Types of Clusters
Types of Clusters

High Performance computer clusters can be designed to perform various types of jobs. Clusters fall into three categories regardless of the specific use:

• High Performance Computing Clusters,
• Load Balancing Clusters, and
• Fail-over Clusters.

Click on the boxes below to learn more.

High Performance Computing Cluster
High Performance Computing Cluster
The compute nodes in a High Performance Computing Cluster are configured to work together as one unit for completion of a job. A large dataset will be divided among the available nodes for computation. Each node will complete its portion of the work on the dataset and the information will be reassembled for the final result.
Load Balancing Cluster
Load Balancing Cluster

The compute nodes in a Load Balancing Cluster are configured to perform individual queries. The head node in the cluster determines which node is available when a request is received and sends the request to that node. Load Balancing Clusters are used in high transactional computing scenarios. Examples include facebook, ebay or online gaming sites.

Fail-Over Cluster
Fail-Over Cluster

A Failover Cluster is a group of servers that work together to provide high-availability of services and/or applications. If one node fails, another node in the cluster will take over its processes without any downtime. Cluster roles are monitored to determine if they are working properly. If not, the processes are restarted or moved to another node.

HPC Applications
HPC Applications
In this section we will explore the primary types of applications that will be running the cluster.