',VERSIONS => 10}    // 保存无版本、有列导入设置、无列导出设置的数据create 'test5',{NAME => 'cf1',VERSIONS => 10}    // 保存有版本、无列导入设置、无列导出设置的数据create 'test6',{NAME => 'cf1',VERSIONS => 10}    // 保存有版本、无列导入设置、有列导出设置的数据create 'test7',{NAME => 'cf1',VERSIONS => 10}    // 保存有版本、有列导入设置、无列导出设置的数据create 'test8',{NAME => 'cf1',VERSIONS => 10}    // 保存有版本、有列导入设置、有列导出设置的数据

 main函数入口：

package GeneralHBaseToHBase;import org.apache.hadoop.util.ToolRunner;public class DriverTest {public static void main(String[] args) throws Exception {// 无版本设置、无列导入设置，无列导出设置String[] myArgs1= new String[]{"test1", // 输入表"test2", // 输出表"0",     // 版本大小数，如果值为0，则为默认从输入表导出最新的数据到输出表"-1",    // 列导入设置，如果为-1 ，则没有设置列导入"-1"// 列导出设置，如果为-1，则没有设置列导出};ToolRunner.run(HBaseDriver.getConfiguration(), new HBaseDriver(),myArgs1);// 无版本设置、有列导入设置，无列导出设置String[] myArgs2= new String[]{"test1","test3","0","cf1:c1,cf1:c2,cf1:c10,cf1:c11,cf1:c14","-1"};ToolRunner.run(HBaseDriver.getConfiguration(), new HBaseDriver(),myArgs2);// 无版本设置，无列导入设置，有列导出设置String[] myArgs3= new String[]{"test1","test4","0","-1","cf1:c1,cf1:c10,cf1:c14"};ToolRunner.run(HBaseDriver.getConfiguration(), new HBaseDriver(),myArgs3);// 有版本设置，无列导入设置，无列导出设置String[] myArgs4= new String[]{"test1","test5","2","-1","-1"};ToolRunner.run(HBaseDriver.getConfiguration(), new HBaseDriver(),myArgs4);// 有版本设置、有列导入设置，无列导出设置String[] myArgs5= new String[]{"test1","test6","2","cf1:c1,cf1:c2,cf1:c10,cf1:c11,cf1:c14","-1"};ToolRunner.run(HBaseDriver.getConfiguration(), new HBaseDriver(),myArgs5);// 有版本设置、无列导入设置，有列导出设置String[] myArgs6= new String[]{"test1","test7","2","-1","cf1:c1,cf1:c10,cf1:c14"};ToolRunner.run(HBaseDriver.getConfiguration(), new HBaseDriver(),myArgs6);// 有版本设置、有列导入设置，有列导出设置String[] myArgs7= new String[]{"test1","test8","2","cf1:c1,cf1:c2,cf1:c10,cf1:c11,cf1:c14","cf1:c1,cf1:c10,cf1:c14"};ToolRunner.run(HBaseDriver.getConfiguration(), new HBaseDriver(),myArgs7);}}

  driver：

package GeneralHBaseToHBase;import org.apache.hadoop.conf.Configuration;import org.apache.hadoop.conf.Configured;import org.apache.hadoop.hbase.client.Put;import org.apache.hadoop.hbase.client.Scan;import org.apache.hadoop.hbase.io.ImmutableBytesWritable;import org.apache.hadoop.hbase.mapreduce.TableMapReduceUtil;import org.apache.hadoop.mapreduce.Job;import org.apache.hadoop.util.Tool;import util.JarUtil;public class HBaseDriver extends Configured implements Tool{public static String FROMTABLE=""; //导入表public static String TOTABLE="";  //导出表public static String SETVERSION=""; //是否设置版本// args => {FromTable,ToTable,SetVersion,ColumnFromTable,ColumnToTable}@Overridepublic int run(String[] args) throws Exception {if(args.length!=5){System.err.println("Usage:\n demo.job.HBaseDriver <input> <inputTable> "+ "<output> <outputTable>"+"< versions >"+ " <set columns from inputTable> like <cf1:c1,cf1:c2,cf1:c10,cf1:c11,cf1:c14> or <-1> "+ "<set columns from outputTable> like <cf1:c1,cf1:c10,cf1:c14> or <-1>");return -1;}Configuration conf = getConf();FROMTABLE = args[0];TOTABLE = args[1];SETVERSION = args[2];conf.set("SETVERSION", SETVERSION);if(!args[3].equals("-1")){conf.set("COLUMNFROMTABLE", args[3]);}if(!args[4].equals("-1")){conf.set("COLUMNTOTABLE", args[4]);}String jobName ="From table "+FROMTABLE+ " ,Import to "+ TOTABLE;Job job = Job.getInstance(conf, jobName);job.setJarByClass(HBaseDriver.class);Scan scan = new Scan();// 判断是否需要设置版本if(SETVERSION != "0" || SETVERSION != "1"){scan.setMaxVersions(Integer.parseInt(SETVERSION));}// 设置HBase表输入：表名、scan、Mapper类、mapper输出键类型、mapper输出值类型TableMapReduceUtil.initTableMapperJob(FROMTABLE, scan, HBaseToHBaseMapper.class, ImmutableBytesWritable.class, Put.class, job);// 设置HBase表输出：表名，reducer类TableMapReduceUtil.initTableReducerJob(TOTABLE, null, job);// 没有 reducers，  直接写入到 输出文件    job.setNumReduceTasks(0);         return job.waitForCompletion(true) ? 0 : 1;        }private static Configuration configuration;public static Configuration getConfiguration(){if(configuration==null){/** * TODO 了解如何直接从Windows提交代码到Hadoop集群 *      并修改其中的配置为实际配置 */configuration = new Configuration();configuration.setBoolean("mapreduce.app-submission.cross-platform", true);// 配置使用跨平台提交任务configuration.set("fs.defaultFS", "hdfs://master:8020");// 指定namenodeconfiguration.set("mapreduce.framework.name", "yarn"); // 指定使用yarn框架configuration.set("yarn.resourcemanager.address", "master:8032"); // 指定resourcemanagerconfiguration.set("yarn.resourcemanager.scheduler.address", "master:8030");// 指定资源分配器configuration.set("mapreduce.jobhistory.address", "master:10020");// 指定historyserverconfiguration.set("hbase.master", "master:16000");configuration.set("hbase.rootdir", "hdfs://master:8020/hbase");configuration.set("hbase.zookeeper.quorum", "slave1,slave2,slave3");configuration.set("hbase.zookeeper.property.clientPort", "2181");//TODO 需export->jar file ; 设置正确的jar包所在位置configuration.set("mapreduce.job.jar",JarUtil.jar(HBaseDriver.class));// 设置jar包路径}return configuration;}}

  mapper：

package GeneralHBaseToHBase;import java.io.IOException;import java.util.ArrayList;import java.util.HashMap;import java.util.HashSet;import java.util.Map.Entry;import java.util.NavigableMap;import org.apache.hadoop.conf.Configuration;import org.apache.hadoop.hbase.Cell;import org.apache.hadoop.hbase.KeyValue;import org.apache.hadoop.hbase.client.Put;import org.apache.hadoop.hbase.client.Result;import org.apache.hadoop.hbase.io.ImmutableBytesWritable;import org.apache.hadoop.hbase.mapreduce.TableMapper;import org.apache.hadoop.hbase.util.Bytes;import org.slf4j.Logger;import org.slf4j.LoggerFactory;public class HBaseToHBaseMapper extends TableMapper<ImmutableBytesWritable, Put> {Logger log = LoggerFactory.getLogger(HBaseToHBaseMapper.class);private static int versionNum = 0;private static String[] columnFromTable = null;private static String[] columnToTable = null;private static String column1 = null;private static String column2 = null;@Overrideprotected void setup(Context context)throws IOException, InterruptedException {Configuration conf = context.getConfiguration();versionNum = Integer.parseInt(conf.get("SETVERSION", "0"));column1 = conf.get("COLUMNFROMTABLE",null);if(!(column1 == null)){columnFromTable = column1.split(",");}column2 = conf.get("COLUMNTOTABLE",null); if(!(column2 == null)){columnToTable = column2.split(",");}}@Overrideprotected void map(ImmutableBytesWritable key, Result value,Context context)throws IOException, InterruptedException {context.write(key, resultToPut(key,value));}/*** * 把key，value转换为Put * @param key * @param value * @return * @throws IOException */private Put resultToPut(ImmutableBytesWritable key, Result value) throws IOException {HashMap<String, String> fTableMap = new HashMap<>();HashMap<String, String> tTableMap = new HashMap<>();Put put = new Put(key.get());if(! (columnFromTable == null || columnFromTable.length == 0)){fTableMap = getFamilyAndColumn(columnFromTable);}if(! (columnToTable == null || columnToTable.length == 0)){tTableMap = getFamilyAndColumn(columnToTable);}if(versionNum==0){                        if(fTableMap.size() == 0){            if(tTableMap.size() == 0){  for (Cell kv : value.rawCells()) {put.add(kv);// 没有设置版本，没有设置列导入，没有设置列导出}return put;} else{return getPut(put, value, tTableMap); // 无版本、无列导入、有列导出}} else {if(tTableMap.size() == 0){return getPut(put, value, fTableMap);// 无版本、有列导入、无列导出} else {return getPut(put, value, tTableMap);// 无版本、有列导入、有列导出}}} else{if(fTableMap.size() == 0){if(tTableMap.size() == 0){return getPut1(put, value); // 有版本，无列导入，无列导出}else{return getPut2(put, value, tTableMap); //有版本，无列导入，有列导出}}else{if(tTableMap.size() == 0){return getPut2(put,value,fTableMap);// 有版本，有列导入，无列导出}else{return getPut2(put,value,tTableMap); // 有版本，有列导入，有列导出}}}}/*** * 无版本设置的情况下，对于有列导入或者列导出 * @param put * @param value * @param tableMap * @return * @throws IOException */private Put getPut(Put put,Result value,HashMap<String, String> tableMap) throws IOException{for(Cell kv : value.rawCells()){byte[] family = kv.getFamily();if(tableMap.containsKey(new String(family))){String columnStr = tableMap.get(new String(family));ArrayList<String> columnBy = toByte(columnStr);if(columnBy.contains(new String(kv.getQualifier()))){put.add(kv); //没有设置版本，没有设置列导入，有设置列导出}}}return put;}/*** * (有版本，无列导入，有列导出)或者(有版本，有列导入，无列导出) * @param put * @param value * @param tTableMap * @return */private Put getPut2(Put put,Result value,HashMap<String, String> tableMap){NavigableMap<byte[], NavigableMap<byte[], NavigableMap<Long, byte[]>>> map=value.getMap();        for(byte[] family:map.keySet()){        if(tableMap.containsKey(new String(family))){        String columnStr = tableMap.get(new String(family));        log.info("@@@@@@@@@@@"+new String(family)+"   "+columnStr);ArrayList<String> columnBy = toByte(columnStr);        NavigableMap<byte[], NavigableMap<Long, byte[]>> familyMap = map.get(family);//列簇作为key获取其中的列相关数据                for(byte[] column:familyMap.keySet()){                              //根据列名循坏                    log.info("!!!!!!!!!!!"+new String(column));                if(columnBy.contains(new String(column))){                NavigableMap<Long, byte[]> valuesMap = familyMap.get(column);                    for(Entry<Long, byte[]> s:valuesMap.entrySet()){//获取列对应的不同版本数据，默认最新的一个                    System.out.println("***:"+new String(family)+"  "+new String(column)+"  "+s.getKey()+"  "+new String(s.getValue()));                    put.addColumn(family, column, s.getKey(),s.getValue());                    }                    }                }        }                    }return put;}/*** * 有版本、无列导入、无列导出 * @param put * @param value * @return */private Put getPut1(Put put,Result value){NavigableMap<byte[], NavigableMap<byte[], NavigableMap<Long, byte[]>>> map=value.getMap();        for(byte[] family:map.keySet()){                NavigableMap<byte[], NavigableMap<Long, byte[]>> familyMap = map.get(family);//列簇作为key获取其中的列相关数据            for(byte[] column:familyMap.keySet()){                              //根据列名循坏                NavigableMap<Long, byte[]> valuesMap = familyMap.get(column);                for(Entry<Long, byte[]> s:valuesMap.entrySet()){                //获取列对应的不同版本数据，默认最新的一个                put.addColumn(family, column, s.getKey(),s.getValue());                }            }        }        return put;}// str => {"cf1:c1","cf1:c2","cf1:c10","cf1:c11","cf1:c14"}/*** * 得到列簇名与列名的k,v形式的map * @param str => {"cf1:c1","cf1:c2","cf1:c10","cf1:c11","cf1:c14"} * @return map => {"cf1" => "c1,c2,c10,c11,c14"} */private static HashMap<String, String> getFamilyAndColumn(String[] str){HashMap<String, String> map = new HashMap<>();HashSet<String> set = new HashSet<>();for(String s : str){set.add(s.split(":")[0]);}Object[] ob = set.toArray();for(int i=0; i<ob.length;i++){String family = String.valueOf(ob[i]);String columns = "";for(int j=0;j < str.length;j++){if(family.equals(str[j].split(":")[0])){columns += str[j].split(":")[1]+",";}}map.put(family, columns.substring(0, columns.length()-1));}return map;}private static ArrayList<String> toByte(String s){ArrayList<String> b = new ArrayList<>();String[] sarr = s.split(",");for(int i=0;i<sarr.length;i++){b.add(sarr[i]);}return b;}}

  程序运行完之后，在hbase shell中查看每个表，看是否数据导入正确：

  test2：(无版本、无列导入设置、无列导出设置)

test3 (无版本、有列导入设置("cf1:c1,cf1:c2,cf1:c10,cf1:c11,cf1:c14")、无列导出设置)

 test4(无版本、无列导入设置、有列导出设置("cf1:c1,cf1:c10,cf1:c14"))

 

 test5(有版本、无列导入设置、无列导出设置)

 

 

 test6(有版本、有列导入设置("cf1:c1,cf1:c2,cf1:c10,cf1:c11,cf1:c14")、无列导出设置)

 

  test7(有版本、无列导入设置、有列导出设置("cf1:c1,cf1:c10,cf1:c14"))

test8(有版本、有列导入设置("cf1:c1,cf1:c2,cf1:c10,cf1:c11,cf1:c14")、有列导出设置("cf1:c1,cf1:c10,cf1:c14"))