Using Hazelcast for clustering

By default, clustered implementations of Spotfire Server use the Hazelcast distributed data grid product to support data clustering.

Hazelcast requires practically no configuration, and in most cases is a sufficient option for clustering.

However, Hazelcast is an unsecure option. To enable data exchange through Hazelcast, a port (by default, 5701) must be open on each Spotfire Server. These ports are not protected by any TLS; Hazelcast uses plain TCP/IP connections for the data exchange between servers.
Important: If you do implement clustering with Hazelcast, the firewalls should be configured for maximum security and, ideally, the ports should be open only to other Spotfire Server instances.
Note: If you have multiple network interfaces on your Spotfire Servers, you may need to configure Hazelcast to Bind to Any Network Interface. To do this, open a command line and export the active server configuration by using the export-config command. Then run the following command:
config set-config-prop --name=clustering.hazelcast.bind-on-any-interface --value=true
Then import the configuration back to the Spotfire database by using the import-config command, and restart the Spotfire Servers sequentially. For additional information on executing commands, see Executing commands on the command line.

If your implementation requires secure connections between the servers in a cluster, you can install TIBCO ActiveSpaces® and configure Spotfire Server to use it for secure TCP/TLS transport. For details, see Using ActiveSpaces for clustering.

For general information about Spotfire Server clusters, see Clustered server deployments.