dongeforever commented on issue #3461: URL: https://github.com/apache/rocketmq/issues/3461#issuecomment-962378450
The raft consensus considers consistency firstly。 In a 2n+1 group，the failure tolerance is n. If more than n nodes are failed, the left nodes may provide corrupted data. Anyway, "Read from a slave" can be an option, both in "master-slave mode" and "dledger mode". but need take care of some cases: * the max offset in slave may be smaller than the currently committed offset, which may result in an out-of-range exception, then cause the committed offset to be reset. -- 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]
