Cisco Cisco UCS Director 4.0 Information Guide
2 © 2014 Cisco and/or its affiliates. All rights reserved. This document is Cisco Public Information.
— Hiroshi Matsuda
Head of the cloud
architecture department, new
solutions division, enterprise
solutions operations unit
NEC Networks & System
Integration Corporation
Utilizing Cisco Unified Computing System (UCS) B-Series Blade Servers with Intel Xeon
processors, the company was able to create a dynamic infrastructure, managed by Cisco
UCS Director, as the infrastructure foundation to support their cloud environment.
processors, the company was able to create a dynamic infrastructure, managed by Cisco
UCS Director, as the infrastructure foundation to support their cloud environment.
Solution
Streamlined implementation and management processes
Hiroshi Matsuda, head of the cloud architecture department, new solutions division,
and enterprise solutions operations unit, oversaw deploying and managing the shared
infrastructure. Matsuda deployed Cisco UCS B-Series Blade Servers because they
are designed for large-scale virtualization—a feature that set the servers apart from
the competition. When combined with Cisco UCS Director, the company was able to
dramatically shorten the time to deploy and manage new large-scale environments.
Hiroshi Matsuda, head of the cloud architecture department, new solutions division,
and enterprise solutions operations unit, oversaw deploying and managing the shared
infrastructure. Matsuda deployed Cisco UCS B-Series Blade Servers because they
are designed for large-scale virtualization—a feature that set the servers apart from
the competition. When combined with Cisco UCS Director, the company was able to
dramatically shorten the time to deploy and manage new large-scale environments.
Unlike traditional blade servers, the Cisco UCS B-Series Servers significantly streamline
server management as well as energy consumption. NESIC had not used blade servers in
the past because of their tendency to increase the number of Ethernet and fibre channel
switches required for each blade server chassis. According to Matsuda, Cisco UCS
B-Series Blade Servers do not require any additional switches but rather utilize a fiber
interconnect for the Ethernet and fibre channel connections which greatly increases
efficiency while lowering the workload on server teams.
server management as well as energy consumption. NESIC had not used blade servers in
the past because of their tendency to increase the number of Ethernet and fibre channel
switches required for each blade server chassis. According to Matsuda, Cisco UCS
B-Series Blade Servers do not require any additional switches but rather utilize a fiber
interconnect for the Ethernet and fibre channel connections which greatly increases
efficiency while lowering the workload on server teams.
NESIC also deployed the Cisco UCS B-Series Servers for their unique service profiles.
Normally, IT staff would perform unique settings on each server to identify the hardware
and enable staff to manage and operate multiple servers from a single location. With
the combination of Cisco UCS Director and UCS Manager, NESIC is able to manage
all servers, and their fibre interconnects, from a single pane of glass—greatly reducing
management complexity.
Normally, IT staff would perform unique settings on each server to identify the hardware
and enable staff to manage and operate multiple servers from a single location. With
the combination of Cisco UCS Director and UCS Manager, NESIC is able to manage
all servers, and their fibre interconnects, from a single pane of glass—greatly reducing
management complexity.
“For example, if we need to add eight servers to the system, normally we would need
to upgrade the firmware and set the BIOS settings for each individual server,” says Matsuda.
“To connect to the SAN storage through fibre channel, we needed to check the World
Wide Name (WWN) of the host bus adapter (HBA) before setting the fibre channel settings.
Deploying a server could be a complex and time-consuming process with lots of room
for error.”
to upgrade the firmware and set the BIOS settings for each individual server,” says Matsuda.
“To connect to the SAN storage through fibre channel, we needed to check the World
Wide Name (WWN) of the host bus adapter (HBA) before setting the fibre channel settings.
Deploying a server could be a complex and time-consuming process with lots of room
for error.”
“On the other hand, with Cisco UCS B-Series Blade Servers, we can create service
profiles before we even receive the physical servers,” says Matsuda. “We simply run
the service profiles once the equipment has arrived, and the servers are immediately
ready for installation.”
profiles before we even receive the physical servers,” says Matsuda. “We simply run
the service profiles once the equipment has arrived, and the servers are immediately
ready for installation.”
Working with Cisco UCS service profiles and UCS Director, NESIC has reduced the time
needed to deploy servers by 50%. Since equipment can be deployed quickly and when
it is needed, the company doesn’t need to purchase and install equipment to meet future
demands. As a result, the company can reduce costs and idle servers.
needed to deploy servers by 50%. Since equipment can be deployed quickly and when
it is needed, the company doesn’t need to purchase and install equipment to meet future
demands. As a result, the company can reduce costs and idle servers.
The network connection between each server chassis and the fabric interconnect achieves
Ethernet speeds of at least 10 Gbs, with two lines used for redundancy. Since the fabric
interconnect uses Fibre Channel over Ethernet (FCoE) for fibre channel connections, the
IP network, fibre channel, and NFS can all pass through a single cable. Furthermore, by
distributing the 10 Gbps bandwidth logically and partitioning the bandwidth, NESIC can
identify issues in the virtual environment more easily and achieve stable operations even
within a complex network environment.
Ethernet speeds of at least 10 Gbs, with two lines used for redundancy. Since the fabric
interconnect uses Fibre Channel over Ethernet (FCoE) for fibre channel connections, the
IP network, fibre channel, and NFS can all pass through a single cable. Furthermore, by
distributing the 10 Gbps bandwidth logically and partitioning the bandwidth, NESIC can
identify issues in the virtual environment more easily and achieve stable operations even
within a complex network environment.