Provisioning and System Management Software — Platform Computing

 
To achieve the greatest performance and maximize overall capabilities when deploying small- to large-scale clusters, you must factor the right combination of provisioning and system management software and tools into the overall operating environment.

Platform Cluster Manager automates system initialization and software deployment with a complete range of tools:

• Device drivers
• Installers
• Cluster management tools
• Resource and application monitoring
• Interconnect support
• Platform Lava, a powerful open source job scheduler
• Options such as diskless nodes and image- or packet-based provisioning

The Platform Cluster Manager — Dell™ Edition features a unique kit architecture designed to automatically install and configure itself. This kit helps simplify cluster environments, from software installation through system maintenance throughout the system life cycle. The kit includes:

• Cacti kit — reporting tools for gathering and graphing node performance metrics
• Ganglia kit — tools for resource monitoring
• High-performance computing (HPC) kit — collection of tools, libraries and utilities
• Lava kit — powerful workload scheduler, compatible with Platform Load Sharing Facility (LSF)
• LSF kit — gold standard in commercial workload schedulers
• Nagios® kit — package for host, service and network monitoring
• NTOP kit — solution for monitoring network bandwidth and analyzing traffic
• OpenFabrics Enterprise Distribution (OFED) kit — collection of drivers to support servers, storage and facilities such as IP over InfiniBand (IPoIB)
• Dell kit — utilities, drivers and open manage agents to streamline installation and management

Platform HPC Suite — Simplifying Cluster Lifecycle Management

Platform HPC Suite of provisioning software enables automatic software deployment and management for cluster-based architectures and applications. Building on the powerful and highly robust Platform LSF workload scheduler, Platform HPC Suite now integrates a broader array of precertified industry-leading management tools into one easy-to-deploy, centralized framework. Platform HPC Suite is specifically designed to provide a complete provisioning and management solution for deploying, optimizing and managing systems designed for maximum performance and advanced applications and analysis.

Exclusive to Dell, Platform HPC Suite is customized to the wide array of resource pools and is orderable in two distinct models:

Platform HPC Workgroup — Dell Edition is designed for small workgroups, departmental and mid-sized systems to dramatically improve operational performance within a simplified foundation package and limited to 32 nodes, it is a combination of:

• Platform Cluster Manager — a cluster management solution that is simple to manage and deploy
• Platform LSF Workgroup Edition — a sophisticated workload scheduler
• Platform MPI™ — industry-leading Message Passing Interface (MPI) implementation
• Platform Adaptive Cluster — turns static clusters and grids into dynamic, shared-HPC cloud environments

Platform HPC Enterprise — Dell Edition, an extension to the Workgroup edition, is designed for simplified deployment and monitoring, while maximizing total utilization and performance capabilities required for large-scale, multiuser environments and applications. It combines all of the Workgroup components with the following:

• Platform Report, Track, Monitor (RTM) — comprehensive and flexible operational dashboard for workload management
• Platform Application Center — flexible interface to HPC clusters

The result is a complete cluster solution that helps provide superior performance, reliability and price, while removing the complexity associated with clusters based on Linux.