BlockingQueue定时或达到一定条数读取队列

最新推荐文章于 2023-05-11 17:43:51 发布

ZYJ_2012

最新推荐文章于 2023-05-11 17:43:51 发布

阅读量4.2k

点赞数 4

分类专栏： java

本文链接：https://blog.csdn.net/ZYJ_2012/article/details/53738692

版权

java 专栏收录该内容

7 篇文章 1 订阅

订阅专栏

谷歌工具类maven：

<dependency>
    <groupId>com.google.guava</groupId>
    <artifactId>guava</artifactId>
    <version>18.0</version>
</dependency>

消费队列线程：

public class LogsConsumer implements Runnable{

    /** 日志 **/
    private static final Logger LOG = LoggerFactory.getLogger(LogsConsumer.class);

    private volatile boolean isRunning = true;

    private BlockingQueue queue;

    public LogsConsumer(BlockingQueue queue) {
        this.queue = queue;
    }

    /**
     * 消费者
     */
    public void run() {
        while (isRunning) {
            try {
                //每次取出的数据存放到logsDatas里
                List<String> logsDatas = new ArrayList<String>();
                //每次到1000条数据才进行入库，等待1分钟，没达到1000条也继续入库
                Queues.drain(queue, logsDatas, 1000, 1, TimeUnit.MINUTES);
                this.insertLogs(logsDatas);
            }catch (Exception e){
                LOG.error("LogsConsumer drain queues error:",e);
            }

        }
    }

    /**
     *批量添加方法
     * @param logsDatas
     */
    private void insertLogs(List<String> logsDatas) throws Exception {
        if (!CollectionUtils.isEmpty(logsDatas)){
            Client client = ElasticsearchClientUtil.getClient();
            BulkRequestBuilder bulkRequest = client.prepareBulk();
            for (String logsData:logsDatas){
                bulkRequest.add(client.prepareIndex("statistics", "pvanduv").setSource(logsData));
            }
            bulkRequest.get();
        }
    }

}

启动线程：

BlockingQueue queue = BlockingQueueFactory.getQueue();
LogsConsumer consumer1 = new LogsConsumer(queue);
LogsConsumer consumer2 = new LogsConsumer(queue);
ExecutorService service = Executors.newCachedThreadPool();
service.execute(consumer1);
service.execute(consumer2);