Installing a Cluster with a Separated Configuration Database
This procedure creates two clusters: one for the server nodes and one for configuration database nodes. The nodes in the server cluster point to the servers in the configuration database cluster for their configuration data. The nodes in the configuration database cluster replicate their data between each other.
The following diagram illustrates this installation scenario:
To install a separate configuration database, you first install the configuration database on servers in a cluster, then install the server on nodes in a separate cluster. At least one configuration database must be running before you can install the server.
Note: In the case of a separated database configuration cluster setup, as shown here, you can apply load balancing only to the Spectrum server cluster. You cannot apply load balancing to the separated database cluster.
-
Place the Spectrum Technology Platform installer on the server where you want to install the configuration database.
Note: We recommend a minimum of three nodes per cluster.
- Run installdb.sh.
- Install the configuration database on each node in the cluster.
-
Configure property settings on each node in the cluster.
-
Start each configuration database node in the cluster consecutively.
Run the DatabaseDirectory/server/bin/server.start script.Note: Verify that the configuration database is fully started before continuing by opening the log file DatabaseDirectory/repository/logs/debug.log.
-
Install Spectrum Technology Platform on each node in the cluster.
For instructions, see Installing a New Server.Important:
- Do not start the server when the installation completes. You must complete the remaining steps before starting the server.
- When installing each server select the Server only option in the installer and provide the host and port of one or more of the configuration database servers. You can find the port in the file SpectrumDirectory/Database/repository/spectrum-container.properties. The port is specified in the
spectrum.repository.server.connector.bolt.port
property.
-
On each node, configure settings in the
SpectrumDirectory/server/conf/spectrum-container.properties
file.
-
On each node, configure settings for Search index.
-
Confirm the following setting on each node in the cluster.
- SpectrumDirectory/server/conf/spectrum-container.properties
spectrum.repository.cluster.mode=CAUSAL
Note: Thespectrum.repository.cluster.mode
setting shown here is for the Spectrum Technology Platform server running in a cluster.
- SpectrumDirectory/server/conf/spectrum-container.properties
-
Start each server in the cluster.
Run the SpectrumDirectory/server/bin/server.start script to start the Spectrum Technology Platform server.For more information, see Starting a Cluster.
-
On each node, open the SpectrumDirectory/index/elasticsearch.template file, insert the hash character
#
to comment out the following line:# cluster.initial_master_nodes: node-1,node-2,node-3,...
Important: This setting must be commented out after you start the cluster for the first time. It should remain commented out for all subsequent restarts of the cluster, as Elasticsearch handles adding and removing nodes after the cluster is started initially.