China
Africa
Explore chapters and articles related to this topic
Storage Acceleration
Published in Heqing Zhu, Data Plane Development Kit (DPDK), 2020
iSCSI (Internet SCSI) is the network storage standard specification developed by IETF. It is used to connect storage devices through an Internet Protocol. By transferring SCSI commands and data over an IP network, iSCSI completes the data transfer between network nodes and enhances remote data management. The excellent data transmission capability of the iSCSI protocol is a key factor driving the SAN market. IP network is everywhere, and iSCSI can enable the data transfer on LANs, WANs, and even Internet, which removes the regional restrictions on storage service. iSCSI depends on TCP/IP network protocol, and SCSI command and data are encapsulated by iSCSI and TCP/IP packets, as shown in Figure 16.2.
The Ethernet Advantage in Networking
Published in James Aweya, Designing Switch/Routers, 2023
High-performance Ethernet switches in conjunction with IP can also provide the high-speed interconnect required for virtualization, inter-process communication (IPC) channels for computational clusters and Grids, and SANs with iSCSI. iSCSI SANs and NAS solutions based on the multi-gigabit Ethernet LAN interconnects are now widely deployed from smaller to large data centers. Ethernet-based storage can be readily assimilated into the existing Ethernet network management environment without requiring additional management tools, or the training necessary for using special-purpose protocols, such as Infiniband and Fibre Channel.
Requirements of Metropolitan Area Network Services
Published in Nam-Kee Tan, MPLS for Metropolitan Area Networks, 2004
ISCSI is a networking protocol for SCSI-3 traffic over TCP/IP Ethernet. This protocol allows block-level storage data to be transported over common IP networks, enabling end users to access the storage network anywhere in the enterprise. ISCSI extends over unlimited distances and can be implemented as a converged storage network in the metro environment through the use of VPNs to support multiple-user data applications within a single network infrastructure.
Towards complex dynamic fog network orchestration using embedded neural switch
Published in International Journal of Computers and Applications, 2021
K. C. Okafor, G. C. Ononiwu, Sam Goundar, V. C. Chijindu, C. C. Udeze
Clearly, large cloud network models as well as Big data systems will continue to have challenges in managing traffic workloads. To address the congestion problem in Figure 1, this paper leverages an increasingly adopted feature in modern HDN integration with its switching fabrics adapted to the Fog switching component (SD-ANN switch). This is based on Fiber channel (over Ethernet or IP, FCoE/IP). The high speed link offers lossless data delivery especially when used to connect data storage interfaces (e.g. internet small computer system interface iSCSI) to servers. In this regard, intelligent QoS implementation creates an opportunity to solve workload congestion problems. With a feedback placement different from the rate based schemes, it is possible to localize majority of the traffic in HDN Top-of-Rack (ToR) switch. The effect is to compensate or balance the outgoing traffic with feedback loop, thus resolving congestion effectively. The proposed neural network switch design (with an intelligent feedback loop) is used for enhanced congestion management in this regard. It is based on dynamic feedback control (DFC) model on a neural network (for the flow management problem) within HDN core switch. In this case, receding horizon plays a significant role in isolated congested reference regions at the Fog layer involving the neural switch.