并发list性能对比

并发list性能对比

代码演示：

package com.dwz.concurrent;

import java.util.ArrayList;
import java.util.Collection;
import java.util.Collections;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.concurrent.ConcurrentLinkedQueue;
import java.util.concurrent.CopyOnWriteArrayList;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.atomic.AtomicInteger;

public class ConcurrentListPerformanceTest {
    static class Entry{
        int threshold;//线程数
        long ms;//运行的平均时间
        public Entry(int threshold, long ms) {
            this.threshold = threshold;
            this.ms = ms;
        }
        
        @Override
        public String toString() {
            return "Count:" + threshold + ",ms:" + ms;
        }
    }
    
    private final static Map<String, List<Entry>> summary = new HashMap<>() ;
    
    public static void pressureTest(final Collection<String> list, int threshold) throws InterruptedException {
        System.out.println("Start pressure testing the map ["+list.getClass()+"] use the threshold ["+threshold+"]");
        long totalTime = 0L;
        final int MAX_THRESHOLD = 50000;
        for(int i = 0; i < 5; i++) {
            final AtomicInteger counter = new AtomicInteger(0);
            long startTime = System.nanoTime();
            ExecutorService executorService = Executors.newFixedThreadPool(threshold);
            
            for(int j = 0; j <= threshold; j++) {
                executorService.execute(() -> {
                    for(int x = 0; x < MAX_THRESHOLD && counter.getAndIncrement() < MAX_THRESHOLD; x++) {
                        Integer randomNumber = (int) Math.ceil(Math.random() * 600000);
                        list.add(String.valueOf(randomNumber));
                    }
                });
            }
            
            executorService.shutdown();
            executorService.awaitTermination(2, TimeUnit.HOURS);
            long endTime = System.nanoTime();
            long period = (endTime - startTime) / 1000000L;
            System.out.println(MAX_THRESHOLD + " element added in " + period + " ms.");
            
            totalTime += period;
        }
        
        List<Entry> entries = summary.get(list.getClass().getSimpleName());
        if (entries == null) {
            entries = new ArrayList<>();
            summary.put(list.getClass().getSimpleName(), entries);
        }
        entries.add(new Entry(threshold, (totalTime/5)));
        System.out.println("For the map ["+list.getClass()+"] the average time is " + (totalTime/5) + " ms.");
    }
    
    public static void main(String[] args) throws InterruptedException {
        for(int i = 10; i <= 100;) {
            pressureTest(new ConcurrentLinkedQueue<>(), i);
            pressureTest(new CopyOnWriteArrayList<>(), i);
            pressureTest(Collections.synchronizedList(new ArrayList<>()), i);
            i += 10;
        }
        
        summary.forEach((k, v) -> {
            System.out.println(k);
            v.forEach(System.out::println);
            System.out.println("=============================");
        });
    }
}

执行结果：

SynchronizedRandomAccessList
Count:10,ms:17
Count:20,ms:18
Count:30,ms:19
Count:40,ms:44
Count:50,ms:25
Count:60,ms:27
Count:70,ms:22
Count:80,ms:32
Count:90,ms:53
Count:100,ms:29
=============================
ConcurrentLinkedQueue
Count:10,ms:32
Count:20,ms:15
Count:30,ms:16
Count:40,ms:77
Count:50,ms:37
Count:60,ms:29
Count:70,ms:52
Count:80,ms:20
Count:90,ms:22
Count:100,ms:38
=============================
CopyOnWriteArrayList
Count:10,ms:6017
Count:20,ms:6619
Count:30,ms:7577
Count:40,ms:8506
Count:50,ms:8007
Count:60,ms:7811
Count:70,ms:7067
Count:80,ms:6442
Count:90,ms:6684
Count:100,ms:7048
=============================

结果表明CopyOnWriteArrayList在插入数据的情况下速度确实很低