jemc commented on issue #2848: URL: https://github.com/apache/kvrocks/issues/2848#issuecomment-2851516853
As far as I can tell the only commands which can induce failover are: - `CLUSTERX SETNODES [...]` (used by the kvrocks controller to sync cluster state when in Clustered mode, including master/replica relationships) - `SLAVEOF [...]` (used when _not_ in Clustered mode) What's not clear to me is whether simply calling the `SLAVEOF` command on the current master node is guaranteed to fully replicate all acknowledged writes before switching to being a replica of the other node. In Redis this is _not_ safe to do, in that acknowledged writes may be lost. KVRocks doesn't seem to have any documentation or indication of how to do a controlled failover gracefully. -- 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]
