Question-61: In case of virtual nodes how does tokens are assigned for each node?

Answer: In case of virtual node either an allocation algorithm is used or a random selection algorithm is used to specify the number of tokens on distributed nodes within the datacenter. However, remember that all systems/nodes in the datacenter must use the same token allocation algorithm.

Question-62: In case of single token architectural how tokens are evenly divided?

Answer: In case of single token architecture to ensure the data is evenly divided across the node in the cluster you must enter values in the “initial_token” parameter in the cassandra.yaml file and this needs to be done for each node.

Question-63: If I have used one partitioner to determine the token, then I can use another partitioner or replace the partitioner?

Answer: Please note that partitioner is not compatible with each other. Data partitioned with one partitioner cannot be easily converted to support by another partitioner. 

Question-64: Can you explain what is the use of Snitch?

Answer: Snitch determines which datacenter and racks on Node belongs to. It is the Snitch responsibility to inform the database about the network topology, so that request can be routed efficiently and distribution of replicas happens smoothly by grouping machine in to datacenter and racks. All the nodes in the cluster must use the same snitch. It is also recommended that you should not have more than one replica in the same rack. 

Question-65: What is the advantage of using dynamic snitching?

Answer: Dynamic snitching is the default one which uses the dynamic snitch layer that monitors read latency and routes request away from poorly performing nodes.