[ad_1]
InfiniBand® is a high-performance network and storage system that uses serial links and high-speed processors. It can address 64,000 nodes and deliver speeds of up to 2.5 Gbps. InfiniBand® offers redundancy for increased reliability and can be distributed geographically. Switched fabric communication links provide overall device bandwidths required by supercomputers.
InfiniBand® is a network communication and storage system that makes use of serial links and high-speed processors to address performance and reliability issues for high-performance computing. It is capable of addressing 64,000 nodes and can deliver speeds of up to 2.5 gigabits per second (Gbps). This speed is around 2,500 million bits per second (Mbps), which can download a 1 gigabit file in less than 4 seconds.
The high-speed serial links now available, such as optical fiber, make InfiniBand® possible in part. In the early days of computing, serial communications were considered too slow for high data rates, and computer buses, which were mostly parallel, were the only choice. Fast serial links and serial buses are very convenient and the points of failure on the hardware are fewer. In a 32-bit parallel bus, for example, there are at least 32 sets of bit drivers, and very often push contacts between interface cards are involved. The probability of failure is therefore higher in parallel devices in this respect.
InfiniBand® also offers redundancy for increased reliability. Storage devices and hosts in InfiniBand® can be distributed geographically for optimized protection. For example, a large database application that requires 24 x 7, or 100%, uptime may consider many techniques including site redundancy, server redundancy, and storage redundancy.
Site redundancy replicates an entire application site running concurrently with the primary application site, which at first may seem like a waste of computing resources and computer networks. Wisdom of approach is observed when an operation at the primary application site is partially or totally disrupted. A backup site that always works in conjunction with the primary site is then promoted as the primary site and continues to process until the primary site is up and online.
Server redundancy replicates the primary server to ensure continuous operation in the event of a failure. Whether collocated or remote from the primary, the secondary server can be running parallel so that any serious failure in the primary server will not interrupt the service. In disk mirroring, the server’s local storage is written to two storage systems, while in site mirroring, the backup site does the work of the primary server and synchronizes with it.
A switched fabric communication link is capable of providing the overall device bandwidths required by supercomputers. Fabric refers to a relatively large network of hosts and network devices that are interconnected to each other. In a switched fabric, high-speed remote storage and fast remote host services are enhanced by switching technologies that reduce processing overhead and collisions as data packets are sent across the network.
[ad_2]