NonStop NS1000 Planning Guide (H06.08+)

4 Integrity NonStop NS1000 System Description
This section describes the Integrity NonStop NS1000 system. For information about installing
the Integrity NonStop NS1000 server hardware, refer to the NonStop NS1000 Hardware Installation
Manual.
NonStop Architecture
The Integrity NonStop NS1000 server achieves full software fault tolerance by running the
NonStop operating system on HP Integrity rx2620 servers. This includes:
• Software fault tolerance and fault isolation
• Dynamic workload balancing
• Linear scalability
• Application virtualization
• Cluster programming transparency
The Integrity NonStop NS1000 server can be configured with two, four, six, or eight processors,
communicate with other servers using Expand-over-IP, and achieves ServerNet connectivity
using ServerNet PCI adapter cards installed in the HP Integrity rx2620 servers. Every piece of
hardware is duplicated to prevent any single point of failure; in the event of a processor failure,
another processor takes over. The Integrity NonStop NS1000 uses the ServerNet fabric and the
same modular I/O as other Integrity NonStop NS-series servers.
Blade Complex
The basic building block of the NonStop NS1000 system is the blade complex, which consists of
rx2620 servers that are also referred to as blade elements. Each blade element houses one
microprocessor called a processor element (PE).
All input to and output from each blade element occurs through a ServerNet PCI adapter card
located in each blade element. The ServerNet PCI adapter card interfaces with the ServerNet
fabrics via the VIO enclosure.
This diagram provides an overview of the modular NS1000 system and shows four blade elements
with ServerNet connections, two VIO enclosures (the I/O hardware), and links to external I/O
and storage:
NonStop Architecture 37