This is James, I am not quiet understanding contact points when some specific situation like below,
For one instance, One universe has 3 nodes and replication setting is 3. and contactpoints in my code is (node1,node2,node3). For cassandra, I figure it will contact any ONE node and get same query result
https://teddyma.gitbooks.io/learncassandra/content/client/which_node_to_connect.html, so technically, my contactpoint can be only node1, and still get same result.
If its still 3 nodes but replication is 1, is there any difference, between if my contact point is (node1,node2,node3) or only (node1)? If I only have 1 node as contact point, can I get same result when I use 3 nodes? or it could be nothing result because data is stored on one the others 2 nodes.
Contact like (node1,node2,node3), YB will pick up a better node to connect? what is the advantage we should or no matter use multiple contact points?
Is it good enough for me to check if we can ping node IP to decide this node is healthy or not?
Thanks in advance