如何将Lucene索引写入Hadoop1.x的HDFS系统,本篇散仙将介绍上将索引写在Hadoop2.x的HDFS上,写入2.x的Hadoop相对1.x的Hadoop来说要简单的说了,因为默认solr(4.4之后的版本)里面自带的HDFSDirectory就是支持2.x的而不支持1.x的,使用2.x的Hadoop平台,可以直接把solr的corejar包拷贝到工程里面,即可使用建索引,散仙,是在eclipse上使用eclipse插件来运行hadoop程序,具体要用到的jar包,除了需要用到hadoop2.2的所有jar包外,还需增加lucene和solr的部分jar包,截图如下,散仙本次使用的是Lucene4.8.1的版本:


如何将Lucene索引写入Hadoop1.x的HDFS系统
具体的代码如下:

Java代码 如何将Lucene索引写入Hadoop1.x的HDFS系统 如何将Lucene索引写入Hadoop1.x的HDFS系统如何将Lucene索引写入Hadoop1.x的HDFS系统
  1. package com.mapreduceindex;  
  2. import org.apache.hadoop.conf.Configuration;  
  3. import org.apache.hadoop.fs.FileSystem;  
  4. import org.apache.hadoop.fs.Path;  
  5. import org.apache.lucene.analysis.Analyzer;  
  6. import org.apache.lucene.analysis.standard.StandardAnalyzer;  
  7. import org.apache.lucene.document.Document;  
  8. import org.apache.lucene.document.Field.Store;  
  9. import org.apache.lucene.document.StringField;  
  10. import org.apache.lucene.document.TextField;  
  11. import org.apache.lucene.index.DirectoryReader;  
  12. import org.apache.lucene.index.IndexReader;  
  13. import org.apache.lucene.index.IndexWriter;  
  14. import org.apache.lucene.index.IndexWriterConfig;  
  15. import org.apache.lucene.index.Term;  
  16. import org.apache.lucene.queryparser.classic.QueryParser;  
  17. import org.apache.lucene.search.IndexSearcher;  
  18. import org.apache.lucene.search.Query;  
  19. import org.apache.lucene.search.ScoreDoc;  
  20. import org.apache.lucene.search.TopDocs;  
  21. import org.apache.lucene.store.Directory;  
  22.   
  23. import org.apache.lucene.util.Version;  
  24. import org.apache.solr.store.hdfs.HdfsDirectory;  
  25.    
  26. import org.wltea.analyzer.lucene.IKAnalyzer;  
  27.   
  28.   
  29.    
  30.   
  31. /** 
  32.  *  
  33.  * 将索引存储在Hadoop2.2的HDFS上 
  34.  * 
  35.  * @author qindongliang 
  36.  * QQ技术交流群: 
  37.  * 1号群: 324714439 如果满员了请加2号群 
  38.  * 2号群: 206247899 
  39.  *  
  40.  *  
  41.  * **/  
  42. public class MyIndex {  
  43.   
  44.     public static  void createFile()throws Exception{  
  45.           
  46.           
  47.         Configuration conf=new Configuration();  
  48.          FileSystem fs=FileSystem.get(conf);    
  49.          Path p =new Path("hdfs://192.168.46.32:9000/root/abc.txt");    
  50.          fs.createNewFile(p);    
  51.          //fs.create(p);    
  52.          fs.close();//释放资源    
  53.          System.out.println("创建文件成功.....");    
  54.           
  55.     }  
  56.       
  57.       
  58.     public static void main(String[] args)throws Exception {  
  59.         //createFile();  
  60.         //long a=System.currentTimeMillis();  
  61.        // add();  
  62.         // long b=System.currentTimeMillis();  
  63.         // System.out.println("耗时: "+(b-a)+"毫秒");  
  64.             query("8");  
  65.         //delete("3");//删除指定ID的数据  
  66.     }  
  67.       
  68.       
  69.       
  70.     /*** 
  71.      * 得到HDFS的writer 
  72.      *  
  73.      * **/  
  74.     public static IndexWriter  getIndexWriter() throws Exception{  
  75.           
  76.         Analyzer  analyzer=new IKAnalyzer(true);  
  77.         IndexWriterConfig    config=new IndexWriterConfig(Version.LUCENE_48, analyzer);  
  78.         Configuration conf=new Configuration();  
  79.        
  80.         conf.set("fs.defaultFS","hdfs://192.168.46.32:9000/");  
  81.         //conf.set("mapreduce.framework.name", "yarn");    
  82.         //conf.set("yarn.resourcemanager.address", "192.168.46.32:8032");   
  83.         //Path p1 =new Path("hdfs://10.2.143.5:9090/root/myfile/my.txt");  
  84.         //Path path=new Path("hdfs://10.2.143.5:9090/root/myfile");  
  85.         Path path=new Path("hdfs://192.168.46.32:9000/qin/myindex");  
  86.         //HdfsDirectory directory=new HdfsDirectory(path, conf);  
  87.           
  88.         HdfsDirectory directory=new HdfsDirectory(path, conf);  
  89.           
  90.         IndexWriter writer=new IndexWriter(directory, config);  
  91.           
  92.         return writer;  
  93.           
  94.     }  
  95.       
  96.       
  97.     public static void add()throws Exception{  
  98.             
  99.         IndexWriter writer=getIndexWriter();      
  100.            
  101. //      Document doc=new Document();  
  102. //      doc.add(new StringField("id", "3", Store.YES));  
  103. //      doc.add(new StringField("name", "lucene是一款非常优秀的全文检索框架", Store.YES));  
  104. //      doc.add(new TextField("content", "我们的工资都不高", Store.YES));  
  105. //      Document doc2=new Document();  
  106. //      doc2.add(new StringField("id", "4", Store.YES));  
  107. //      doc2.add(new StringField("name", "今天天气不错呀", Store.YES));  
  108. //      doc2.add(new TextField("content", "钱存储在银行靠谱吗", Store.YES));  
  109. //        
  110. //      Document doc3=new Document();  
  111. //      doc3.add(new StringField("id", "5", Store.YES));  
  112. //      doc3.add(new StringField("name", "没有根的野草,飘忽的命途!", Store.YES));  
  113. //      doc3.add(new TextField("content", "你工资多少呀!", Store.YES));  
  114. //       writer.addDocument(doc);  
  115. //       writer.addDocument(doc2);  
  116. //      writer.addDocument(doc3);  
  117.         for(int i=6;i<10000;i++){  
  118.             Document doc=new Document();  
  119.             doc.add(new StringField("id", i+"", Store.YES));  
  120.             doc.add(new StringField("name""lucene是一款非常优秀的全文检索框架"+i, Store.YES));  
  121.             doc.add(new TextField("content""今天发工资了吗"+i, Store.YES));  
  122.             writer.addDocument(doc);  
  123.             if(i%1000==0){  
  124.                 writer.commit();  
  125.             }  
  126.         }  
  127.         //writer.forceMerge(1);  
  128.         writer.commit();  
  129.         System.out.println("索引3条数据添加成功!");  
  130.          writer.close();  
  131.     }  
  132.       
  133.     /*** 
  134.      * 添加索引 
  135.      *  
  136.      * **/  
  137.     public static void add(Document d)throws Exception{  
  138.         
  139.         IndexWriter writer=getIndexWriter();      
  140.         writer.addDocument(d);  
  141.          writer.forceMerge(1);  
  142.         writer.commit();  
  143.         System.out.println("索引10000条数据添加成功!");  
  144.         writer.close();  
  145.     }  
  146.       
  147.     /** 
  148.      * 根据指定ID 
  149.      * 删除HDFS上的一些数据 
  150.      *  
  151.      *  
  152.      * **/  
  153.     public static void delete(String id)throws Exception{  
  154.           
  155.           
  156.         IndexWriter writer=getIndexWriter();  
  157.         writer.deleteDocuments(new Term("id", id));//删除指定ID的数据  
  158.         writer.forceMerge(1);//清除已经删除的索引空间  
  159.         writer.commit();//提交变化  
  160.           
  161.         System.out.println("id为"+id+"的数据已经删除成功.........");  
  162.           
  163.           
  164.     }  
  165.       
  166.       
  167.     public static void query(String queryTerm)throws Exception{  
  168.         System.out.println("本次检索内容:  "+queryTerm);  
  169.         Configuration conf=new Configuration();  
  170.         conf.set("fs.defaultFS","hdfs://192.168.46.32:9000/");  
  171.         //Path p1 =new Path("hdfs://10.2.143.5:9090/root/myfile/my.txt");  
  172.     //  Path path=new Path("hdfs://192.168.75.130:9000/root/index");  
  173.         Path path=new Path("hdfs://192.168.46.32:9000/qin/myindex");  
  174.         Directory directory=new HdfsDirectory(path, conf);  
  175.         IndexReader reader=DirectoryReader.open(directory);  
  176.         System.out.println("总数据量: "+reader.numDocs());  
  177.         long a=System.currentTimeMillis();  
  178.         IndexSearcher searcher=new IndexSearcher(reader);  
  179.         QueryParser parse=new QueryParser(Version.LUCENE_48, "content"new IKAnalyzer(true));  
  180.           
  181.          Query query=parse.parse(queryTerm);  
  182.           
  183.          TopDocs docs=searcher.search(query, 100);  
  184.            
  185.      System.out.println("本次命中结果:   "+docs.totalHits+"  条" );  
  186.          for(ScoreDoc sc:docs.scoreDocs){  
  187.              System.out.println("评分:  "+sc.score+"  id : "+searcher.doc(sc.doc).get("id")+"  name:   "+searcher.doc(sc.doc).get("name")+"   字段内容: "+searcher.doc(sc.doc).get("content"));  
  188.                
  189.          }  
  190.         long b=System.currentTimeMillis();  
  191.         System.out.println("第一次耗时:"+(b-a)+" 毫秒");  
  192.     //  System.out.println("============================================");  
  193. //      long c=System.currentTimeMillis();  
  194. //         query=parse.parse(queryTerm);  
  195. //            
  196. //         docs=searcher.search(query, 100);  
  197. //       System.out.println("本次命中结果:   "+docs.totalHits+"  条" );  
  198. //       for(ScoreDoc sc:docs.scoreDocs){  
  199. //             
  200. //           System.out.println("评分:  "+sc.score+"  id : "+searcher.doc(sc.doc).get("id")+"  name:   "+searcher.doc(sc.doc).get("name")+"   字段内容: "+searcher.doc(sc.doc).get("content"));  
  201. //             
  202. //       }  
  203. //      long d=System.currentTimeMillis();  
  204. //      System.out.println("第二次耗时:"+(d-c)+" 毫秒");  
  205.           
  206.          reader.close();  
  207.          directory.close();  
  208.            
  209.          System.out.println("检索完毕...............");  
  210.        
  211.           
  212.           
  213.           
  214.     }  
  215.       
  216.       
  217.       
  218.       
  219. }  
package com.mapreduceindex;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.lucene.analysis.Analyzer;
import org.apache.lucene.analysis.standard.StandardAnalyzer;
import org.apache.lucene.document.Document;
import org.apache.lucene.document.Field.Store;
import org.apache.lucene.document.StringField;
import org.apache.lucene.document.TextField;
import org.apache.lucene.index.DirectoryReader;
import org.apache.lucene.index.IndexReader;
import org.apache.lucene.index.IndexWriter;
import org.apache.lucene.index.IndexWriterConfig;
import org.apache.lucene.index.Term;
import org.apache.lucene.queryparser.classic.QueryParser;
import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.ScoreDoc;
import org.apache.lucene.search.TopDocs;
import org.apache.lucene.store.Directory;

import org.apache.lucene.util.Version;
import org.apache.solr.store.hdfs.HdfsDirectory;
 
import org.wltea.analyzer.lucene.IKAnalyzer;


 

/**
 * 
 * 将索引存储在Hadoop2.2的HDFS上
 *
 * @author qindongliang
 * QQ技术交流群:
 * 1号群: 324714439 如果满员了请加2号群
 * 2号群: 206247899
 * 
 * 
 * **/
public class MyIndex {

	public static  void createFile()throws Exception{
		
		
		Configuration conf=new Configuration();
		 FileSystem fs=FileSystem.get(conf);  
         Path p =new Path("hdfs://192.168.46.32:9000/root/abc.txt");  
         fs.createNewFile(p);  
         //fs.create(p);  
         fs.close();//释放资源  
         System.out.println("创建文件成功.....");  
		
	}
	
	
	public static void main(String[] args)throws Exception {
		//createFile();
		//long a=System.currentTimeMillis();
	   // add();
	 	// long b=System.currentTimeMillis();
	 	// System.out.println("耗时: "+(b-a)+"毫秒");
		    query("8");
		//delete("3");//删除指定ID的数据
	}
	
	
	
	/***
	 * 得到HDFS的writer
	 * 
	 * **/
	public static IndexWriter  getIndexWriter() throws Exception{
		
		Analyzer  analyzer=new IKAnalyzer(true);
 		IndexWriterConfig    config=new IndexWriterConfig(Version.LUCENE_48, analyzer);
 		Configuration conf=new Configuration();
 	 
 		conf.set("fs.defaultFS","hdfs://192.168.46.32:9000/");
		//conf.set("mapreduce.framework.name", "yarn");  
		//conf.set("yarn.resourcemanager.address", "192.168.46.32:8032"); 
 		//Path p1 =new Path("hdfs://10.2.143.5:9090/root/myfile/my.txt");
 		//Path path=new Path("hdfs://10.2.143.5:9090/root/myfile");
 		Path path=new Path("hdfs://192.168.46.32:9000/qin/myindex");
 		//HdfsDirectory directory=new HdfsDirectory(path, conf);
 		
 		HdfsDirectory directory=new HdfsDirectory(path, conf);
 		
 		IndexWriter writer=new IndexWriter(directory, config);
 		
 		return writer;
		
	}
	
	
	public static void add()throws Exception{
	      
		IndexWriter writer=getIndexWriter();	
		 
//		Document doc=new Document();
//		doc.add(new StringField("id", "3", Store.YES));
//		doc.add(new StringField("name", "lucene是一款非常优秀的全文检索框架", Store.YES));
//		doc.add(new TextField("content", "我们的工资都不高", Store.YES));
//		Document doc2=new Document();
//		doc2.add(new StringField("id", "4", Store.YES));
//		doc2.add(new StringField("name", "今天天气不错呀", Store.YES));
//		doc2.add(new TextField("content", "钱存储在银行靠谱吗", Store.YES));
//		
//		Document doc3=new Document();
//		doc3.add(new StringField("id", "5", Store.YES));
//		doc3.add(new StringField("name", "没有根的野草,飘忽的命途!", Store.YES));
//		doc3.add(new TextField("content", "你工资多少呀!", Store.YES));
// 		 writer.addDocument(doc);
//		 writer.addDocument(doc2);
//		writer.addDocument(doc3);
		for(int i=6;i<10000;i++){
			Document doc=new Document();
			doc.add(new StringField("id", i+"", Store.YES));
			doc.add(new StringField("name", "lucene是一款非常优秀的全文检索框架"+i, Store.YES));
			doc.add(new TextField("content", "今天发工资了吗"+i, Store.YES));
			writer.addDocument(doc);
			if(i%1000==0){
				writer.commit();
			}
		}
		//writer.forceMerge(1);
		writer.commit();
		System.out.println("索引3条数据添加成功!");
		 writer.close();
	}
	
	/***
	 * 添加索引
	 * 
	 * **/
	public static void add(Document d)throws Exception{
      
		IndexWriter writer=getIndexWriter();	
		writer.addDocument(d);
		 writer.forceMerge(1);
		writer.commit();
		System.out.println("索引10000条数据添加成功!");
		writer.close();
	}
	
	/**
	 * 根据指定ID
	 * 删除HDFS上的一些数据
	 * 
	 * 
	 * **/
	public static void delete(String id)throws Exception{
		
		
		IndexWriter writer=getIndexWriter();
		writer.deleteDocuments(new Term("id", id));//删除指定ID的数据
		writer.forceMerge(1);//清除已经删除的索引空间
		writer.commit();//提交变化
		
		System.out.println("id为"+id+"的数据已经删除成功.........");
		
		
	}
	
	
	public static void query(String queryTerm)throws Exception{
		System.out.println("本次检索内容:  "+queryTerm);
		Configuration conf=new Configuration();
		conf.set("fs.defaultFS","hdfs://192.168.46.32:9000/");
 		//Path p1 =new Path("hdfs://10.2.143.5:9090/root/myfile/my.txt");
 	//	Path path=new Path("hdfs://192.168.75.130:9000/root/index");
 		Path path=new Path("hdfs://192.168.46.32:9000/qin/myindex");
		Directory directory=new HdfsDirectory(path, conf);
		IndexReader reader=DirectoryReader.open(directory);
		System.out.println("总数据量: "+reader.numDocs());
		long a=System.currentTimeMillis();
		IndexSearcher searcher=new IndexSearcher(reader);
		QueryParser parse=new QueryParser(Version.LUCENE_48, "content", new IKAnalyzer(true));
		
		 Query query=parse.parse(queryTerm);
		
		 TopDocs docs=searcher.search(query, 100);
		 
 	 System.out.println("本次命中结果:   "+docs.totalHits+"  条" );
		 for(ScoreDoc sc:docs.scoreDocs){
			 System.out.println("评分:  "+sc.score+"  id : "+searcher.doc(sc.doc).get("id")+"  name:   "+searcher.doc(sc.doc).get("name")+"   字段内容: "+searcher.doc(sc.doc).get("content"));
			 
		 }
		long b=System.currentTimeMillis();
		System.out.println("第一次耗时:"+(b-a)+" 毫秒");
	//	System.out.println("============================================");
//		long c=System.currentTimeMillis();
//		   query=parse.parse(queryTerm);
//			
//		   docs=searcher.search(query, 100);
//		 System.out.println("本次命中结果:   "+docs.totalHits+"  条" );
//		 for(ScoreDoc sc:docs.scoreDocs){
//			 
//			 System.out.println("评分:  "+sc.score+"  id : "+searcher.doc(sc.doc).get("id")+"  name:   "+searcher.doc(sc.doc).get("name")+"   字段内容: "+searcher.doc(sc.doc).get("content"));
//			 
//		 }
//		long d=System.currentTimeMillis();
//		System.out.println("第二次耗时:"+(d-c)+" 毫秒");
		
		 reader.close();
		 directory.close();
		 
		 System.out.println("检索完毕...............");
	 
		
		
		
	}
	
	
	
	
}


使用IK的分词器,建立索引完毕后,在HDFS上的索引如下截图:

如何将Lucene索引写入Hadoop1.x的HDFS系统
检索数据时,第一次检索往往比较慢,第一次之后因为有了Block Cache,所以第二次,检索的速度非常快,当然这也跟你机器的配置有关系:

Java代码 如何将Lucene索引写入Hadoop1.x的HDFS系统 如何将Lucene索引写入Hadoop1.x的HDFS系统如何将Lucene索引写入Hadoop1.x的HDFS系统
  1. 本次检索内容:  8  
  2. WARN - NativeCodeLoader.<clinit>(62) | Unable to load native-hadoop library for your platform... using builtin-java classes where applicable  
  3. 总数据量: 9994  
  4. 本次命中结果:   1  条  
  5. 评分:  4.7582965  id : 8  name:   lucene是一款非常优秀的全文检索框架8   字段内容: 今天发工资了吗8  
  6. 第一次耗时:261 毫秒  
  7. ============================================  
  8. 本次命中结果:   1  条  
  9. 评分:  4.7582965  id : 8  name:   lucene是一款非常优秀的全文检索框架8   字段内容: 今天发工资了吗8  
  10. 第二次耗时:6 毫秒  
  11. INFO - HdfsDirectory.close(97) | Closing hdfs directory hdfs://192.168.46.32:9000/qin/myindex  
  12. 检索完毕...............  
本次检索内容:  8
WARN - NativeCodeLoader.<clinit>(62) | Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
总数据量: 9994
本次命中结果:   1  条
评分:  4.7582965  id : 8  name:   lucene是一款非常优秀的全文检索框架8   字段内容: 今天发工资了吗8
第一次耗时:261 毫秒
============================================
本次命中结果:   1  条
评分:  4.7582965  id : 8  name:   lucene是一款非常优秀的全文检索框架8   字段内容: 今天发工资了吗8
第二次耗时:6 毫秒
INFO - HdfsDirectory.close(97) | Closing hdfs directory hdfs://192.168.46.32:9000/qin/myindex
检索完毕...............




为什么要使用Hadoop建索引? 使用Hadoop建索引可以利用MapReduce分布式计算能力从而大大提升建索引的速度,这一点优势很明显,但美中不足的是在Hadoop上做检索,性能却不怎么好,虽然有了块缓存,但是如果索引被按64M的块被切分到不同的节点上,那么检索的时候,就需要跨机器从各个块上扫描,拉取命中数据,这一点是很耗时的,目前,据散仙所知,还没有比较好的部署在Hadoop上的分布式检索方案,但毫无疑问的是建索引的能力,确实很给力,后面散仙会写如何使用MapReduce来并行构建Lucene索引,其实既然单机版的都可以完成,那么稍微改造下变成MapReduce作业

相关文章: