NonStop NS14000 Series Planning Guide (H06.13+)

In the event of a processor fault in either a duplex or triplex processor, the failed component within
a NonStop Blade Element (processor element, power supply, and so forth) or the entire Blade
Element can be replaced while the system continues to run. A single NonStop NS14000 series
system can have up to two NonStop Blade Complexes for a total of eight processors. Processors
communicate with each other and with the system I/O over dual ServerNet fabrics.
In the term ServerNet fabric, the word fabric is significant because it contrasts with the concept of
a bus. A bus provides a single, fixed communications path between start and end points. A fabric
is a complex web of links between electronic routers that provide a large number of possible paths
from one point to another. Two communications fabrics, the X and Y ServerNet, provide redundant,
fault-tolerant communications pathways. If a hardware fault occurs on one of the ServerNet fabrics,
communication continues on the other with hardware fault recovery transparent to all but the lowest
level of the OS.
Input/output components tap into the ServerNet fabrics through Fibre Channel and Ethernet
ServerNet functionality in the VIO enclosures in the preceding diagram of the NonStop NS14000
series system. The Fibre Channel and Ethernet functionality in the VIO enclosures provides the
system I/O to Fibre Channel storage devices and gigabit Ethernet communications networks.
Although the hardware architecture differs from one series of NonStop servers to another, the
NonStop NS14000 series system can be networked with other NonStop systems using the same
message system and the same network software.
104 NonStop NS14000 Series System Architecture