I see in your network topology, it used P2P network. Every node would store a full node catalog which is a global view of the entire network’s nodes. Does anybody get a bench mark of a big cockroach DB network? How does it works under this condition?
That’s a great question! The answer really depends on your workload, but we can say we have customers running clusters with hundreds of nodes in production. A good example is Baidu, we have a case study on their usage of CockroachDB here. They are serving massive amounts of data as well as processing thousands of queries per second.
Let me know if you have any other questions!
Thanks Ron for your quick feedback.