I found that the gossip always bring a lot of network consumes, is anyway to modify it? like, do it in multi-raft way, split the nodes in cluster into some groups, and the groups has master nodes to gossip with other master nodes, and the groups use gossip inside too.
Comment From: madolson
@Odysseusailoon Hey, we have two existing issues for this: https://github.com/redis/redis/issues/8948 <- high level tracking for improving cluster https://github.com/redis/redis/issues/3929 <- Your specific comment.
We've been looking at a couple of ideas in this space including reducing failover messages to smaller groups, like multi-raft. I'm going to close it since I think the conversation should happen elsewhere, but please feel free to re-open if you believe your comment wasn't addressed.