You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@rocketmq.apache.org by GitBox <gi...@apache.org> on 2020/04/01 15:04:37 UTC

[GitHub] [rocketmq] yongliangcode edited a comment on issue #1910: CPU spike on slave node

yongliangcode edited a comment on issue #1910: CPU spike on slave node
URL: https://github.com/apache/rocketmq/issues/1910#issuecomment-607302787
 
 
   > @yongliangcode 非常详细的描述,看上去调整后只在slave 上面出现了上述状况,原因可能有下面几点:
   > (1)四个slave 应该是同时部署在了两个ecs上面,所以相对io压力会比主大一些,顺序写也会受影响,从监控上面看,内存也几乎满了。
   > (2)看了下内核参数的影响,有个地方有些问题:vm.dirty_ratio应该要大于vm.dirty_background_ratio,试下把vm.dirty_ratio改为25,vm.dirty_background_ratio不变。
   > (3)尝试下升级操作系统,好像是有个操作系统bug也会导致相关的情况,大于2.7的内核对于内存的使用会更加激进一些,当然vm. min_free_kbytes 这些参数也不起作用了。
   > 
   > 服务器如果那时候还进得去的话,能不能看下飙高的线程:)
   
   谢谢杜总大佬回复,有一次没有挂掉,上去有查看过,以下是截图,比较诡异是jps进程/java进程占用CPU飙高,进程号还轮换变化。
   另外升级Linux内核版本,建议使用哪个版本?2.7以上?
   
   ![image](https://user-images.githubusercontent.com/46984218/78152262-24264800-746c-11ea-95b4-4eb5dd52f89b.png)
   
   ![image](https://user-images.githubusercontent.com/46984218/78152430-5e8fe500-746c-11ea-827f-3f9f3d635888.png)
   
   ![image](https://user-images.githubusercontent.com/46984218/78152692-bc243180-746c-11ea-81d2-fce86ea47b44.png)
   
   ![image](https://user-images.githubusercontent.com/46984218/78152723-c6dec680-746c-11ea-979b-997a90ed6822.png)
   
   
   

----------------------------------------------------------------
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.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services