IBM DCN Redbook (Validated Reference Design)
Upcoming SlideShare
Loading in...5
×
 

IBM DCN Redbook (Validated Reference Design)

on

  • 320 views

IBM published in Q2 of calendar 2013, a Redbook, entitled, IBM and Cisco: Together for a World Class Data Center. A team of distinguished engineers from IBM and Cisco collaborated on this extensive ...

IBM published in Q2 of calendar 2013, a Redbook, entitled, IBM and Cisco: Together for a World Class Data Center. A team of distinguished engineers from IBM and Cisco collaborated on this extensive document. The Redbook is analogous to Cisco's Validated Design guidelines and is considered the best practice in designing Data Center Networks.

Statistics

Views

Total Views
320
Views on SlideShare
296
Embed Views
24

Actions

Likes
0
Downloads
7
Comments
0

1 Embed 24

http://datacenternetworkingibm-cisco.ning.com 24

Accessibility

Categories

Upload Details

Uploaded via as Microsoft PowerPoint

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Processing…
Post Comment
Edit your comment
  • In this chapter we address critical network implications, which means we describe specificnetwork issues and requirements in virtualized Data Centers and elaborate on those issuesby providing guidance and exploring possible solutions.The process to develop the network design can no longer be done in isolation. It has toinclude key requirements from the other areas of the Data Center infrastructure includingserver, storage, security and applications. The input can be gathered across six key areas ofarchitectural considerations: Design scope - means developing the central design principles based on present andplanned application workloads, server and storage platform virtualization capabilities, ITservices for internal and external customers, and anticipated schedules for growth andinvestment. Organization - means addressing the existing boundaries between the network, serversand storage to establish better interfaces for operational teams to work effectively invirtualized and dynamic environments; creating a comprehensive design by requiring thenetwork team to develop a greater operational awareness of server and storagevirtualization technologies and capabilities. Technology - means choosing from current and emerging networking technology optionsto make the best decision for executing your IT strategy while minimizing businessdisruption. Management - means deciding on the right level of integration with the IT managementenvironment to both enhance current operations and provide a path to emerging,policy-driven network configuration and provisioning standards. Security - means balancing the introduction of evolving capabilities, such as virtualizationand provisioning, to provide the optimal relationship between business demands and risktolerance. Best practices for security management and integration can create acomprehensive view of the physical and virtualized infrastructure to provide robustsecurity, visibility and segmentation. Automation - means the automated approach to network and security configurations withthe provisioning and orchestration of servers, storage and facilities.The use of virtualized servers implies change in some characteristics of the Data Centernetwork. Beyond much more complex scalability issues due to a much higher number ofservers (now virtualized), there is also the need to have mechanisms to assist the network inadapting dynamically to VM migration between different physical servers, which might beplaced in geographically dispersed Data Centers.Issues in the Layer 3 domain for highly virtualized data center networks also exist and need tobe addressed. For example, due to the larger Layer 2 domains the placement of Layer 3needs to be carefully designed.In this chapter, the following design consideration are addressed in more detail: The Modular Data Center Network Design concept Data Center Fabric Characteristics and Solutions IPv6 and the Data Center Data Center Network Throughput Considerations Multi-Tenancy Virtualized Workload Mobility between Data Centers High Availability and Fast Failover First Hop Redundancy Protocols Network Services - Load Balancing, Firewall, WAN Acceleration and Traffic Analysis IP-based Storage Requirements Regulatory Requirements Consistency between Physical and Virtual Network Differentiated service levels on a shared infrastructure (QoS) Data Center Cabling
  • 3.1.1 The PoD conceptThe introduction of a Point of Delivery (PoD) in relation to Modern Data Center design issomething which can be observed more and more, but the term PoD is often used in differentvariations with different meanings.A PoD is, in general, a physical repeatable pattern, and its components maximize themodularity, scalability, and manageability of Data Centers. The concept of a design pattern isnothing new in computer or network architecture, but the term PoD is often used in differentvariations, especially in combination with cloud services. As the scope of this book is far more8105 03 Requirements for Designing Networks.fm Draft Document for Review May 9, 2013 6:54 am136 IBM and Cisco: Together for a World Class Data Centerthan designing a Data Center network for Cloud services, the definition of a PoD in this bookmight differ to documents which are referencing a virtualized, multi-tenant and fullyautomated Data Center design for specific services like Cloud services or massively scalableData Centers.We define a PoD as a physical, modular building block in a Data Center with the followingcharacteristics: Collection of cabinets with a common set of network equipment Server and network hardware are co-resident in the same PoD Network functionality is typically Layer 2 only in the form of network access/aggregation Cabinets in the PoD are typically contiguous but are not limited to this due to special sitecharacteristics All server NIC to network port connections stay within the PoD to optimize cabling All PoDs are connected to a Data Center Network CoreThe characteristics for a PoD defined above are the common ones. It is recommended todefine further possible and more granular characteristics, but this should be defined on acase by case basis depending on the Data Center specifics. The following ones should beconsidered additionally: Use of Top of Rack (ToR), Middle of Rack (MoR) or Bottom of Rack (BoR) installation ofnetwork devices Use of low cost copper cables for 10GE server ports PoD internal placement of storage and storage network devices Use and placement of cable management patch panels Rack power and weight requirements Network device airflow should comply with Data Center hot/cold isle requirements Server types and connectivity in regards to the number of physical 1GE and 10GEinterfaces to allow an efficient use of network resourcesThe following examples show different types of PoDs for illustration. All have the samecommon characteristics in slightly different deployment scenarios. For simplification, thestorage showed in the pictures is utilizing Ethernet, but no native Fiber Channel.The first example, in Figure 3-1 on page 137, shows a PoD in which the storage devices areimplemented in each PoD. Four physical network devices are installed ToR to provideredundant 10GE and 1GE ports for the servers.The benefits of this approach are: Servers and storage can be expanded by a rack Cables and storage traffic are closed in the PoDThe drawbacks are: May not make efficient use of storage Distributed storage (data store) management
  • Server virtualization and the use of virtual machines is profoundly changing data centerdynamics. Most organizations are struggling with the cost and complexity of hosting multiplephysical servers in their data centers. The expansion of the data center, a result of bothscale-out server architectures and traditional “one application, one server” sprawl, hascreated problems in housing, powering, and cooling large numbers of under-utilized servers.In addition, IT organizations continue to deal with the traditional cost and operationalchallenges of matching server resources to organizational needs that seem ever changing.Virtual machines can significantly mitigate many of these challenges by enabling multipleapplication and operating system environments to be hosted on a single physical server whilemaintaining complete isolation between the guest operating systems and their respectiveapplications. Hence, server virtualization facilitates server consolidation by enablingorganizations to exchange a number of under-utilized servers for a single highly utilizedserver running multiple virtual machines.
  • Data center design is an act of constant planning. The core infrastructure design, power andcooling, cabling, and location must be carefully analyzed prior to installation and reviewed inan ongoing basis. The services in a data center are often seen as add-ons that are simplyconnected when needed. The concept of add-on services is a good one and in fact servicesshould be able to be added as requirements change and the data center grows. But whenthese devices are in the traffic path for critical applications and data, careful thought must begiven to how they connect, where they connect, how they scale, and how they handle failure.Security is often considered as an after-thought in many architecture designs. In reality, it iseasier in the long run if security is considered as part of the core requirements—and not asan add-on. But this depends on several business-related factors and the drivers for security inthe data center might change over time due to new application rollouts, compliancerequirements, acquisitions, and—not the least of which—security breeches.This chapter will highlight security solutions that are tailored to meet the securityrequirements of a modern data center environment.
  • In this chapter we will describe the possibilities and benefits of Converged networking. TheLAN and SAN environments have existed independently for a long time, but in recent yearswe have seen significant cost and performance benefits by mixing these two types ofnetworks together on a single infrastructure. We will describe the architecture, functionalityand benefits of using Fibre Channel over Ethernet (FCoE) to converge the LAN and SANtogether.It is beyond the scope of this chapter to cover attaching BladeCenters or IBM PureFlex andIBM Flex systems to the network. Refer to Chapter 2, “IBM and Cisco Building Blocks for theData Center” on page 29 and Chapter 4, “Data Center Physical Access Layer Evolution” onpage 193 for reference information on BladeCenter or Pure Flex, or consult thehttp://www.ibm.com web site.We will explore topics such as: Converged - what does that really mean? How is FCoE different from FC and iSCSI? Data Center Bridging Business drivers for FCoEFigure 7-20 shows FCoE at the Access Layer with Nexus 5x00 and FC uplink to the SAN. Theservers connect to the Nexus 5x00 using their CNA's. If we consider Server 1, the FCoEtraffic from the port of the CNA that is going to Fabric A (left link from the server) istransported on VLAN 20, marked in Red. The FCoE traffic from the port of the CNA that isgoing to Fabric B (right link from the server) is transported on VLAN 30, marked in Red aswell. VLAN 20 is only present on the left link between the server 1 and the Nexus 5x00 on theleft. VLAN 30 is only present on the right link between server 1 and the Nexus 5x00 on theright. This way Fabric A and Fabric B are completely separated. The Ethernet traffic istransported on VLAN 10 in this example, and is present on both links from the server to theNexus 5x00.
  • This chapter introduces the main challenges Network architects have to face when trying todesign and implement networks in a Data Center environment and elaborates on the Ciscosolutions that can be leveraged to overcome these shortcomings. First we will introduce the traditional three tiered data center network model and explainwhy traditional Ethernet networks cannot meet the requirements Modern Data Centersbring in terms of efficiency, flexibility, scalability and manageability. After explaining the reasons why Ethernet networks have to evolve and where the driversfor change come from we will explore the available Cisco solutions that can be leveragedto overcome the shortcomings we have introduced at the beginning of the chapter. These new Cisco data center networking solutions possess different degrees of maturityand are positioned differently in terms of use cases and scope. We will explore thesesolutions in detail in this chapter, also highlighting what type of standard based solutionscan be used and elaborating on the key differences. The solutions are divided in:– Multi-chassisEtherchannel techniques: Port Channels, virtual Port Channels,Enhanced virtual Port Channels, virtual Port Channel +.– Data Center Network Fabrics: Cisco FabricPath and TRILL– Data Center Network overlays techniques: OTV, VPLS and LISP.
  • In this chapter we introduce the NX-OS Network Operating System and the ways it enablesus to build world class data centers. NX-OS was built with self-healing, modularity andscalability in mind. NX-OS is the operating system for Cisco’s Nexus line of products, as wellas Cisco’s MDS product line. Since NX-OS provides administrators with a consistentoperating system for both network and storage devices it allows them to simplify the designand operation of the entire data center fabric.
  • Today’s data center network infrastructure is evolving from static to dynamic and technologiessuch as cloud combined with highly virtualization and distributed workload applications aredominating the network backbone infrastructure. Aside, threats due to malicious attachmentsare increasing and both businesses and consumers are looking at service providers toincorporate more defenses into the network. All these factors are pushing organizationstowards a complex network infrastructure and as the data center environment becomes moredistributed, complexity increases exponentially and manageability becomes an even greaterchallenge.A converged, dynamic infrastructure can help you address today's challenges and tomorrow'sopportunities. Smarter data center solutions from IBM and Cisco can increase resourceutilization, support business continuity, provide higher availability and faster provisioning, andhelp you achieve productivity gains through virtualization, optimization, energy stewardshipand flexible sourcing. IBM and Cisco together can help you to plan and deploy a responsivenetworking infrastructure that adapts quickly, cost effectively and securely to rapidly changingdemands.In this chapter we discuss how recent efficient, standards based innovations in data centernetworking can meet the agile requirements, such as security and performance, of the nextgeneration data center. We describe how improved economics of IT can be achieved byconnecting servers and storage with a high-speed and intelligent network fabric that issmarter, faster, greener with more simplified management. We also, describe some of theproducts available from both IBM and Cisco that will lead to better management andautomation, and how they work together to provide an effective, all-encompassing solutionthat harnesses the creative and intellectual abilities of both companies. In addition we brieflyintroduce some of the important considerations for comprehensive data center management.
  • Software Defined Networking aims at enhancing the programmability of the networkinginfrastructure so it can be easily consumed by application and IT services developers acrossdifferent networking areas, vendors and implementations without requiring custom softwaredevelopment and usage of scripting languages.While networks have supported significant innovations in compute and storage, networkingtechnologies have not tracked the expanding levels of virtualization and programmability ofthese technologies. Networks thus are increasing in complexity due to the increased demandfor multi-tenancy, higher bandwidths, and the emergence of on-demand resource andapplication models in the Cloud. Software-Defined Networking (SDN) tries to address thesechallenges by altering the traditional paradigm for network control. By decoupling the controland data planes, introducing programmability, centralizing intelligence, and abstractingapplications from the underlying network infrastructure, highly scalable and flexible networkscan be designed that readily adapt to changing business needs.In order to achieve this levelof virtualization and programmability, though, the network has to overcome traditional barriersand evolve so that the whole network can be exposed as an easy to use, programmableinterface.With increased network programmability more granular network control (at the session, user,or device level), and improved automation and management can be achieved.Integral to the SDN value proposition is not just transforming the way networks are designedand operated but also ensuring a tight coupling between the network and the rest of the ITinfrastructure (Servers and Storage), IT Service Management and Applications. This waythere can be a closer synchronization between the network, rest of IT and the business,turning the network into an enabler instead of being perceived as a show-stopper: this is thehigh level goal of the Software Defined Networking paradigm.It is difficult to define Software Defined Networking in an agreed upon fashion among all theinvolved actors as different industry players will have different viewpoints and approaches tothe subject, but from a high level point of view SDN can be defined as a set of tools, protocolsand technologies that enable:OpenDaylight Project: Since the writing of this book the OpenDaylight project has beenannounced.SAN FRANCISCO, April 8, 2013 – The Linux Foundation, the nonprofit organizationdedicated to open source development and technologies, today announced the founding ofthe OpenDaylight Project, a community-led and industry-supported open sourceframework that will accelerate adoption, foster new innovation and create a more open andtransparent approach to Software-Defined Networking (SDN).Why OpenDaylight? The adoption of new technologies and pursuit of programmablenetworks has the potential to significantly improve levels of functionality, flexibility andadaptability of mainstream datacenter architectures. To leverage this abstraction to itsfullest requires the network to adapt and evolve to a Software-Defined architecture. One ofthe architectural elements required to achieve this goal is a Software-Defined-Networking(SDN) platform that enables network control and programmability.IBM and Cisco are Platinum members of this project. For more information as to themission, refer to:http://www.opendaylight.org/Chapter 11. Software Defined Networking and Cisco Open Network Environment 549Draft Document for Review May 9, 2013 6:55 am 8105 11 Cisco Open Network Environment.fm Separation of the network data plane (packet forwarding) and the control plane (logic,network state and policies) so that the latter can be centralized, usually as softwarerunning on a server. Abstraction of network topology and functions so these can be consumed by andorchestrated together with servers, storage and applications. The network to become more programmable, via open APIs - Application ProgrammingInterface.Another important benefit of the SDN architecture is enhanced automation, allowing networksto accommodate highly elastic and rapidly changing demands of users or cloud-basedapplications. Cloud-based applications can now be managed through intelligent orchestrationand provisioning systems, beyond the compute and storage space and including the network.SDN can be used for many purposes, including simplifying network control and management,automating network virtualization services, and providing a platform from which to build agilenetwork services. To accomplish these goals, SDN leverages IETF network virtualizationoverlays and the ONF OpenFlow standards, but it’s not limited to them. SDN is an evolvingconcept and the technologies and standards enabling SND are evolving rapidly.11.1.2 Areas of collaboration between IBM and CiscoAs introduced in the previous section SDN is not just a new networking technology but it ispromising a radical shift in terms of the degree of integration between networks and the restof IT infrastructure and applications.
  • From top to bottom the functional areas are:Data Center Edge : this block allows external access to the Data Center resourcesExternal Security Services : this block enforces security policies on traffic to and from the Data Center Network Services : this block provides all the networking services required for the Data Center except packet forwarding and security services.Internal Security Services : this block enforces security policies on traffic within the Data Center Data Center LAN : this block provides packet forwarding services between clients and servers and between servers in the Data CenterData Center SAN : this block provides connectivity for block level I/O (as opposed to file level I/O) between server and storage end points within the Data CenterData Center Servers : this block provides Data Center computing power resources . Server attach to the LAN and SAN via NICs, or CNAs and to the SAN using or HBAs.Data Center Storage : this block provides Data Center data storage resources. Storage is connected to the SAN using the FC protocol.Data Center Management and Support : this block enables to control, monitor and remotely support Data Center resources
  • The main purpose of a DMZ configuration is to protect the critical data and business logic in the Data Center environment from unauthorized access. An outer firewall pair is used between the public Internet and the Web facing server or servers processing theinternet originated requests that originate onto the company Web site. An inner firewall pair is also used between the Web facing server and the application servers to which it is forwarding requests. Company data resides behind the inner firewall. The area between the two firewalls gives the DMZ configuration its name. Additional firewalls can further safeguard access to databases holding administrative and application data. Load Balancers and other appliances may also be placed at this level.A company's security policy will dictate what is allowed from a network architecture standpoint. This security policy will dictate the physical security requirements for Internet Access and any external security services.

IBM DCN Redbook (Validated Reference Design) IBM DCN Redbook (Validated Reference Design) Presentation Transcript

  • SG24-8105 Joseph Robinson IBM Distinguished Engineer joerobin@us.ibm.com Page 1 IBM Internal
  • Abstract This IBM Redbooks publication is an IBM and Cisco collaboration that articulates how IBM and Cisco can bring the benefits of their respective companies to the modern data center. It documents the architectures, solutions and benefits that can be achieved by implementing a data center based on IBM server, storage and integrated systems, with the broader Cisco network. We describe how to design a state-of-the art data center and networking infrastructure combining Cisco and IBM solutions. The objective is to provide a reference guide for customers looking to build an infrastructure that is optimized for virtualization, is highly available, interoperable, and is efficient in terms of power and space consumption. It will explain the technologies used to build the infrastructure, provide use cases and guidance on deployments IBM Internal
  • Authors Left to Right  Joe Robinson (IBM)  Santiago Freitas (Cisco)  Holger Mueller (IBM)  Michele Girola (IBM)  Peter Clemens (IBM)  Andy Sholomon (Cisco)  Ray O'Hanlon (Cisco)  Veerendra Para (IBM)  Pall Beck (IBM)  Jon Tate (IBM) Missing  Jason Gmitter (Cisco)  Jason Walker (Cisco) Page 3 IBM Internal
  • Agenda  Chapter 1. The Modern Data Center  Chapter 2. IBM and Cisco Building Blocks for the Data Center  Chapter 3. Designing Networks for the Modern Data Center  Chapter 4. Data Center Physical Access Layer Evolution  Chapter 5. Cisco Virtual Networking Solutions  Chapter 6. Application Networking Solutions and Data Center Security (Physical and Virtual)  Chapter 7. Convergence of LAN and SAN - Fibre Channel over Ethernet  Chapter 8. Overcoming the limitations of traditional Data Center Networks  Chapter 9. NX-OS Network Operating System  Chapter 10. Data Center Network Management and Automation  Chapter 11. Software Defined Networking and Cisco Open Network Environment  Chapter 12. IBM and Cisco Next Generation Architectures for Small, Medium and Large Data Centers Page 4 IBM Internal
  • Chapter 1. The Modern Data Center  Data Center Drivers and Trends – Cloud Computing – Virtualization – Big Data – Analytics – Mobility – Security  Architecture Evolution This book will focuses on the Enterprise Virtualized Data Centers, even if the lines between different models tend to blur. Page 5 IBM Internal
  • Chapter 2. IBM and Cisco Building Blocks for the Data Center  IBM PureSystems  IBM Distributed Virtual Switch 5000V  IBM System Networking  IBM Power Systems  IBM System z  IBM iDataPlex  IBM System x  IBM Smarter Storage  Disk Systems and Virtualization  Tape Systems  IBM Tivoli Storage Productivity Center  Cisco Local Area Network Hardware  Cisco Fixed form factor switching platforms  IBM BladeCenter Switching platforms  Cisco Storage Area Network and FCoE Hardware  Power and Cooling Requirements Page 6 IBM Internal
  • Chapter 3. Designing Networks for the Modern Data Center  The Modular Data Center Network Design concept  Data Center Fabric Characteristics and Solutions  IPv6 and the Data Center  Data Center Network Throughput Considerations  Multi-Tenancy  Virtualized Workload Mobility between Data Centers  High Availability and Fast Failover  First Hop Redundancy Protocols  Network Services - Load Balancing, Firewall, WAN Acceleration and Traffic Analysis  IP-based Storage Requirements  Regulatory Requirements  Consistency between Physical and Virtual Network  Differentiated service levels on a shared infrastructure (QoS)  Data Center Cabling Page 7 IBM Internal
  • A PoD is a physical, modular building block in a Data Center with the following characteristics:  Collection of cabinets with a common set of network equipment  Server and network hardware are co-resident in the same PoD  Network functionality is typically Layer 2 only in the form of network access/aggregation  Cabinets in the PoD are typically contiguous but are not limited to this due to special site characteristics  All server NIC to network port connections stay within the PoD to optimize cabling  All PoDs are connected to a Data Center Network Core IBM Internal 8
  • Chapter 4. Data Center Physical Access Layer Evolution  Introduction to Server Access Layer  Traditional Server Access Layer  Physical Cabling and Data Center Layout  End-of-Row (EoR), Middle-of-Row (MoR) and Top-of-Rack (ToR)  Fabric Extender (FEX) Solutions for the Access Layer  Integrated Switching (Blade Switches and Pass-Through Modules)  Flex System (Integrated switches and Pass-through Modules)  IBM System x (Physical and Virtualized)  IBM System p IBM Internal 9
  • Chapter 5. Cisco Virtual Networking Solutions  The Impact of Server Virtualization to the Data Center Network  Cisco Virtual Networking  Nexus 1000v  Cisco Cloud Services Router 1000V Series IBM Internal 10
  • Chapter 6. Application Networking Solutions and Data Center Security (Physical and Virtual)  Network technologies that improve application performance and security  Cisco Application Control Engine (ACE) / SLB Considerations and Solutions  Wide Area Application Services (WAAS)  Cisco APPNAV Building Blocks  Virtual WAAS  ASA Series Security in the Data Center  Cisco Security Manager (CSM)  Cisco TrustSec  Virtualized Network Services Security Services in a virtualized Data Center IBM Internal 11
  • Chapter 7. Convergence of LAN and SAN - Fibre Channel over Ethernet  Converged Networking in Data Centers  Business Benefits and Customer Drivers for Unified Fabric  But what is FCoE all about, is it SAN or LAN?  Fibre Channel over Ethernet architecture  Multi-Hop Fibre Channel over Ethernet  iSCSI technology  Cisco and IBM Solutions for FCoE  Operational Models Enabling FCoE  Acronyms IBM Internal 12
  • Chapter 8. Overcoming the limitations of traditional Data Center Networks  Traditional Data Center Network model and its limitations  Multi-chassis Etherchannel Techniques  Data Center Network Fabrics  Data Center Network Overlays IBM Internal 13
  • Chapter 9. NX-OS Building Blocks  NX-OS Building Blocks  Virtual Device Contexts Overview  Security  IPv6 Support  Advanced System Management  In-Service Software Upgrade  Control Plane and Data Plane Separation  Graceful Restart and Non Stop Routing  NX-OS Fast Failover Features VDC Diagram IBM Internal 14
  • Chapter 10. Data Center Network Management and Automation  Data Center Network Management Requirement  Cisco Prime Data Center Network Manager LAN and SAN  Cisco Virtual Network Management Center  Cisco Access Control System (ACS)  Cisco Prime Infrastructure  Cisco Network Services Manager  Cisco Security Manager (CSM)  Cisco Prime Network Registrar (CNR)  Network Capabilities Enabling Application Performance Management  IBM Tivoli Solutions IBM Internal 15
  • IBM Tivoli Network Manager functional layers Page 16 IBM Internal
  • Chapter 11. Software Defined Networking and Cisco Open Network Environment  Software Defined Networking Overview  Cisco Open Network Environment Software Defined Networking Framework IBM Internal 17
  • Chapter 12. IBM and Cisco Next Generation Architectures for Small, Medium and Large  IBM and Cisco Data Center Objectives  IBM and Cisco Data Center Component Model  Data Center Requirements  Data Center Use Cases  Small Data Center Use Case  Medium Data Center Use Case  Large Data Center Use Case  WAN Edge Block Design for Small, Medium and Large Data Centers  DMZ Design  Services supporting Next Generation Data Center Architectures  IBM Global Technology Services for Modern Data Centers  Cisco Data Center Networking Services IBM Internal 18
  • Data Center Architecture Page 19 IBM Internal
  • Use Cases Page 20 IBM Internal
  • Small Data Center Use Case Design Concept Page 21 IBM Internal
  • Medium Data Center Use Case Design Concept Page 22 IBM Internal
  • Large Data Center Use Case Design Concept Page 23 IBM Internal
  • DMZ design  The main purpose of a DMZ configuration is to protect the critical data and business logic in the Data Center environment from unauthorized access. Page 24 IBM Internal
  • Questions Page 25 IBM Internal