Perform Forward Only Bindings

Let’s assume you started with a single node which is node1 and have data on node1. The
node1 is being used actively for transactional operations and is the only node you have at
the moment which is taking all the transactions. Now you have setup node2 because
node1 alone is not able to take the complete application workload and you wish to connect
node2 to node1 so that node2 can take some of the workload of node1.
In order to do so, you must run the add-node command on node1 and not node2.

blobcity>add-node node2

When the above command is run, it is treating node1 as the primary node and node2 as
the new node to be added into the cluster. All datastore and collection schemas from
node1 will get replicated to node2. Only the schema is, the data is not. The node 2 will
receive data based on the replication settings of each individual data collection.

Running the query on node2 to connect it to node1 will not work as node1 already has
data, so the connection would not be permitted.

In case you have loaded data onto node2 after setting it up, or changed any of the user
permission or added new users into the node2, the command for connecting node1 to
node2 will not work, as now node2 is not a fresh node, but has data pre-loaded.

The forward binding concepts are equally important when adding backup clusters to
existing clusters.