About clusters

supports clustering.

A cluster consists of a minimum of three instances of and its supporting external services; file shares, a PostgreSQL database or database cluster, and at least one load balancer, such as HAProxy.

clustering uses additional technologies to support the cluster:

• GlusterFS for file shares. Other file systems, such as NFS can be used instead of GlusterFS.
• Consul to provide action locking as needed.
• RabbitMQ to provide a fast, reliable messaging bus.
• HAProxy as a load balancer. Alternate load balancers can be used instead of HAProxy.

In a cluster, both the PostgreSQL database and the deployment of Splunk Enterprise are externalized from the instances. This allows you to scale your database and Splunk Enterprise deployments separately from the nodes.

Splunk SOAR (On-premises) clusters add a level of system redundancy and scaling to your deployment. However, SOAR does not have the ability to automatically scale, or automatically add or remove cluster nodes through external systems such as Kubernetes, AWS, or Azure.

Before creating a cluster, work with your delivery team representative to assess your needs and design your cluster.

Why build a Cluster?

Clustering addresses several important needs:

• Clustering adds horizontal scaling for workloads, allowing for increased capacity.
• Clustering adds redundancy for the platform. One or more cluster nodes can fail and you still have a functioning deployment of .
• Clustering removes system downtime for upgrades or maintenance. You can upgrade individual cluster nodes without taking the entire deployment offline.

Building a cluster

Clusters are built from unprivileged installations, where required services are provided by servers external to . Each node is converted from a TAR file installation using the make_cluster_node.pyc script. See Create a cluster using an unprivileged installation.