zoukankan      html  css  js  c++  java
  • 多线程处理慢sql查询小笔记~

    多线程处理慢sql查询以及List(Array)的拆分

    系统数据量不大,但是访问速度特别慢,使用多线程优化一下!!!

    优化结果:访问时间缩短了十几秒  25s --> 8s

    一、List的拆分:Iterables.partition

    注意: 引入的包为google名下的

        Iterables.partitions(List list , int size);  // list表示需要被拆分的集合,size表示拆分后的集合的元素个数

       例如:size=3,表示会拆分为2个集合,第一个集合中的元素为{1,2,3},第二个集合中的元素为{4,5}

          size=2,表示会拆分为3个集合,分别为{1,2} {3,4} {5}  以此类推

    ps: 集合的合并:

    List list = new ArrayList();

    List list1 = new ArrayList();

     list.addAll(list1);

    import com.google.common.collect.Iterables;  

    List<String> strList = new ArrayList<String>();
    strList.add(1);
    strList.add(2);
    strList.add(3);
    strList.add(4);
    strList.add(5);
    Iterable<List<String>> splitStringlList = Iterables.partition(strList, 3); //3表示拆分后的集合元素的数量

    二、Array的拆分:Arrays.asList(ArrayObj)

    先将数组转换为List,然后再按照第一方法去拆分。

    String[] validRiderArray = [1,2,3,4,5];
    ArrayList<String> validArrayList = new ArrayList<String>(Arrays.asList(validRiderArray));

    三、多线程处理类:

    Runtime.getRuntime().availableProcessors();//返回到Java虚拟机的可用的处理器数量
    package com.chinalife.proposal.common.util;
    
    import java.util.concurrent.Callable;
    import java.util.concurrent.ExecutorService;
    import java.util.concurrent.Executors;
    import java.util.concurrent.Future;
    
    import org.springframework.stereotype.Component;
    
    import com.google.common.util.concurrent.ThreadFactoryBuilder;
    /**
     * 固定线程个数的线程池工具类
     * */
     
    public class SimpleFixedThreadPoolUtil {
        /** 默认线程个数 */
        private static final int DEFAULT_THREAD = Runtime.getRuntime().availableProcessors();//返回到Java虚拟机的可用的处理器数量
        /** 线程池个数: 默认为: Runtime.getRuntime().availableProcessors() * 2 */
        public static int THREADPOOL_NUM = DEFAULT_THREAD * 2;
        private static ExecutorService executor = Executors.newFixedThreadPool(THREADPOOL_NUM,
                new ThreadFactoryBuilder().setNameFormat("SimpleFixedThreadPool- %d: ").build());
    
        /**
         * 提交任务,任务执行后无返回值
         * 
         * @param task
         */
        public static void addTask(Runnable task) {
            executor.execute(task);
        }
    
        /**
         * 提交具有返回值的任务
         * @param task
         * @return
         */
        public static <T> Future<T> addTask(Callable<T> task) {
            Future<T> future = executor.submit(task);
            return future;
        }
    
        /**
         * 关闭线程池
         */
        /*public void shutdown() {
            executor.shutdown();
        }*/
    }

    四、多线程处理慢查询使用示例:

    JunitTest:

    /**
         * 多线程处理慢sql查询示例
         * @throws Exception
         */
        @Test
        public void testAddTaskCallables() throws Exception {
            String role = "Agent";
            String nodeid = "2";
            String subnodeid = "3";
            long time = System.currentTimeMillis();
            List<OnlProposalRolePrdRelation> rolePrdlist = webProposalService.getAvailableProductListByRoleAndID(role,nodeid, subnodeid);
            Iterable<List<OnlProposalRolePrdRelation>> result = Iterables.partition(rolePrdlist, 30);
            Iterator<List<OnlProposalRolePrdRelation>> iterator = result.iterator();
            List<Future<List<OnlProposalRolePrdRelation>>> futures = new ArrayList<>();
            while (iterator.hasNext()) {
                List<OnlProposalRolePrdRelation> listTemp = iterator.next();
                Future<List<OnlProposalRolePrdRelation>> future = SimpleFixedThreadPoolUtil
                        .addTask((Callable<List<OnlProposalRolePrdRelation>>) () -> {
                            List<OnlProposalRolePrdRelation> relations = null;
                            try {
                                relations = webProposalService.filterByCriteria(listTemp, 20, "1", "1", "0");
                                return relations;
                            } catch (Exception e) {
                                e.printStackTrace();
                            }
                            return relations;
                        });
                futures.add(future);
            }
            List<OnlProposalRolePrdRelation> proposalResultList = new ArrayList<OnlProposalRolePrdRelation>();
            for (Future<List<OnlProposalRolePrdRelation>> future : futures) {
                while (!future.isDone()) {
                    Thread.sleep(100);
                }
                List<OnlProposalRolePrdRelation> relations = future.get();
                System.out.println(relations.size());
                proposalResultList.addAll(relations);
    
            }
            logger.debug("time 3 =" + (System.currentTimeMillis() - time) / 1000);
            System.out.println("stop");
        }

     实际应用:

            ArrayList<String> validArrayList = new ArrayList<String>(Arrays.asList(validRiderArray)); 
            
            //step1 method start
            //List<TblProposalRiderAttr> tblProposalRiderAttrList = getValidRiderListStep1(validRiderArray, parameter,tblProposalProductSetting, tPresentrelationForRider );
            Iterable<List<String>> splitValidArrayList = Iterables.partition(validArrayList, 15);
            Iterator<List<String>> iterator = splitValidArrayList.iterator();
            List<Future<List<TblProposalRiderAttr>>> tblRiderFutures = new ArrayList<Future<List<TblProposalRiderAttr>>>();
            while (iterator.hasNext()) {
                List<String> tmpValidArrayList = iterator.next();
                //调用多线程执行  future为单个list的返回结果; tblRiderFutures为所有返回结果的集合
                Future<List<TblProposalRiderAttr>> future = SimpleFixedThreadPoolUtil
                        .addTask((Callable<List<TblProposalRiderAttr>>) () -> {
                            List<TblProposalRiderAttr> tmpTblRiderList = null;
                            try {
                                //调用step1方法进行筛选附加产品信息  进行多线程处理的方法
                                tmpTblRiderList = getValidRiderListStep1(tmpValidArrayList, parameter,tblProposalProductSetting,tPresentrelationForRider);
                                return tmpTblRiderList;
                            } catch (Exception e) {
                                e.printStackTrace();
                            }
                            return tmpTblRiderList;
                        });
                tblRiderFutures.add(future);
            }
            //将筛选后的结果添加到集合中
            List<TblProposalRiderAttr> tblProposalRiderAttrList = new ArrayList<TblProposalRiderAttr>();
            for (Future<List<TblProposalRiderAttr>> future : tblRiderFutures) {
                while (!future.isDone()) {
                    try {
                        Thread.sleep(100);
                    } catch (InterruptedException e) {
                        e.printStackTrace();
                    }
                }
                try {
                    tblProposalRiderAttrList.addAll(future.get());
                } catch (InterruptedException e) {
                    e.printStackTrace();
                } catch (ExecutionException e) {
                    e.printStackTrace();
                }
            }
            //step1 method  end

    参考链接:

    https://www.cnblogs.com/dolphin0520/p/3949310.html

    https://blog.csdn.net/qq1137623160/article/details/79772505

  • 相关阅读:
    从Hello World说起(Dart)到“几乎所有东西都是Widget”小部件。
    C#开发者工具网
    sqlitestudio
    jstat命令 Java Virtual Machine Statistics Monitoring Tool
    ProxyPass与ProxyPassReverse及ProxyPassMatch的概述
    IBM MQ介绍
    sun misc unsafe类的使用
    Android Webview实现文件下载功能
    使用OpenSSL生成私钥 Private Key 以及根据Private Key创建证书
    阿里巴巴的面试
  • 原文地址:https://www.cnblogs.com/DFX339/p/9884930.html
Copyright © 2011-2022 走看看