zhixinwen opened a new pull request, #3092: URL: https://github.com/apache/kvrocks/pull/3092
If we want no data loss, we have to turn on `rocksdb.write_options.sync`. This would create a big bottleneck for replication throughput because of its single threaded nature. However, with ack based replication impelemented, the replica does not necessarily have to sync for every write. It just needs to sync before it acks. This PR adds an option `replication-group-sync`. If both `replication-group-sync` and `rocksdb.write_options.sync` are on, replication would call sync once before ack. This would still guarantee all the acked changes would have no data loss, but the replication throughput would be much higher. -- 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]
