[GitHub] [rocketmq] yylstudy opened a new issue #3781: 压测生产者内存溢出

Fri, 21 Jan 2022 02:17:48 -0800 


yylstudy opened a new issue #3781:
URL: https://github.com/apache/rocketmq/issues/3781


   1. Please describe the issue you observed:
   压测生产者内存溢出
   2. Please tell us about your environment:
   操作系统：centos
   broker版本：v_4.9.2
   rocketmq-client版本：4.9.2
   
   创建生产者代码
   
   asyncSenderExecutor 80个线程，队列长度为50000，线程池拒绝策略为CallerRunsPolicy
   callbackExecutor，80个线程，队列长度为50000，线程池拒绝策略为：AbortPolicy
   
![image](https://user-images.githubusercontent.com/26402953/150509710-c29b6ac1-e498-4a7c-9875-0b2c7e0fb8dd.png)
   
   异步发送消息
   
![image](https://user-images.githubusercontent.com/26402953/150507857-9f747543-2613-4b59-bf20-ea2f015d3248.png)
   
   当压测到1亿数据量时，生产消息很慢
   查看生产者日志，invokeSync call timeout错误
   
   查看生产者gc时间，发现一直在fullgc
   
![image](https://user-images.githubusercontent.com/26402953/150508149-fbef16ef-9229-4996-8277-0e8bd161a21e.png)
   
   导出dump文件，发现大量的CompletableFuture对象
   
![image](https://user-images.githubusercontent.com/26402953/150508712-593414c3-2deb-482f-bee0-e215e79397ed.png)
   
   请问下这个怎么解决，是不是因为单个客户端请求量太大了导致回调队列太长，但是我配置了
   callbackExecutor 线程池拒绝策略为：AbortPolicy 也没有效果。请问下怎么解决呢 谢谢
   
   
   
   
   
   
   
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to