HP BL Cluster Platform
HP BL Cluster Platform

The compute node is the heart of your high performance cluster. The application(s) to be run and the size of your cluster will be the determining factor in the type of node, the number of nodes and the path to future expansion.

NordStar Group offers three computing node platforms by Hewlett Packard and Hitachi. NSG can help you select the best compute node platform to meet your requirements based on the proposed use of the cluster and your budget.

C7000 Chassis
C7000 Chassis

C7000 Chassis
The Hewlett Packard line of chassis is available in two options, the c3000 and c7000, (Click on the link below for a detailed comparison).

HP Blade chassis are designed for and have built in redundancy to maximize uptime. Each HP chassis purchased will come with the maximum number of fans and power supplies to ensure that you not only have full redundancy but also provide you with an easy upgrade path.

HP C Class Chassis, HP Chassis Redundant Features, HP BL460 Server, HP Switches

Head Node
DL 360

Head Node
The head node's functions include the OS provisioning, network management, scheduling, job submission and control in smaller clusters. This node is the heart of the cluster and facilitates compute node communication and job processing along with DHCP, DNS and NFS.

HP DL360, HP DL Servers

Job Node
Job Node

Job Node
The Job Node, as the name implies is responsible for cluster job submission, management and monitoring. In smaller clusters, this function can be performed by the head node. When multiple jobs are being sent to the cluster, the job node places the jobs in a queue, allocates resources and initiates the jobs on the compute nodes. The job node can also schedule jobs to run at a certain time or when system resources become available.

HP DL Redundant Features

Storage Node
Storage Node

Storage Node
The storage node organizes, pulls and delivers the raw data from the storage system to complete a job.

HP Connectivity

Storage
Storage

Storage
High Performance Computing storage is very different than your typical enterprise storage systems. HPC Applications tend to use large amounts of data for a short period of time. HPC storage system needs to offer a large capacity accessible at high speed and to be highly expandable. They also need to have the ability to share the same data with multiple servers at the same time.

HP P2000 MSA, Hitachi