How do I simplify my
network infrastructure
and eliminate the
redundant elements?
Part of the Choose Your Own Infrastructure eBook series.
Cable sprawl. Not enough throughput. The headaches of
managing separate networks. Soaring power and cooling costs.
Redundancy in switches, adapters, and cables.
This is the reality for most data centers today. For years,
businesses have been creating a series of separate LAN and
storage area networks (SAN) as they react to new demands
and business situations. Data centers are facing a new set
of external demands such as global availability, regulatory
compliance, expectations of the empowered user, and new
metrics such as energy efficiency have been added to the
mix. The result can be a complex, expensive data center
infrastructure that is a huge challenge to manage.
So how do you eliminate all these redundant elements
and simplify the management of your network?
The answer lies in merging LAN and SAN (Ethernet and Fibre
Channel) traffic all in one unified cable, and then consolidating
servers and applications through virtualization. We’ve created
this eBook to help you start to understand the challenges and
choices you’ll face when looking at solutions for simplifying
your data center.
How do I simplify my network infrastructure and eliminate the redundant elements?
1
TOP
1
2
3
4
5
What you are
faced with.
The challenges
that you face
Companies are being forced to continually to expand and invest in
their IT infrastructure, as well as making their core functions more IT
dependent, because they simply don’t have a choice if they want
to stay competitive. There are a whole host of reasons for this, from
providing borderless “anytime, anywhere” access that users demand
to improving customer service and throughput. And in theory, the
guiding principle when adding IT elements to address these business
drivers is to create a scalable and sustainable data center.
We’ve mentioned a few of the challenges facing IT professionals
when it comes to managing specialized, parallel networks, but let’s
take a closer look at the specifics, because understanding these
challenges is critical when it comes to making the right choices for
your infrastructure.
However, the traditional “piecemeal” approach to building data
center infrastructure is becoming a barrier to running efficient data
center. Typically the approach yields low asset utilization, poor
energy efficiency, and increasing complexity. This helps to exacerbate
a situation where companies are running out of space, power and
cooling capacity, budget, and qualified staff. This traditional approach
also poses major challenges in terms of availability and agility (think
throughput for next gen apps, time to provision, and so on). At the
physical level, it creates big redundancies in terms adapters, and
switches, not to mention the harmful cabling mess that results from
connecting all the disparate elements.
Simply put, having a variety of different networks creates a range
of problems that taken together add up to an inefficient data center
and an expensive management challenge.
In light of this, it has emerged that best approach to building a
scalable, sustainable data center that supports business goals is to
consolidate and virtualize servers, storage, and network resources.
Operating Costs
The time spent on management in these environments can be
enormous. Just imagine the resources spent in an increased server
environment configuring ports, cables, multiple host bus adapters
(HBA’s) and network interface cards (NIC’s) in response to increased
demand and new business functions. There is also considerable
capital expense associated with consistently adding new network
equipment, not to mention the number of modules, adapters,
switches and cables that are required to pull it all together.
Cable Sprawl
A virtualized network environment requires more NIC’s and cables
per physical machine for multiple redundancies over copper, as well
as network connections. This means you’ve got cable everywhere,
and it creates a highly complex and expensive environment. And
let’s not forget that cables can create air dams in the data center
floor that act prevent cool air from circulating, leading to increased
heating problems and HVAC costs.
Throughput
Today, the typical data center is populated with servers that are
connected to the network with gigabit ethernet NIC’s and separate
fibre channel HBA’s. When you move to the virtualized environment
necessary to provide the necessary bandwidth for today’s load’s
and applications, a single gigabit ethernet link can quickly
become overwhelmed. From a networking perspective, virtualized
environments typically necessitate a transition to 10 gigabit
ethernet as the default mechanism for attaching servers and
soon to be 40 gigabit ethernet.
Management
It is in inescapable truth that in a traditional, piecemeal siloed
environment, the variety of devices and their different operating
systems can create a huge management headache. Just managing
the SAN can require a full-time employee. These kinds of
environments also create big training issues when it comes to new
IT employees, and makes cross-training and collaboration between
storage and network teams extremely difficult. All of this adds up to
a chunky, slow and expensive infrastructure management process.
The ultimate goal is to have the Server, LAN, and SAN teams to
work together to simplify and reduce the duplication of efforts in
managing the data center.
Availability & Agility
In a typical data center, because of the complexity of the
environment, downtime has to be scheduled for upgrades and
changes, and that just isn’t practical for most of today’s organizations.
Usually, introducing a new application or service, it takes 90 minutes
(at least) to provision a physical server, and you have to provision
the server, storage and networking connection separately. It isn’t
difficult to see how the complexity factor runs counter to today’s
continuity expectations.
Energy Consumption
With all of the equipment required to make the typical data center
run, the energy costs can be enormous. Not only do all the various
servers require lots of power to run, there is also a significant cost
incurred in cooling all this equipment.
How do I simplify my network infrastructure and eliminate the redundant elements?
2
TOP
1
2
3
4
5
Convergence: leading the way
towards a Cisco FCoE Solution
Obviously, there is no silver bullet when it comes to solving
the typical data center challenges, but one of the best guiding
concepts is convergence. You need to merge your data center
server, storage, and network resources to increase asset, energy,
and budget efficiency while eliminating redundancies and
unnecessary management complexity.
And when it comes to enabling convergence, there is one
solution that is making huge inroads helping organizations
simplify their data center infrastructure: Fibre Channel over
Ethernet (FCoE).
I/O Consolidation in the data center, allowing fibre channel and
ethernet networks to share a single Converged Network Adapter
(CNA) in a server, connected to Cisco Nexus switch allows better
efficiency of resources and energy in an integrated unified data
center. An important pillar of this consolidated approach and a
need for a Unified Fabric is Fibre Channel over Ethernet (FCoE).
FCoE enhances server virtualization initiatives with the
availability of standard server I/O, which supports the LAN
and all forms of ethernet-based storage networking, eliminating
specialized networks from the data center. FCoE is an industry
standard developed by the same standards body that creates
and maintains all Fibre Channel standards. FCoE is specified
under INCITS as FC-BB-5.
How it works: A deeper look
Let’s take a closer look at some of the specific applications
of FCoE and convergence solutions at work with designing a
network with designing a network hierarchy with Cisco Nexus
Family of switches Figure 1 hierarchy with Cisco Nexus Family of
switches, and how that translates into some tangible benefits for
any organization.
What is Fiber Channel
over Ethernet?
Fibre channel over ethernet (FCoE) is the next evolution of
ethernet and fibre channel networking where LAN and SAN
traffic are merged onto a one link is called unified fabric. Unified
fabric reduces the number of adapters, cables and devices,
resulting in savings that can extend the life of the data center.
Figure 1
Converging the Access Layer
The first logical step in reducing the number of host adapters
and discrete switching platforms is to converge the access-layer
LAN and SAN switches into a single set of Converged Network
Adapter (CNA’s) host adapters and single switching platform.
Access-layer convergence significantly reduces capital expenditures
(CapEx) and operating expenses (OpEx), and extending convergence
beyond the access layer increases those benefits. By simply adding
FCoE-capable line cards in the Cisco Nexus 5000 Series it provides
unified port functionality that combines 1G or 10G Ethernet or any
iSCSI storage , including layer 2 and layer 3 traffic, and Fibre Channel
storage traffic all in one common data center platform, where the
need for dedicated Fibre Channel hardware is reduced. This approach
largely benefits IT administrators as they manage to reap the benefits
of convergence while using the same management tools that they
use on the LAN to monitor their mission-critical applications in the
SAN. Leverage a Nexus 2000 Series as a Fabric Extender (FEX) or
external I/O modules top of rack (ToR) device uplinked to Nexus
5000 Series End of Row (EoR) switch that brings 10G connectivity
to all your server ports in a rack, reducing the cost of cabling in
the cabinets, provide better air circulation, and a single point of
management. Replace the VMware virtual switches that have limited
capabilities with virtual Nexus 1000v switches in VSphere that ensures
secure network connectivity for all the VM’s in the data center. With
Nexus 1000V VN-LINK, Cisco enhances the VM networking abilities
and helps vMotion since the port configurations move with hosts and
network policies are preserved after vMotion.
Cisco also provides scalability for Fibre Channel traffic to a dedicated
MDS 9100, 9200 Fibre Channel SAN switches or MDS 9500 Multilayer
Directors Switches for more stringent requirements for large data
center environments.
How do I simplify my network infrastructure and eliminate the redundant elements?
3
TOP
1
2
3
4
5
Converging the Distribution/Aggregation/Core Layer
Nexus 7000 Series switch comprise Cisco flagship modular switching
platform for data center Cisco Unified Fabric deployment. The provide
an end-to-end solution for data center core, aggregation, and high
density EoR and ToR server connectivity in a single platform and
designed for 10G, 40G, and 100G scalability to more than 15 Tbps of
throughput. Cisco Nexus 7000 series in conjunction with Nexus 2000
Series FEX scales up to 1,500 ports per Nexus 7000 with a single
point of management , significantly simplifying management as
illustrated in Figure 2
Converged Cisco Nexus Network Design
Additionally, by delivering a Unified Fabric (i.e., consolidating LAN,
SAN, and server clustering traffic) the Cisco Nexus 5000 Series can
eliminate redundant infrastructure (interfaces, cables, switches, etc.)
and simplify operations.
Simplified management, increased productivity
With specific features embedded in the hardware, operating system,
and network management facilities, the Cisco Nexus Family provides
component- and system-level operational continuity. Increased
network stability and fewer service disruptions mean that your
employees have access to the resources they need, when they need
them, and your operations staff can handle problems in the most
efficient way possible.
Increased business agility
IT can dynamically respond to changing business demands by the
rapid provisioning of application and infrastructure services from
shared pools of consolidated computing, storage, and network
resources. Data center consolidation helps streamline operational
procedures and enable policy-based control.
Figure 2
The Cisco Nexus Family Benefits
The flexible, responsive data center infrastructure delivered with the
Cisco Nexus Family can help customers align IT assets and business
priorities by providing the following benefits:
Lower cost of operation, elimination of redundancies
IT can significantly reduce capital costs through consolidation
and increased utilization of previously separate resources, reduce
operational expenses through simplification and automation of
administrative tasks, and reduce cooling and power costs through
improved utilization and an inherently more efficient system design.
Enhanced business resilience
IT can respond to disruptions or attacks by protecting infrastructure,
applications, and data with the highest levels of availability, pervasive security, and business continuance. The Cisco Nexus 7000 Series
is designed with a zero-service-loss architecture, so capacity and
capability can be added without any downtime.
Improved Throughput, Optimized service levels
IT can optimize business service levels by providing secure, accelerated access to data center-hosted applications and information from
anywhere, at any time, across the organization.
Investment protection and efficient use of capital
Cisco Data Center 3.0 emphasizes investment protection and the
incremental incorporation of new technologies and infrastructure, so
investment is more closely aligned with business need. The Cisco
Nexus Family is designed to support the entire lifecycle of the nextgeneration data center. The Cisco Nexus Family will allow capacity
enhancements to be added in a granular, cost-effective manner.
Technical Nexus Benefits in the Data Center
Nexus-OS …Cisco keeps making enhancements
• Ability to perform internal packet captures
• Data plane and control plane separation
• Checkpoints and rollbacks
• Issue commands from anywhere
• Capture to text and append to configs
• SNMP + XML management
• Can be paired with DCNM and/or Fabric Manager
• Many other improvements over I/O
So Long Spanning Tree!
• Eliminating loops in the data center costs bandwidth
• Several technologies offered with Nexus are changing the way
topologies are used, not sacrificing bandwidth and redundancy
• Virtual PortChannels (vPC)
• Fabric Shortest Path First
• TRILL*
• Cisco Fabric Path
*Transparent Redundant Interconnection of a Lot of Links.
>> See Google Tech Talk “Routing without tear; Bridging without
danger” on YouTube
How do I simplify my network infrastructure and eliminate the redundant elements?
4
TOP
Case
Study
1
2
3
4
5
Hospital Reduces Data Center Costs
through Comprehensive Upgrade
Challenge
Salem Hospital is a regional medical center, offering medical services
and technology to people in and around Oregon’s Willamette Valley. A
not-for profit hospital, it is one of the largest of Oregon’s 57 acute care
hospitals and operates the state’s busiest emergency department.
The hospital needed to upgrade its main data center to help ensure
business continuity, reduce energy consumption, and accommodate
larger data volume from electronic medical records (EMR) and the
picture archiving and communications system (PACS).
“We could meet all of those needs by virtualizing our data center
servers,” says Ken Kudla, chief information officer for Salem Hospital.
“First, we would need 10 Gigabit Ethernet.”
Solutions
The Salem Hospital IT department built Cisco® Data Center 3.0
architecture, working with Phoenix Health Systems for strategic
planning, and collaborating on implementation with NetXperts, a Cisco
Healthcare Select Partner with over 12 years of experience in the
healthcare industry.
“We proposed the Cisco Nexus family because it provides the
throughput and scalability needed for PACS, while not affecting the
availability of EMR and other applications critical to patient care,”
says Dave Norinsky of NetXperts. “The NX-OS provides a common,
stable, and manageable operating system for both the Nexus and
MDS platforms, minimizing lifetime support costs.”
One pair of Cisco Nexus 5020 Switches supports Fibre Channel over
Ethernet (FCoE). They connect over lossless 10 Gigabit Ethernet to the
hospital’s newer servers, which host multiple healthcare applications
and Microsoft Exchange, which is scheduled for implementation in
mid-2010. The Cisco Nexus 5020 Switches send storage traffic to a
Cisco MDS 9500 Multilayer Director and data traffic to the Cisco Nexus
7000 Switch at the core (Figure 1).
Another pair of Cisco Nexus 5020 Switches connects to existing
Gigabit Ethernet servers, by way of Cisco Nexus 2000 Fabric Extenders.
The main benefits of the Cisco Data Center Solutions for Salem
Hospital include:
• Lower capital costs: The Salem Hospital campus includes multiple
buildings. Rather than purchasing additional switches for the
distribution layer, the IT group used the virtual device context (VDC)
feature of the Cisco Nexus 7000 Switch to create a separate virtual
switch that acts as the distribution-layer switch.
• Lower cabling costs: FCoE support in the Cisco Nexus 5020 Switch
supports a unified fabric for server data and storage traffic, reducing
the number of server cables from four to two. What’s more, the
Cisco Nexus 5020 Switches are deployed close enough to the
servers to use relatively inexpensive Twinax cabling.
• Support for new generation of healthcare applications: With 10
Gigabit Ethernet and a unified fabric, Salem Hospital can confidently
expand its PACS and EMR programs. The EMR database grew to
60 gigabytes in the first 18 months. “We can also fully utilize the
faster input/output operations per second for our servers,” says
Kudla.
• Simplified management: All Cisco Nexus switches use the same
operating system, NX-OS, minimizing training requirements. And all
Cisco Nexus 2000 Fabric Extenders are managed through a single
Cisco Nexus 5020 Switch.
• Streamlined IT organization: The hospital IT department uses Cisco
Fabric Manager Software to manage both the Cisco Nexus 5020 and
Cisco MDS 9500 Series Multilayer Director, and saves 3 to 4 hours
daily to manage the SAN. “Our networking and storage teams
can now cross-train and collaborate more effectively,” says Kudla.
“We no longer need to dedicate a full-time employee to SAN
management.”
• Increased availability: The IT department can upgrade the Cisco
Nexus switches without interrupting hospital applications because
of the in-service software upgrades (ISSU) feature.
Scheduling downtime for upgrades in a 24-hour healthcare
environment is not practical, and the ISSU feature helps to ensure
business continuity.
• Reduced energy consumption: Before the data center upgrade,
Salem Hospital was using more than 80 percent of available
energy, the maximum recommended. “We have reduced our
energy consumption while also increasing reliability and data
throughput compared to our previous solution,” says Kudla.
• Increased agility: When the hospital introduces a new IT service,
the application team can provision a virtual server in just 30
minutes, compared to 90 minutes for a physical server. Rather
than separately provisioning a server, storage, and networking
connection, the team simply provisions the virtual machine (VM),
and then everything else happens automatically. Up to 50 VMs
reside on each server, reducing equipment and operational costs.
Salem Hospital plans to implement the Cisco Unified Computing
System and begin using Cisco Nexus 1000V, a software switch on a
server. The Cisco Nexus 1000V will give the hospital IT department the
flexibility to move virtual machines across the server farm, preserving
their settings. The network administrator will manage the Cisco Nexus
1000V switch alongside the other Cisco Nexus switches, offloading the
burden from the server administrator.
“We have reduced our energy consumption while also
increasing reliability and data throughput compared to our
previous solution.”
— Ken Kudla, Chief Information Officer, Salem Hospital
How do I simplify my network infrastructure and eliminate the redundant elements?
5
TOP
1
2
3
4
5
Softchoice
can help
When it comes to converged infrastructure we offer the resources
and technical expertise to realize the benefits of the latest in this
innovation. Our difference? An assessment-led approach aimed
at providing a clear understanding of the risks, opportunities, and
payback period for a given solution. From there we offer deep
expertise in solution design and delivery, including technology
architecture, project management and comprehensive implementation support. And behind this dedicated team our licensing and
product specialist work to ensure the technology itself is acquired
in the most efficient and cost-effective manner possible. At every
step our goal is to enhance performance and speed time to market while lowering overall business costs.
Contact our networking experts at [email protected] to learn
more about consolidating your network.
>> Read about Softchoice’s Network Assessment
Know your network:
Softchoice’s Network
Discovery Assessment
Knowing the status of your network assets can greatly improve business up-time and help you anticipate issues long before they occur. That’s
why we created Softchoice’s Network Discovery Assessment – an end-to-end process designed to give you practical, actionable insights into
the most complicated networking challenges.
Issue Identification
MAINTENANCE
Coverage & Contracts
Uncovered Devices
Orphaned Contracts
NETWORK
DISCOVERY
Device Inventory
HARDWARE
End-of-Life/ End-of-Sale
Support Status
IOS Issues
CONFIGURATION
PSIRT Violations
Configuration Errors
STEP 1
STEP 2
Prescription & Treatment
MAINTENANCE
COVERAGE & CO-TERM
CONFIGURATION
RECOMMENDATIONS
CONFIGURATION
REMEDIATION
STEP 3
Step 1: Discovery
It all begins with inventory. Either by leveraging your existing network management investments or by deploying one of several network discovery
options, Softchoice collects, organizes and presents a digestible overview of your network from top to bottom.
Step 2: Analysis & issue identification
Next we uncover the status of your maintenance coverage, the lifecycle of network devices and any other equipment and configuration issues.
We focus on the areas that require attention without subjecting you to “analysis paralysis.” With Softchoice Axis, our online IT Asset Management
portal, understanding the complexities of your network has never been easier.
Step 3: Recommend and act
Once we identify the issues affecting your network, the next step is to act on the findings. Whether you want to do this yourself, or you’re looking
for help from an on-site resource, Softchoice can help. From easing maintenance management, charting equipment replacement strategies or
optimizing your network, Softchoice can design, deliver and even deploy the right solution for you.
Discover your options
The first step is speaking with a Softchoice representative to find out if this web-based
assessment is right for you. Get in touch today to request a discovery call with one of our
technology experts.
Connect with us today.
1.800.268.7638 | www.softchoice.com