Clustering HPC solution : High Performance Computing solution for businesses and organizations.
Mandriva Linux Clustering is a Mandriva clustering solution based on research product CLIC. It's been designed to offer research laboratories an affordable, fast-to-deploy, easy-to-use, heavy-calculation systems based on Clustering.
Features and hardware requirements
Low latency and high-bandwidth are made possible by the use of the SCI or Infiniband connectics between nodes.
Thanks to PXE technology, Mandriva Linux Clustering can be quickly installed. Complete deployment is possible in only a few minutes with the help of Clusterscripts and Kadeploy utilities. URPMI -- the Mandrivalinux package management system and dependency solver -- has been parallelized to allow automated software updating of all nodes simultaneously.
To guarantee an optimal user experience, Mandriva Linux Clustering provides a graphical environment plus numerous administration tools which allow the Cluster administrator to add and remove nodes from a simple clic; create virtual partitions; assign users to partitions; plus graphically configure the entire system. Graphical monitoring of the system is performed via a dedicated tool named "Ganglia".
In addition, Mandriva Linux Clustering offers the ability to use two separate network interfaces, thereby allowing the ability to differentiate the administration network from the computing network.
Based on the popular Mandriva Linux distribution, Mandrivalinux Clustering benefits from a superior range of hardware support. Each node automatically adapts itself depending on the hardware used (SCSI, network, filesystem, etc.), thus allowing superior support for heterogeneous environments. Additionally, Mandriva Linux Clustering supports up to 4 GB of RAM for each node (16 GB for Opteron).
Mandriva Linux Clustering ships as a turn-key clustering environment. Based on the high-performance Mandrivalinux operating system, this product offers SCI and Infiniband drivers, several clustering messaging transit layers (Mpich, Lam, PVM), graphical monitoring tools, parallelized tools (URPMI, Ka-tools, Gexec) plus numerous mathematical libraries.