Ø I keep hearing that the minimum number of Cassandra nodes required to achieve Quorum consensus is 4 I wonder why not 3? In fact, many container deployments by default seem to deploy 4 nodes. Can anyone shine some light on this?
I think it may be due to the following (note : I am assuming, here, a “vnode” cluster) a) When using 3 nodes, and QUORUM, the cluster can tolerate the loss of a node, but in that case, each of the remaining nodes will have a +50% workload b) When using 4 nodes, in case of the same loss, each of the remaining nodes will have (approximately) a +33% workload Option (a) will impact more the cluster stability than (b). Dominique [@@ THALES GROUP INTERNAL @@] De : Kant Kodali [mailto:k...@peernova.com] Envoyé : samedi 17 décembre 2016 22:21 À : user@cassandra.apache.org Objet : quick questions I keep hearing that the minimum number of Cassandra nodes required to achieve Quorum consensus is 4 I wonder why not 3? In fact, many container deployments by default seem to deploy 4 nodes. Can anyone shine some light on this? What happens if I have 3 nodes and replication factor of 3 and consistency level: quorum? I should be able to achieve quorum level consensus right. If Total node = 3, RF=2 and consistency level = Quorum. Then I understand the quorum level consensus is not possible because the number of replica nodes here are 2. This also brings up another question does number of replica nodes always have to be an odd number to achieve quorum level consensus? If so, what happens when a replica node goes down ? it would still serve the requests but the quorum level consensus is not possible? Thanks kant
