[GitHub] [rocketmq] ticsmtc6 opened a new issue #1602: org.apache.rocketmq.store.ha.HAConnection.ReadSocketService#run this.selector.select(1000) cost 5000+ms

[GitBox <gi...@apache.org>]

ticsmtc6 opened a new issue #1602: org.apache.rocketmq.store.ha.HAConnection.ReadSocketService#run this.selector.select(1000) cost 5000+ms
URL: https://github.com/apache/rocketmq/issues/1602
 
 
   1. Please describe the issue you observed:
   
   - What did you do (The steps to reproduce)?
   4 master 4 slave scale out to 10 master 10 slave, i found many TIMEOUT_CLEAN_QUEUE exception, in store.log i found a lot of master-slave sync error(do sync transfer other node, wait return, but failed, topic:).  so i added a log around of HAConnection.ReadSocketService#run this.selector.select(1000) as follow:
   
   **Long beginTime = System.currentTimeMillis();
   this.selector.select(1000);
   log.info("[ADD DEBUG LOG] master read socket service select eclipse: " + (System.currentTimeMillis() - beginTime));**
   
   then i found a lot of log shows this line cost exceed 5000ms
   
   - What did you expect to see?
   Selector.select() method shoud block at most 1000ms
   
   - What did you see instead?
   A lot of times Selector.select() method cost exceeds 5000ms
   
   2. Please tell us about your environment:
   rocketmq version:4.2.0
   Aliyun 16C64G 10m-10s
   
   3. Other information (e.g. detailed explanation, logs, related issues, suggestions how to fix, etc):
   Only new added 6m-6s report this exception.
   When TIMEOUT_CLEAN_QUEUE exception occurs, I found some TCP retransmission，Is there a connection between the two？

----------------------------------------------------------------
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