Please note that JavaScript and style sheet are used in this website,
Due to unadaptability of the style sheet with the browser used in your computer, pages may not look as original.
Even in such a case, however, the contents can be used safely.
A major international ICT provider, NTT Communications, is using ProgrammableFlow® technology in this way. Operating data centers around the world providing cloud services, the provider must be ready to reconfigure the resources used by its customers on an almost instant basis, together with delivering a reliable, state-of-the-art service. With ProgrammableFlow® technology, a logical centralized controller incorporating redundancy and robustness can reconfigure a worldwide network to provide a client with a virtual “data center” almost immediately.
ProgrammableFlow® technology brings about considerable savings in capital expenditure and operational expenses (CAPEX and OPEX), as major logistics operator Nippon Express has found, being able to achieve substantial reductions in both. Indeed, NEC estimates that for a 1,000-server data center, both capital and operational expenditure can be slashed by 50% compared to conventional networks. Nippon Express has been able to cut its network switch rack space requirements from 32U to 10U, and thereby power consumption, from 14kW to 2.5kW, and perhaps most dramatically, has been able to bring its network management in-house, with the benefit that outsourcing services were no longer required. Not only have costs been reduced, but the time needed to fail over following a network outage has been cut from nearly one minute to one second.
For Kanazawa University Hospital, ProgrammableFlow® has brought other benefits. The hospital is a constantly changing network environment. In the medical networking field, flexibility of ports and switches is of paramount importance. Equipment used for patient diagnosis and treatment is frequently moved, and the different departments using the equipment have varied network security policies, as do the different devices themselves. Moving or adding such network-connected equipment effectively was previously extremely difficult. Now, through the implementation of ProgrammableFlow®, when a medical device is connected to the LAN, it is assigned to the appropriate virtual network, and the appropriate policies are automatically applied. This occurs regardless of the LAN port to which the device is connected, thus realizing universal connectivity.
Typically, the “refresh cycle” for networking hardware in an enterprise is between three and five years. With an open SDN architecture, networks can be refreshed and their capabilities expanded at any time, with substantial reductions in CAPEX, and the cycle is broken. Furthermore, since the network can be dynamically reconfigured, it is easier for the “new” network configuration to be run, tested and debugged in parallel with the existing network before going live into production status. A software-defined network can be modified at any time there is an organizational change within the enterprise. MACs (Moves, Adds and Changes) applied to such a network need not cause interruptions to the existing network’s traffic flow.
VLANs in a data center using a traditional silo model use “appliance resources” such as firewalls, load balancers and so on inefficiently, and hence there are “bumps” and “steps” in the capital costs incurred through expansion as the number of switches increases and the VLANs hit the maximum device numbers (typically 2,000 or 4,000). In a ProgrammableFlow® implementation, these infrastructure appliances can be shared more easily, meaning that not only is the total capital expenditure reduced as the expansion proceeds, but costs rise in a smooth linear fashion, proportional to the number of switches.