Go语言视频零基础入门到精通

admin

需要将生产环境上Infoxmix里的数据原封不动的Copy到另一台 Oracle数据库服务器上，然后对Copy后的数据作些漂白处理。为了将人为干预的因素降到最低，在系统设计时采用java代码对数据作Copy,思路 如图：



    首 先在代码与生产库间建立一个Connection，将读取到的数据放在ResultSet对象，然后再与开发库建立一个Connection。从 ResultSet取出数据后通过TestConnection插入到开发库，以此来实现Copy。代码写完后运行程序，速度太慢了，一秒钟只能Copy 一千条数据，生产库上有上亿条数据，按照这个速度同步完要到猴年马月呀,用PreparedStatement批处理速度也没有提交多少。我想能不能用多 线程处理，多个人干活总比一个人干活速度要快。
    假设生产库有1万条数据，我开5个线程，每个线程分2000条数据，同时向开发库里插数据，Oracle支持高并发这样的话速度至少会提高好多倍，按照这 个思路重新进行了编码，批处理设置为1万条一提交，统计插入数量的变量使用 java.util.concurrent.atomic.AtomicLong，程序一运行，传输速度飞快CPU利用率在70%~90%，现在一秒钟可 以拷贝50万条记录，没过几分钟上亿条数据一条不落地全部Copy到目标库。

在查询的时候我用了如下语句
String queryStr = "SELECT * FROM xx";ResultSet coreRs = PreparedStatement.executeQuery(queryStr);
实习生问如果xx表里有上千万条记录，你全部查询出来放到ResultSet, 那内存不溢出了么？Java在设计的时候已经考虑到这个问题了，并没有查询出所有的数据，而是只查询了一部分数据放到ResultSet,数据“用完”它 会自动查询下一批数据，你可以用setFetchSize(int rows)方法设置一个建议值给ResultSet,告诉它每次从数据库Fetch多少条数据。但我不赞成，因为JDBC驱动会根据实际情况自动调整 Fetch的数量。另外性能也与网线的带宽有直接的关系。
相关代码

1. package com.dlbank.domain;import java.sql.Connection;import java.sql.PreparedStatement;import java.sql.ResultSet;import java.sql.Statement;import java.util.List;import java.util.concurrent.atomic.AtomicLong;import org.apache.log4j.Logger;/***<p>title: 数据同步类 </p>*<p>Description: 该类用于将生产核心库数据同步到开发库</p>*@author Tank Zhang*/public class CoreDataSyncImpl implements CoreDataSync {private List<String> coreTBNames; //要同步的核心库表名private ConnectionFactory connectionFactory;private Logger log = Logger.getLogger(getClass());private AtomicLong currentSynCount = new AtomicLong(0L); //当前已同步的条数private int syncThreadNum;  //同步的线程数@Overridepublic void syncData(int businessType) throws Exception {for (String tmpTBName : coreTBNames) {log.info("开始同步核心库" + tmpTBName + "表数据");// 获得核心库连接Connection coreConnection = connectionFactory.getDMSConnection(4);Statement coreStmt = coreConnection.createStatement();//为每个线程分配结果集ResultSet coreRs = coreStmt.executeQuery("SELECT count(*) FROM "+tmpTBName);coreRs.next();//总共处理的数量long totalNum = coreRs.getLong(1);//每个线程处理的数量long ownerRecordNum =(long) Math.ceil((totalNum / syncThreadNum));log.info("共需要同步的数据量："+totalNum);log.info("同步线程数量："+syncThreadNum);log.info("每个线程可处理的数量："+ownerRecordNum);// 开启五个线程向目标库同步数据for(int i=0; i < syncThreadNum; i ++){StringBuilder sqlBuilder = new StringBuilder();//拼装后SQL示例//Select * From dms_core_ds Where id between 1 And 657398//Select * From dms_core_ds Where id between 657399 And 1314796//Select * From dms_core_ds Where id between 1314797 And 1972194//Select * From dms_core_ds Where id between 1972195 And 2629592//Select * From dms_core_ds Where id between 2629593 And 3286990//..sqlBuilder.append("Select * From ").append(tmpTBName).append(" Where id between " ).append(i * ownerRecordNum +1).append( " And ").append((i * ownerRecordNum + ownerRecordNum));Thread workThread = new Thread(new WorkerHandler(sqlBuilder.toString(),businessType,tmpTBName));workThread.setName("SyncThread-"+i);workThread.start();}while (currentSynCount.get() < totalNum);//休眠一会儿让数据库有机会commit剩余的批处理(只针对JUnit单元测试,//因为单元测试完成后会关闭虚拟器，使线程里的代码没有机会作提交操作);//Thread.sleep(1000 * 3);log.info( "核心库"+tmpTBName+"表数据同步完成，共同步了" + currentSynCount.get() + "条数据");}}// end for looppublic void setCoreTBNames(List<String> coreTBNames) {this.coreTBNames = coreTBNames;}public void setConnectionFactory(ConnectionFactory connectionFactory) {this.connectionFactory = connectionFactory;}public void setSyncThreadNum(int syncThreadNum) {this.syncThreadNum = syncThreadNum;}//数据同步线程final class WorkerHandler implements Runnable {ResultSet coreRs;String queryStr;int businessType;String targetTBName;public WorkerHandler(String queryStr,int businessType,String targetTBName) {this.queryStr = queryStr;this.businessType = businessType;this.targetTBName = targetTBName;}@Overridepublic void run() {try {//开始同步launchSyncData();} catch(Exception e){log.error(e);e.printStackTrace();}}//同步数据方法void launchSyncData() throws Exception{// 获得核心库连接Connection coreConnection = connectionFactory.getDMSConnection(4);Statement coreStmt = coreConnection.createStatement();// 获得目标库连接Connection targetConn = connectionFactory.getDMSConnection(businessType);targetConn.setAutoCommit(false);// 设置手动提交PreparedStatement targetPstmt =targetConn.prepareStatement("INSERT INTO " + targetTBName+" VALUES (?,?,?,?,?)");ResultSet coreRs = coreStmt.executeQuery(queryStr);log.info(Thread.currentThread().getName()+"'s Query SQL::"+queryStr);int batchCounter = 0; //累加的批处理数量while (coreRs.next()) {targetPstmt.setString(1, coreRs.getString(2));targetPstmt.setString(2, coreRs.getString(3));targetPstmt.setString(3, coreRs.getString(4));targetPstmt.setString(4, coreRs.getString(5));targetPstmt.setString(5, coreRs.getString(6));targetPstmt.addBatch();batchCounter++;currentSynCount.incrementAndGet();//递增if (batchCounter % 10000 == 0) { //1万条数据一提交targetPstmt.executeBatch();targetPstmt.clearBatch();targetConn.commit();}}//提交剩余的批处理targetPstmt.executeBatch();targetPstmt.clearBatch();targetConn.commit();//释放连接connectionFactory.release(targetConn, targetPstmt,coreRs);}}}   
2.   

复制代码