Crazylychee opened a new issue, #9616:
URL: https://github.com/apache/rocketmq/issues/9616

   ### Before Creating the Enhancement Request
   
   - [x] I have confirmed that this should be classified as an enhancement 
rather than a bug/feature.
   
   
   ### Summary
   
   Currently, metadata synchronization between the master and slave nodes uses 
scheduled RPCs. In controller mode, a RocketMQ master and slave node experience 
a metadata loss window of approximately 12 seconds, while in standard 
master-slave synchronization, this window increases to approximately 60 
seconds. This can lead to many unpredictable issues. Our goal is to use system 
topics to implement incremental metadata synchronization, thereby reducing the 
window of data inconsistency.
   
   I have written the proposal and you can see it by the link below:
   Chinese version:
   
https://m0sgzrp00oj.feishu.cn/docx/CvbedEJXto7A2BxfYVXcLXATnbd?from=from_copylink
   
   We will implement this proposal in multiple phases:
   Phase 1
   - [x] Complete metadata change notifications
   
   Phase 2
   - [x] Complete scheduled storage and deletion of master node snapshots
   - [x] Complete slave node burst snapshot synchronization and incremental 
synchronization
   - [ ] Complete adaptation of this functionality to the JSON and RocksDB 
storage formats
   
   Phase 3
   - [ ] Complete relevant unit tests, integration tests, and chaos tests
   - [ ] Complete test reports
   
   ### Motivation
   
   null
   
   ### Describe the Solution You'd Like
   
   null
   
   ### Describe Alternatives You've Considered
   
   null
   
   ### Additional Context
   
   _No response_


-- 
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