We're running DSE v4.8.5 (which uses Cassandra 2.1.13), and we need to migrate to the community edition of Cassandra v3.11 and then retire the old DSE servers.
Someone (it might have been me) suggested the idea of adding the new C* 3.11 nodes to the existing DSE (C* 2.1) cluster and then retiring the old nodes once the new nodes are caught up, but I'm not sure what the risks of running 2.x and 3.x together are.
I've found zero-downtime upgrade guides which talk about updating each node in turn, but I'm not entirely sure what difference it makes if the server is upgraded to 3.11 or replaced with a new server running 3.11. It seems to me that if it works for an upgraded instance, it should also work for a new instance.
So, my question to the community is: What might happen if we try to add a C* 3.11 node to a cluster running C* 2.1, with the intention of retiring the 2.1 nodes?
