The back end layer of our Web and e-mail hosting and e-commerce services consists of Network Solutions' clustered hosting solution, as well as third-party applications integrated into our product, such as domain registration. However, Network Solutions Operating Systems (IQOS) treats each backend resource as a black box owned by a third-party vendor. In order to communicate, IQOS creates requests through the applications and services provided by the vendor. This guarantees that a service runs the way the vendor intended it to, and our partners can sell their services with confidence.
Network Solutions' back-end clustered hosting solution contains multiple layers to provide a redundant, scalable solution. We work with premier hardware and software vendors, always choosing names that our private label customers know and trust. In addition, IQOS allows Network Solutions' platform to be managed centrally, so customers are not sharing one server, they are sharing the entire environment. Servers can be specialized to handle specific functions most efficiently, even though all services will appear to be in one place to the end user.
The network layer consists of Cisco and F5 equipment, and provides infrastructure for all TCP/IP traffic.
The Application Layer runs on Hewlett-Packard and IBM Blade servers, and hosts all core services and applications. Each blade is comparable to a server in a traditional environment, but it is smaller and more power efficient. Because the blades sit in enclosures, there is less risk of cabling failures. And if a blade fails, Network Solutions can easily provision another one, using our Linux and Windows images and a fully tested application set, so all services are back online within minutes, with no loss of service.
The data layer provides a RAID-protected data storage environment accessible centrally by the application layer. Network Solutions uses Network Appliance filers and iSCSI technology to host customer content. Storing the data centrally has numerous benefits. If a blade or server fails, we just bring another one up and tell it where to go find the data. It takes only minutes, instead of the hours it used to take, to restore data from a failed server. Servers are all-purpose so it's more efficient to put storage in a specialized place and let the server manage requests, exactly what it does best. Finally, iSCSI makes each hard drive look like one big file. Therefore, it's easy to move or copy that file, for backup and disaster recovery purposes.
Finally the management layer provides a secure, encrypted transport for operations, support and provisioning.