动态调整线程池参数

目标

1. 实现动态调整线程池参数
2. 对线程池运行情况进行监控

实现

一，线程池可调整的参数

1. 核心线程数
2. 超时时间
3. 最大线程数
4. 拒绝策略

而队列BlockingQueue因为是final类型，所以没有对外修改入口。但可以通过重写LinkedBlockingQueue并把capacity设置为非final。

二，结合配置中心实现动态调整

这里的配置中心使用Apollo, 通过监听配置中心变化，然后更新线程池配置。示例代码如下：

@Slf4j
@Component
public class DynamicThreadPoolConfig {
    /** 线程执行器 **/
    private volatile ThreadPoolExecutor executor;

    /** 核心线程数 **/
    private Integer corePoolSize = 10;

    /** 最大值线程数 **/
    private Integer maximumPoolSize = 20;

    /** 待执行任务的队列的长度 **/
    private Integer workQueueSize = 1000;

    /** 线程空闲时间 **/
    private Long keepAliveTime = 1000L;

    /** 线程名 **/
    private String threadName;

    private Config config = ConfigService.getConfig("项目配置中心namespace");

    public DynamicThreadPoolConfig() {
        init(config);
    }

    /** * 初始化 */
    private void init(Config config) {
        log.info("线程池初始化中..........");
        if (executor == null) {
            synchronized (DynamicThreadPoolConfig.class) {
                if (executor == null) {
                    String corePoolSizeProperty = config.getProperty("corePoolSize", corePoolSize.toString());
                    log.info("修改前的核心线程池:{}",corePoolSizeProperty);
                    String maximumPoolSizeProperty = config.getProperty("maximumPoolSize", maximumPoolSize.toString());
                    String keepAliveTImeProperty = config.getProperty("keepAliveTime", keepAliveTime.toString());
                    BlockingQueue<Runnable> workQueueProperty = new LinkedBlockingQueue<>(workQueueSize);
                    executor = new ThreadPoolExecutor(Integer.valueOf(corePoolSizeProperty), Integer.valueOf(maximumPoolSizeProperty),
                            Long.valueOf(keepAliveTImeProperty), TimeUnit.MILLISECONDS, workQueueProperty);
                }
            }
        }
    }

    /**
     * 监听到配置中心发生变化后，更新线程池配置
     * @param changeEvent
     */
    @ApolloConfigChangeListener
    public void onChange(ConfigChangeEvent changeEvent){
        log.info("线程池参数配置发生变化,namespace:{}",changeEvent.getNamespace());
            for(String key : changeEvent.changedKeys()){
                ConfigChange change = changeEvent.getChange(key);
                String newValue = change.getNewValue();
                refreshThreadPool(key,newValue);
            }
    }

    /**
     * 更新线程池配置
     * @param key
     * @param newValue
     */
    private void refreshThreadPool(String key, String newValue) {
        if (executor == null) {
            return;
        }
        if (ParamsEnum.CORE_POOL_SIZE.getParam().equals(key)) {
            executor.setCorePoolSize(Integer.valueOf(newValue));
            log.info("修改核心线程数key={},value={}",key,newValue);
        }
        if (ParamsEnum.MAXIMUM_POOL_SIZE.getParam().equals(key)) {
            executor.setMaximumPoolSize(Integer.valueOf(newValue));
            log.info("修改最大线程数key={},value={}", key, newValue);
        }
        if (ParamsEnum.KEEP_ALIVE_TIME.getParam().equals(key)) {
            executor.setKeepAliveTime(Integer.valueOf(newValue), TimeUnit.MILLISECONDS);
            log.info("修改线程空闲时间key={},value={}", key, newValue);
        }
    }

    public ThreadPoolExecutor getExecutor() {
        return executor;
    }
}

@AllArgsConstructor
public enum ParamsEnum {

    CORE_POOL_SIZE("apollo.async.executor.thread.core_pool_size", "核心线程数"),
    MAXIMUM_POOL_SIZE("dynamic.maximumPoolSize", "最大线程数"),
    KEEP_ALIVE_TIME("dynamic.keepAliveTime", "线程空闲时间"),
    ;

    @Getter
    private String param;

    @Getter
    private String desc;

}

三，监控方式

修改线程池有关参数重要，但知道何时修改同样重要，可以考虑间隔一段时间进行采集，通过日志输出，达到临界点后告警。
同样，ThreadPoolExecutor也提供获取线程池相关信息的API：

这里通过一个定时任务进行统计,需要注意的是启动类上需要加上EnableScheduling注解

@Slf4j
@Component
@Async
@ConditionalOnBean(DynamicThreadExecutor.class)
public class ThreadPoolMonitorSchedule {

    @Autowired
    private DynamicThreadExecutor dynamicThreadExecutor;

    @Scheduled(fixedDelay = 2000)
    public void watchThreadPoolInfo(){
        log.info("开始统计线程池相关数据");
        ThreadPoolExecutor threadPoolExecutor = dynamicThreadExecutor.getExecutor();

        BlockingQueue<Runnable> queue = threadPoolExecutor.getQueue();
        //线程活跃度：活跃线程数趋向于maximumPoolSize的时候，代表线程负载趋高。
        log.info("核心线程数：{},活动线程数：{},最大线程数：{},线程池活跃度：{},任务完成数：{}," +
                 "队列大小：{},当前排队线程数：{},队列剩余大小：{},队列使用度：{}",
                threadPoolExecutor.getCorePoolSize(),
                threadPoolExecutor.getActiveCount(),
                threadPoolExecutor.getMaximumPoolSize(),
                divide(threadPoolExecutor.getActiveCount(), threadPoolExecutor.getMaximumPoolSize()),
                threadPoolExecutor.getCompletedTaskCount(),
                (queue.size() + queue.remainingCapacity()),
                queue.size(),
                queue.remainingCapacity(),
                divide(queue.size(), queue.size() + queue.remainingCapacity()));
    }


    private String divide(int num1,int num2){
        return String.format("%1.2f%%",Double.parseDouble(num1+"") / Double.parseDouble(num2+""));
    }
}

/**
 *启动类
 */
@MapperScan({"com.demo.dao"})
@SpringBootApplication
@EnableScheduling
public class DemoApplication {
    public static void main(String[] args) {
        SpringApplication.run(DemoApplication.class, args);
    }
}