Cisco Cisco UCS B200 M4 Blade Server Informationshandbuch
Network reached limits of performance
Galliker operates the software for its various warehouse and logistics systems in
a mainframe environment, which, like everything else, runs in the company’s own
data center at the headquarters in Altishofen. That activity includes, for example,
office automation, email communication, and file services. “The demands made on
our data center are rapidly increasing,” says André Dousse, IT expert in Galliker’s
data center. “And data traffic is also growing at the same fast rate. Our previous
network just couldn’t cope any more; by early 2010, it had definitely reached its
bandwidth limits.”
a mainframe environment, which, like everything else, runs in the company’s own
data center at the headquarters in Altishofen. That activity includes, for example,
office automation, email communication, and file services. “The demands made on
our data center are rapidly increasing,” says André Dousse, IT expert in Galliker’s
data center. “And data traffic is also growing at the same fast rate. Our previous
network just couldn’t cope any more; by early 2010, it had definitely reached its
bandwidth limits.”
The idea of using powerful Cisco Nexus 7000 Series Switches in the core of
the network was suggested to Galliker by Netcloud AG. Based in Winterthur with
subsidiaries in both Berne and Basle, this highly specialized systems company is
one of the most competent Cisco partner in Switzerland; in 2009, in addition to
gaining the “Cisco Gold Partner of the Year” award, the company also won the title
“Cisco Data Center Partner of the Year.” “Quite apart from its various virtualization
functions, Nexus 7000 was the most cost-efficient long-term option for scaling
the network up to a speed of 10 gigabit,” says Markus Michalek, UCS and data
center product specialist for Netcloud at its Basle site. He went on to say: “Initially,
blade servers were not even mentioned in our discussions; the main subject was
the Nexus 7000. However, as this is not just a high-performance switch, but also
a core component in the Cisco Data Center architecture, the discussions soon
became more general. And then at some point, the question arose as to what
direction Galliker’s data center should develop in as a whole.” And that was how
the Cisco Unified Computing System (UCS) came to be mentioned.
the network was suggested to Galliker by Netcloud AG. Based in Winterthur with
subsidiaries in both Berne and Basle, this highly specialized systems company is
one of the most competent Cisco partner in Switzerland; in 2009, in addition to
gaining the “Cisco Gold Partner of the Year” award, the company also won the title
“Cisco Data Center Partner of the Year.” “Quite apart from its various virtualization
functions, Nexus 7000 was the most cost-efficient long-term option for scaling
the network up to a speed of 10 gigabit,” says Markus Michalek, UCS and data
center product specialist for Netcloud at its Basle site. He went on to say: “Initially,
blade servers were not even mentioned in our discussions; the main subject was
the Nexus 7000. However, as this is not just a high-performance switch, but also
a core component in the Cisco Data Center architecture, the discussions soon
became more general. And then at some point, the question arose as to what
direction Galliker’s data center should develop in as a whole.” And that was how
the Cisco Unified Computing System (UCS) came to be mentioned.
Inspiring architectural concept
“UCS goes far beyond all the other blade systems available on the market,”
Michalek to says. “The solution combines high-performance blade servers, VMware
virtualization, and SAN/LAN connections in one single complete system, which is
uniformly administered. I know of no other data center technology that was developed
so systematically with the focus on standardizing virtualization of different IT worlds.
And it is precisely here, in the standardization of virtual servers, storage, and network,
that the enormous cost benefits of Cisco UCS can be found.”
Michalek to says. “The solution combines high-performance blade servers, VMware
virtualization, and SAN/LAN connections in one single complete system, which is
uniformly administered. I know of no other data center technology that was developed
so systematically with the focus on standardizing virtualization of different IT worlds.
And it is precisely here, in the standardization of virtual servers, storage, and network,
that the enormous cost benefits of Cisco UCS can be found.”
A technology session, which also involved the manufacturer Cisco, was arranged so
that the UCS philosophy could be explained in greater detail to Galliker’s IT team.
Dousse clearly remembers how sceptical he was before attending this workshop:
“After a day and a half of intensive discussions, I was then absolutely won over by
the network-focused approach on which UCS is based,” says Dousse. “We have, of
course, already had server virtualization. But an architectural model that is so universal,
where the software is so far abstracted from the hardware, and at all infrastructure
levels: I was just so enthusiastic.” And anyone who is himself enthusiastic about
something can convince others, for example, the people holding the purse strings in
one’s own company. So it was only a matter of a few weeks from the UCS workshop
until the systems were on order.
that the UCS philosophy could be explained in greater detail to Galliker’s IT team.
Dousse clearly remembers how sceptical he was before attending this workshop:
“After a day and a half of intensive discussions, I was then absolutely won over by
the network-focused approach on which UCS is based,” says Dousse. “We have, of
course, already had server virtualization. But an architectural model that is so universal,
where the software is so far abstracted from the hardware, and at all infrastructure
levels: I was just so enthusiastic.” And anyone who is himself enthusiastic about
something can convince others, for example, the people holding the purse strings in
one’s own company. So it was only a matter of a few weeks from the UCS workshop
until the systems were on order.
Toward Unified Fabric
Cisco UCS turns the concept of what is called Unified Fabric into reality. What this
actually means is that “normal” IP data and block-orientated storage data of all kinds
flow along the same cable. This applies to Network Attached Storage (NAS) as well
as storage protocols such as Small Computer System Interface over IP (iSCSI) and
Fibre Channel. All these protocols are transmitted via a modified form of Ethernet,
also referred to as data center bridging (DCB). With Fibre Channel, it is then called
Fibre Channel over Ethernet (FCoE). The Nexus switching platform is incidentally
also based on the Unified Fabric concept.
actually means is that “normal” IP data and block-orientated storage data of all kinds
flow along the same cable. This applies to Network Attached Storage (NAS) as well
as storage protocols such as Small Computer System Interface over IP (iSCSI) and
Fibre Channel. All these protocols are transmitted via a modified form of Ethernet,
also referred to as data center bridging (DCB). With Fibre Channel, it is then called
Fibre Channel over Ethernet (FCoE). The Nexus switching platform is incidentally
also based on the Unified Fabric concept.
© 2011 Cisco and/or its affiliates. All rights reserved. This document is Cisco Public Information..
Page 2 of 3
Solution Overview
Background
Challenge
Solution
Benefits
Markus Michalek from Cisco partner Netcloud
is familiar with the requirements in the Galliker
data center
is familiar with the requirements in the Galliker
data center