January 16, 2018 at 11:09New technology speeds WAN up to 40 Gbps
As cloud technology evolves, there is a growing need for efficient data and server management. An important condition is to simplify infrastructure maintenance by transferring business-generated data to the cloud. Fulfillment of new technical requirements entails a multiple increase in the amount of data transferred between cloud environments through WAN channels. Of primary importance is the creation and implementation of new technologies to accelerate data transfer in WAN-networks. One of these technologies developed by Fujitsu will be discussed in this article.
Using WAN acceleration technology in a cloud environment
Today, most often, data in a WAN is transmitted between cloud environments at speeds from 1 to 10 Gb / s. However, for the proper functioning of modern IT solutions, including the Internet of things and artificial intelligence, an even higher data transfer rate is required. Existing WAN acceleration technologies increase the effective transmission speed by reducing the amount of data transmitted by compression or deduplication.
However, in 10-gigabit WANs, the total amount of information is already reaching its maximum values. Existing acceleration technologies in cloud servers cannot seriously increase the data transfer rate, which becomes the bottleneck of the entire cloud infrastructure. In order to bring the speed of cloud services to a whole new level, you must either use processors with a higher clock speed, or use WAN acceleration technologies that work using different algorithms.
Fujitsu has developed WAN acceleration technology that enables data transmission in cloud environments (in real time) at speeds up to 40 Gb / s. This speed is achieved with the simultaneous use of several 10-gigabit channels. At the same time, Field-Programmable Gate Array (FPGA) matrices installed on the server act as accelerators. The technology is designed to help transfer large amounts of data between cloud environments. The implementation of the new development was made possible thanks to the allocation of a special computing node for processing compression operations. It provides highly parallel operation of the main computing nodes, transmitting data at precisely the specified time, based on the predicted completion of each stage of the calculation.
Below are the features of the technology.
1. FPGA matrix parallelization technology uses dedicated computing nodes
Fujitsu has developed FPGA matrix parallelization technology, which can significantly reduce the processing time required for data compression and deduplication. The result is achieved by deploying dedicated computing nodes and ensuring a high degree of parallelization of computing.
Implementing WAN acceleration using servers equipped with FPGAs
2. Optimizing the distribution of data processing flows between the central processor and FPGAs
Previously, to compress data without loss using duplication, they had to be read twice - before and after determining data duplication. As a result, resource consumption increased, and the speed of systems decreased. The new technology optimizes the use of computing resources by transferring computing load to FPGAs.
Method for reducing resource consumption between CPU and FPGAs
Fujitsu specialists estimated the transfer speed of large volumes of data generated by the business. In a test environment, servers with FPGAs were connected using 10-gigabit channels. In the test simulating data backup, a transmission speed of 40 Gbit / s is fixed. This is a new industry record. As a result, the speed of the system increased by about 30 times compared with a system equipped only with a processor. The new technology will make it possible to create a new generation of cloud services that share large amounts of data from several companies from several geographically dispersed data centers.
Fujitsu plans to implement this technology in the form of an application uploaded to servers with FPGAs. First of all, it can be used in cloud environments. The company's laboratory will continue testing the new development. The commercial use of the technology is scheduled for mid-2018.
Using WAN acceleration technology in a cloud environment
Today, most often, data in a WAN is transmitted between cloud environments at speeds from 1 to 10 Gb / s. However, for the proper functioning of modern IT solutions, including the Internet of things and artificial intelligence, an even higher data transfer rate is required. Existing WAN acceleration technologies increase the effective transmission speed by reducing the amount of data transmitted by compression or deduplication.
However, in 10-gigabit WANs, the total amount of information is already reaching its maximum values. Existing acceleration technologies in cloud servers cannot seriously increase the data transfer rate, which becomes the bottleneck of the entire cloud infrastructure. In order to bring the speed of cloud services to a whole new level, you must either use processors with a higher clock speed, or use WAN acceleration technologies that work using different algorithms.
How does the new technology work?
Fujitsu has developed WAN acceleration technology that enables data transmission in cloud environments (in real time) at speeds up to 40 Gb / s. This speed is achieved with the simultaneous use of several 10-gigabit channels. At the same time, Field-Programmable Gate Array (FPGA) matrices installed on the server act as accelerators. The technology is designed to help transfer large amounts of data between cloud environments. The implementation of the new development was made possible thanks to the allocation of a special computing node for processing compression operations. It provides highly parallel operation of the main computing nodes, transmitting data at precisely the specified time, based on the predicted completion of each stage of the calculation.
Below are the features of the technology.
1. FPGA matrix parallelization technology uses dedicated computing nodes
Fujitsu has developed FPGA matrix parallelization technology, which can significantly reduce the processing time required for data compression and deduplication. The result is achieved by deploying dedicated computing nodes and ensuring a high degree of parallelization of computing.
Implementing WAN acceleration using servers equipped with FPGAs
2. Optimizing the distribution of data processing flows between the central processor and FPGAs
Previously, to compress data without loss using duplication, they had to be read twice - before and after determining data duplication. As a result, resource consumption increased, and the speed of systems decreased. The new technology optimizes the use of computing resources by transferring computing load to FPGAs.
Method for reducing resource consumption between CPU and FPGAs
results
Fujitsu specialists estimated the transfer speed of large volumes of data generated by the business. In a test environment, servers with FPGAs were connected using 10-gigabit channels. In the test simulating data backup, a transmission speed of 40 Gbit / s is fixed. This is a new industry record. As a result, the speed of the system increased by about 30 times compared with a system equipped only with a processor. The new technology will make it possible to create a new generation of cloud services that share large amounts of data from several companies from several geographically dispersed data centers.
Fujitsu plans to implement this technology in the form of an application uploaded to servers with FPGAs. First of all, it can be used in cloud environments. The company's laboratory will continue testing the new development. The commercial use of the technology is scheduled for mid-2018.