Execute the following command:

nodetool gossipinfo

Note the gossip information for both nodes as discussed in the slides.

Notice that even though we executed this command on node1, node1 still knows node2's gossip state. Also notice the node state consists of key-values pairs as discussed in the slides.

Rerun the nodetool gossipinfo command above a few more times and notice the heartbeat values increasing for both nodes.

Run nodetool gossipinfo on node2 instead of node1.

ssh [email protected] "nodetool gossipinfo"

Notice the gossip data is the same.

Terminate your second node by executing:

/home/ubuntu/node2/resources/cassandra/bin/nodetool stopdaemon

Check the gossipinfo on node1. Notice node2's gossip information is still present as it is part of the cluster, but its STATUS state is shutdown.

Restart your second node.

Rerun nodetool gossipinfo on either node, and notice the generation values are the same as before for node1, but are now different for node2 since we restarted node2.

Start cqlsh and execute the following query of the system.peers table which stores some gossip data about a node's peers.

SELECT peer, data_center, host_id, preferred_ip, rack, release_version, rpc_address, schema_version FROM system.peers;

Notice the values here are some of the same values you saw in the terminal. Also notice that a node does not store a row of peer data for itself. By default, cqlsh connects to 127.0.0.1.

Gossip Quiz