voltaire.com
  Networked Computing Infrastructure™ for the Next Generation Data Center

Products

|

GridVision™ Enterprise



 

  • Create and Manage Complete Virtual Data Centers From Sets of Virtual or Physical Server, Storage, and Network Resources
  • Simplifies Enterprise Grid Deployment 

Enterprise Grid Deployment, Extensive Monitoring, grid, grid environments, GridVision Enterprise

Overview
Voltaire GridVision™ Enterprise is a first-of-its kind software solution for the management and automation of heterogeneous and dynamic grid environments in the data center. The software accepts high-level user definitions of data center elements and topologies, translates them to actual configurations, and dynamically maps them to real or virtual server, network, and storage resources. This enables true automation of repetitive IT tasks associated with network, server and storage provisioning to deliver unprecedented efficiency and allows full exploitation of virtualization technologies. Provisioning tasks that used to take days and involve multiple groups within IT can now be reduced to seconds, resources can be used much more efficiently, and new service levels can be achieved.

    Key Features

  • Improves resource utilization and data center efficiency
  • Delivers extensive discovery, monitoring, and diagnostics for server networks
  • Provides simple and aggregate views and management of physical or virtual network, server and storage resources
  • Dynamically creates logical data center environments that consist of independent resources based on high-level user definitions
  • Automates repetitive IT tasks associated with network, server and storage provisioning
  • Provides service oriented infrastructure for grids
  • Integrates with third-party virtualization and management tools through an open model and extensive set of APIs


Dynamically Maps Requirements to Data Center Infrastructure
GridVision Enterprise automatically discovers and monitors servers, storage, and network elements and places them in a resource pool which is made available to the entire enterprise while being managed as a resource. When the user is ready (or the application calls for it), GridVision Enterprise takes the user definition of a single server/cluster, or even a complete data center model and deploys it using resources from the pool. Servers are matched with relevant storage volumes. Server I/O and network topologies are virtualized to match definitions and a variety of services are configured to satisfy the predefined requirements. If more resources are needed, GridVision Enterprise will configure and add them. When a failure occurs, resources can be swapped, and when the task is completed all the resources can be returned to the resource pool.

Next Generation IT Using Model-Based Approach
GridVision Enterprise uses an innovative, object-based model to describe logical building blocks of compute, network, storage elements and how they are connected or used in the grid. Objects and topologies can then be collected into logical data center environments. The GridVision Enterprise model focuses on user or application requirements that can be described in simple terms. GridVision Enterprise objects are stateless and resource independent in a way that allows the re-use and inheritance of definitions, deployment across different technologies (over physical or virtual resources, or both) and simple extensions by users and partners. The model can provide feedback and notification on a variety of physical and logical changes. GridVision Enterprise provides a rich set of web-services APIs and can be accessed via CLI, Web, or SNMP.

Automated Discovery and Extensive Monitoring
GridVision Enterprise performs agent-less discovery of nodes, devices, and connectivity, and constantly monitors fabric/device behavior and performance. Collected data is analyzed to provide a central view of status, utilization and failures. Data is correlated between elements and between physical and virtual/logical resources, and can be presented in multiple views with different summary or grouping levels. The GUI shows both physical and logical views of the data center and data can be exported using variety of APIs. This provides the end user with a scalable plug & play infrastructure that can be operated immediately after assembling the hardware. Monitored data can be used to perform corrective measures and/or to shift resources in order to meet changing demands.

Data Center Virtualization and Service Oriented Infrastructure (SOI)
GridVision Enterprise provides network and I/O virtualization by programming Voltaire's advanced Grid Director™ multi-service switches, which can create virtual LAN or SAN networks across different technologies and create multiple virtual NICs and disks per server. GridVision Enterprise can interact with external server and storage virtualization technologies to provide complete data center virtualization. GridVision Enterprise integrates with server virtualization technologies such as Xen or VMware, with server management tools and with storage virtualization tools through an open API.

By integrating SLA/SLO management, orchestration or home grown tools with GridVision Enterprise, users can now build service-oriented data centers. GridVision Enterprise can be used to dynamically map service objectives to real infrastructure and provide real-time feedback on status, failures, and utilization.

Enterprise Grid Deployment, Extensive Monitoring, grid, grid environments, GridVision Enterprise
GridVision Enteprise architecture

Powerful Integration of Different Fabric Technologies
Voltaire's family of Grid Director multi-service switches provides InfiniBand, Ethernet, and Fiber Channel connectivity. GridVision Enterprise was designed to focus on functional, fabric-independent, physical and logical models. The same user interface is used regardless of the physical media. Virtual servers can boot from iSCSI, InfiniBand or Fibre Channel disks based on whichever is available at the moment. VLANs can be formed across mixed Ethernet and InfiniBand islands. All of this allows true consolidation of data center resources, eliminates deployment challenges, and enables users to maximize the benefit of each technology.