Cisco Unified Computing Systems Update


Published on

The Cisco Unified Computing System represents a radical simplification of traditional architectures, dramatically reducing the number of devices organizations must purchase, deploy, and maintain. Cisco UCS delivers end-to-end optimization for virtualized environments while retaining the ability to support traditional OS and application stacks in physical environments. This session will provide attendees with an update on new features available from both hardware and software perspectives for the Cisco Unified Computing System.

Published in: Technology
1 Like
  • Be the first to comment

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide
  • Trends:Anything as a ServiceOptimize heterogeneous workloadsVM MobilityDynamic Resource AllocationOn-demand ResponseReal-time performance
  • Today’s typical IT environment is characterized by independent, silo’ed technologies and organizations.These silos result in ineffective use of resources and incremental complexity to operations, the net effect of which is to drive up total cost of ownership.
  • Cisco’s Data Center Business Advantage framework offers a different approach. It is based on the 3 key technology pillars of Unified Compute, Unified Network Services, and Unified Fabric. By unifying these 3 elements into a single, cohesive Data Center Fabric, Cisco enables customers to eliminate the silos, optimize use of their resources, simplify operations and lower TCO. Cisco’s Data Center Fabric allows customers to transform the way they consume IT by enabling virtualization and facilitating the move to private, public or hybrid cloud models.
  • Cisco’s Data Center Fabric Vision delivers a flexible, high-performance, secure, shared infrastructure, tightly integrated amongst network, storage, compute and application services coupled with Cisco Life cycle Services and Innovations that facilitate an “enter anywhere, grow anywhere” approach to evolving a customer’s Data Center – focusing on aligning to customer architectural needs, implementation timelines, and ROI requirements. All through an architecture that is:OpenIntegratedFlexibleScalableResilient andSecure
  • At Cisco, we have been on an evolutionary path for over three years. It started with the vision of converged networks.Next phase was when we integrated services into network structure for complete VM visibility to support wide scale virtualizationThen brought in the compute layer to integrate virtualized network and services into the serverMoving forward to next evolution integrated policy based management across fabric lanesSimplifying management vision taking us to single pane of glass management with massive scalability10’s of thousands of server in single L2 domain with more efficient resource utilizationNo one else can deliver across the data center fabric – others have only piece partsCisco supports Heterogeneous workloads with mobilityCisco has vision…with execution
  • The Cisco UCS solution is a set of pre-integrated data center components that substantially reduces the size of the footprint of the systems in the data center.The intelligence to manage the system, called UCS Manager, is embedded in a fabric switchThe UCS Manager is an XML-based application that provides a common GUI for managing servers, storage and networksIt has an Open API that Allows UCSManager to be integrated with other management console systems.The common GUI makes it much easier for various IT disciplines to collaborate on provisioning and managing the data center environment.The UCS Manager eliminates the need for separate server, storage and switch management tools.Each end point in the system is linked back to the UCS ManagerEach segmented element of the UCS system is assigned a Service Profile. The UCS Manager manages and maintains each Service Profile, which is an abstraction of the underling server, storage and switching equipment that makes up each Service ProfileAs a result, UCS Manager Reduces the Time Required to Provision any set of integrated server, storage and switch components down to a matter of minutesBlade servers are based on Intel Nehalem-class processorsFewer Components and Simplified Management Translates into Much Less Cost.
  • Cisco’s Unified Computing System is based on Cisco’s unified approach to the data center that brings together server, storage under a single management fabric.Cisco has worked closely with VMware to extend all the management capabilities of UCS Manager to support virtual servers.This creates the potential to lower systems management costs by allowing staff members to manage a larger number of devices by leveraging the ability to dynamically provision physical and virtual IT assets.UCS Manager has the ability to auto-discover any element of the system.But no element of the system can be invoked until it is assigned to be part of a particular set of Service ProfilesThis approach also makes it easier to see the dependencies between elements of the data center because it manages the whole system at a higher level of abstraction.Fewer components in the system creates a better eco-footprint in terms of both the amount of power consumed and carbon generated.A unique shared memory architecture developed with Intel for Nehalem processors actually increases application performance in a model that also lowers the total cost of ownership.
  • Unified Management DomainManagement embedded in all system elementsAutomatic discovery of all componentsOperates autonomouslyDynamic ProvisioningAllows for complete infrastructure repurposingOffers seamless server mobility: identity and connectivity policyDeploy in minutes, not daysBuilding Block for Dynamic Data CenterComprehensive open XML APIIntegration with 3rd party tools
  • The Cisco Server portfolio continues to grow with the introduction of the C260M2! Expands form factor options in the UCS family to 9 systems. Industry-leading memory and I/O density, combined with 16 SSD or HDD: ideal for local-storage, performance-intensive OLTP and Data Warehousing applications Brings Cisco Extended Memory Technology to the mission critical systems class with double the memory of any 2-socket Intel Westmere EX-based server
  • N7K: scales beyond 15Tbps vs. FC platforms which scale between 2~4TbpsN7K:40G and 100G ready vs. FC roadmap uncertain beyond 16G FCMultiprotocol flexibility to supports iSCSI, NAS, FCoE,FCMDS: Bridge Converged Networks to FC SANs, preserving existing and continued investments in FCMDS: Delivers services such as FCIP distance extension to Converged NetworksMulti-hop FCoE uses VE ports, standardized in FC-BB-5 in July 2009…no dependency on TRILL or FabricPathMulti-hop FCoE scales to 7 hops and 10,000 devices per fabric…the same scale supported for Fibre Channel SANsA single operating system (NX-OS) and a single management tool (DCNM) for both Nexus and MDS platforms
  • The Cisco Unified Fabric Switching portfolio offers one of the broadest data center LAN and SANswitching portfolio’s in the industry spanning from the Hypervisor to the data center core. Cisco Unified Fabric provides the flexibility of high-performance, highly available, highly-scalable networks to serve diverse data centers needs including the lossless requirements to carry storage traffic (FC, FCoE, iSCSI, NAS) over simplified infrastructure based on 10Gb Ethernet. All platforms in the portfolio run on a common operating system - NX-OS. Cisco is the only vendor to provide a single operating system across all PINs in the data center. Having a single operating system across every element of the data center network, provides operational and functional consistency with tight integration across the unified fabric significantly simplifying operations and providing a foundation for accelerated innovations. Adding to this industry leading portfolio, we are announcing the Nexus 3000 family designed for ultra-low latency, wire speed applications such as High Frequency Trading in financial applications. We are also introducing new members to the Nexus 5000 family, the Nexus 5548UP and the 5596UP which offer high density low latency unified ports for server access. We are also announcing the addition of FCoE and Multi-Hop FCoE to the MDS and Nexus 7000 platforms enabling end to end LAN and SAN convergence over a unified fabric.
  • Adapter FEX: Adapter FEX is a technology that extends the current Cisco Fabric Extender solution into the servers. Adapter FEX enables the server adapter to be logically partitioned into multiple vNICS. Each vNIC behaves like a physical NIC port and satisfies the network connectivity need for each application, delivering independent network polices for each VM. Adapter FEX enables the efficient utilization of 10 Gigabit Ethernet bandwidths. The technology is based on the standards architecture, IEEE 802.1Qbh, delivering secure, scalable network monitoring and management for virtualized servers. It enables a single point of management from the access layer switch that manages and controls the end-to-end connectivity between server adapters and the switches in a virtualized environment. VM-FEX: Cisco is extending the switching fabric all the way to the server hypervisor, by providing switching of VM traffic instead of using a software switch within the hypervisor, achieving greater performances and consolidation of virtual access layer and physical access layer.Highlights:Efficient utilization of 10 Gigabit Ethernet bandwidthsPartition the 10 Gigabit Ethernet server adapter to provide dedicated network bandwidth to individual applications or virtual machinesReduce cabling, NICs, power and operational costsSingle point of monitoring and managementMonitor and manage the entire network extending into the virtual machines from the Nexus 5000 Series SwitchesFlexibility and Scalability with SecurityBased on industry standard, IEEE 802.1QbhThe standard enables the bare metal or virtualized environments to scale as the business grows – increased number of applications, virtual machines or number of servers.ACLs implementation is based on tagged data packets instead of MAC addresses. By doing so each data packet is protected and does not impact the overall network by any rogue data packet.
  • LISP is a new routing architecture that enables enterprises and service providers to transform their network architectures to be more scalable and agile for virtualization and cloud computing environments. Unlike current IP routing and addressing architecture that uses a single numbering space for device identity and location, the LISP routing architecture creates a new paradigm by splitting the device identity and its location into two different numbering spaces. This address separation provides global IP address portability and significantly increases routing scalability with on-demand route look up. Customers can take advantage of the LISP architectural benefits to enable: Workload Mobility –IP address portability enables virtual machines and physical hosts to be moved seamlessly across Layer 3 boundaries to other data centers or to cloud service providers while preserving their IP addresses. This enables live global application migration and optimal traffic routing. The user traffic is automatically re-routed to the new location. This make the live VM move non disruptive and transparent to the end user. Scalable and secure Multi-tenancy – The LISP routing architecture eliminates the route table scaling issues by only populating required routes on-demand.IPv6 transition – LISP enables IPv6 to be deployed incrementally over the existing IPv4 infrastructure. This approach allows for an incremental and simpler phased transition to IPv6.Through its participation in standards bodies such as the IETF LISP Working Group, Cisco is committed to the development of the LISP architecture as an industry standard.
  • For those interested in copies of this year’s conference presentations, please visit our Event Landing Page at Here you will find the presentations for download.Lastly, we are interested in your feedback. Please take the time to fill-out the Conference Evaluation Form. If you did not receive an Eval Form with today’s Conference Guide, please see one of our Registration Attendants. The Eval Forms will be used for the prize draws at the Cocktail Reception.
  • Cisco Unified Computing Systems Update

    1. 1. Cisco Unified Computing System Update<br />Presented by:<br />Ivan Kobe<br />Consulting Systems Engineer<br />#CNSF2011<br />
    2. 2. Unified Computing System<br />BUSINESS VALUE<br />SOLUTION<br />DIFFERENTIATION<br />Lower TCO<br />BUSINESS VALUE<br />Virtualization-Aware<br />SOLUTION DIFFERENTIATION<br />SYSTEMS EXCELLENCE<br />Rapid ROI<br />End-to-end Networking<br />Superior Agility<br />TECHNOLOGY INNOVATION<br />Radical Simplification<br />SYSTEMS EXCELLENCE<br />Efficient<br />Agile<br />Cost Effective<br />Transformative<br />TECHNOLOGY INNOVATION<br />RapidDeployment<br />WorkloadMobility<br />UnifiedManagement<br />MaximizedScaling<br />ServiceProfiles<br />Operational Efficiency<br />Open API<br />Unified ITWorkflows<br />VirtualInterfaces<br />ExtendedMemory<br />
    3. 3. Unified Computing System<br />A single system<br /><ul><li>Compute: Industry standard x86
    4. 4. Network: Unified fabric
    5. 5. Virtualization: Control, scale, performance
    6. 6. Storage Access: Wire once for SAN, NAS, iSCSI</li></ul>Embedded management<br /><ul><li>Increase scalability without added complexity
    7. 7. Dynamic resource provisioning
    8. 8. Ability to integrate with broad partner ecosystem</li></ul>Highly efficient <br /><ul><li>Fewer servers, switches, adapters, cables
    9. 9. Lower power and cooling requirements
    10. 10. Fewer people to deploy and manage </li></li></ul><li>New Data Center DemandsImpact of Virtualization and Cloud<br />Dynamic response<br />Anything-as-a-Service<br />VM Mobility<br />Heterogeneous Workloads for ANY APPLICATION<br /> Optimize resource allocation across ANY LOCATION <br /> Respond in real-time atANY SCALE<br />
    11. 11. Traditional IT Silos<br />Network<br />Compute<br />Application Services<br />Storage<br />Inefficient Resource Utilization<br /> Operationally Complex—Higher TCO<br /> Promotes Human Silos<br />
    12. 12. Cisco Approach: Data Center Fabric<br />Unified Fabric<br />Network<br />Compute<br />Application Services<br />Storage<br />Unified Network Services<br />Unified Computing<br /> Optimized Resource Utilization<br /> Operationally Simple—lower TCO<br /> Overcomes IT Silos<br />
    13. 13. Cisco Data Center Fabric Vision<br />Flexible, High-Performance, Secure, Shared Infrastructure<br />Tightly Integrated Network, Storage, Compute and Application Services<br />Cisco Life cycle Services and Innovations that facilitate “enter anywhere, grow anywhere”<br /> Open Integrated Flexible Scalable Resilient<br />Secure<br />
    14. 14. Cisco Data Center Fabric EvolutionIndustry-Leading Transitions since 2008<br />Fabric Foundation to Handle Dynamic Cloud-Based Environments<br />Application Independence<br />Location Freedom<br />Massive Scale<br />2008<br />Evolutionary Innovations Offering Investment Protection<br />
    15. 15. The Benefits of UCS<br />#CNSF2011<br />
    16. 16. Modular Building Blocks<br />
    17. 17. Mgmt Server<br />Server Deployments Today<br />Over the past 10 years<br /><ul><li>More cores, more memory
    18. 18. Proliferation of servers & switches
    19. 19. More switches per server
    20. 20. Management applied, not integrated</li></ul>An “accidental” architecture<br /><ul><li>Evolved by default vs planned design</li></ul>Result: Complexity<br /><ul><li>Many points of management
    21. 21. Difficult to maintain policy coherence
    22. 22. Difficult to secure
    23. 23. Difficult to scale</li></li></ul><li>Mgmt Server<br />Mgmt Server<br />Mgmt Server<br />Our Solution<br /><ul><li>Embed management
    24. 24. Unify fabrics
    25. 25. Optimize virtualization
    26. 26. Remove unnecessary switches, adapters and management modules
    27. 27. Less than 1/3rd the support infrastructure for a given workload</li></li></ul><li>Mgmt Server<br />Our Solution:<br />Unified Computing System<br /><ul><li>A single system that encompasses:</li></ul>Network: Unified fabric<br />Compute: Industry standard x86 components<br />VNlink: Distributed Virtual Switch<br /><ul><li>Efficient Scale</li></ul>Fewer servers with more memory<br /><ul><li>Lower cost</li></ul>Fewer servers, switches, adapters, cables<br />Lower power consumption<br />Fewer points of management<br /><ul><li>Unified management model</li></li></ul><li>Distributed Architecture<br />LAN/SAN Uplinks<br />20Gb/s<br />40Gb/s<br />80Gb/s<br /><ul><li>Wire once for bandwidth, not connectivity
    28. 28. Policy-driven bandwidth allocation
    29. 29. All links can be active all the time
    30. 30. Integrates as a single system into your data center</li></li></ul><li>Consolidated Management, Single Tool<br />Cisco Solution<br />Today<br />Single Embedded Device Manager<br />Mgmt Server<br />Multi-Chassis<br />Connectivity Manager<br />Connectivity Manager<br />Chassis Manager<br />Blade Manager<br /><ul><li>Reduced management costs
    31. 31. Easier to integrate with existing frameworks
    32. 32. Single consistent API for all managed objects</li></li></ul><li>UCS Manager<br />
    33. 33. Embedded Unified Management<br />Existing<br />Customer<br />Mgmt Tools<br />TightlyCoupled Partner<br />Mgmt Tools<br /><ul><li>Unified Management Domain
    34. 34. Automatic discovery
    35. 35. Dynamic Provisioning
    36. 36. Building Block for Dynamic Data Center
    37. 37. Simplify management of infrastructure for ESX clusters and datacenters
    38. 38. One-click configuration of LAN, SAN and firmware parameters</li></ul>XML API<br />Traditional<br />APIs<br />Service Profile: HR-App1<br />Network: HR-VLAN<br />Network QoS: High<br />MAC: 08:00:69:02:01:FC-E<br />WWN: 5080020000075740-3<br />BIOS: Version 1.03<br />Boot Order: SAN, LAN<br />App<br />OS<br />Firmware<br />Network<br />
    39. 39. Zero Touch IntegrationDecouple Complexity & Scale<br />Name: UCS 5108<br />Class: Chassis<br />ID: 234222-33<br />IOM 1: UCS 2104<br />IOM 2: UCS 2104<br />Blade slot occupied: 8<br />Fans: 8<br /><ul><li>Increase capacity, not complexity
    40. 40. New equipment self integrates</li></ul>Physical Inventory<br />Name: UCS 12<br />Class: System<br />ID: 77449-32<br />Chassis: 1<br />- IOM 1: UCS 2104<br />- IOM 2: UCS 2104<br /><ul><li> Blade slots occupied: 8</li></ul>Chassis: 2<br />- IOM 1: UCS 2104<br />- IOM 2: UCS 2104<br /><ul><li> Blade slots occupied: 8</li></ul>Chassis: 3<br />- IOM 1: UCS 2104<br />- IOM 2: UCS 2104<br /><ul><li> Blade slots occupied: 8</li></ul>Chassis: 4<br />- IOM 1: UCS 2104<br />- IOM 2: UCS 2104<br /><ul><li> Blade slots occupied: 8</li></ul>Chassis: 5<br />- IOM 1: UCS 2104<br />- IOM 2: UCS 2104<br /><ul><li> Blade slots occupied: 8</li></ul>Name: UCS 2104<br />Class: FEX<br />ID: 234222-33<br />IOM 1: UCS 2104<br />IOM 2: UCS 2104<br />Blade slot occupied: 8<br />Fans: 8<br />
    41. 41. Zero Touch IntegrationDecouple Complexity & Scale<br /><ul><li>Increase capacity, not complexity
    42. 42. New equipment self integrates
    43. 43. Inventory & status updated</li></ul>Physical Inventory<br />Name: UCS 12<br />Class: System<br />ID: 77449-32<br />Chassis: 1<br />- IOM 1: UCS 2104<br />- IOM 2: UCS 2104<br /><ul><li> Blade slots occupied: 8</li></ul>Chassis: 2<br />- IOM 1: UCS 2104<br />- IOM 2: UCS 2104<br /><ul><li> Blade slots occupied: 8</li></ul>Chassis: 3<br />- IOM 1: UCS 2104<br />- IOM 2: UCS 2104<br /><ul><li> Blade slots occupied: 8</li></ul>Chassis: 4<br />- IOM 1: UCS 2104<br />- IOM 2: UCS 2104<br /><ul><li> Blade slots occupied: 8
    44. 44. Immediately apply existing policies</li></ul>Policy Inventory<br />Service Profile: Default 1<br />Service Profile: HR-App1<br />Chassis: 5<br />- IOM 1: UCS 2104<br />- IOM 2: UCS 2104<br /><ul><li> Blade slots occupied: 8</li></li></ul><li>Programmatic InfrastructureDevelop With The Infrastructure, Not Just On The Infrastructure<br /><ul><li>Comprehensive XML API, standards-based interfaces
    45. 45. Bi-Directional access to physical & logical internals</li></ul>Self Serve portals<br />Management Tools<br />Auditing Tools<br />Direct<br />UCS CLI<br />UCS GUI<br />Customer<br />3rd Party<br />XML API<br />System Status<br />Physical Inventory<br />Logical Inventory<br />
    46. 46. Cisco Server Portfolio<br />Mainstream<br />Intensive and Mission Critical<br />Extended Memory<br />WSM EX!<br />WSM EX!<br />New!<br />WSM EX!<br />Rack-Mount<br />UCS C210 M2<br />UCS C460 M2<br />UCS C260 M2<br /><ul><li>Leading memory & I/O density for OLTP, Data WH
    47. 47. 2S WSM EX, Extended Memory, 16 SSD/HDD (9.6 TB)
    48. 48. 2S Licensing @ 4S RAS/Extensibility / Performance</li></ul>UCS C250 M2<br />UCS C200 M2<br />Extensibility<br />UCS B250M2<br />UCS B440M2<br />UCS B230 M2<br />UCSB200M2<br />Blade<br />2S EX<br />2S EP<br />4S EX<br />Cisco UCS: Many Form Factors, One System<br />
    49. 49. Expanded Memory Architecture <br /><ul><li>Nehalem Memory Architecture
    50. 50. 3 Channels
    51. 51. 3 Banks per Channel
    52. 52. All Channel Banks should be populated equally for optimum performance. Populating Channel Banks differently is possible, however, results in unbalanced memory performance.
    53. 53. Different Memory bus speeds dependant on how many DIMMs are populated.
    54. 54. 1333MHz
    55. 55. 1066MHz
    56. 56. 800MHz</li></ul>Physical View<br />Integrated Memory Controller<br />Slot 5<br />Slot 4<br />Slot 3<br />Slot 2<br />Slot 1<br />Slot 0<br />
    57. 57. Expanded Memory Architecture <br />Logical View<br />Physical View<br />Integrated Memory Controller<br />Integrated Memory Controller<br />
    58. 58. New: Cisco UCS C260M2Bringing Cisco Extended Memory Technology to Mission Critical Class Systems<br />Ideal for local OLTP and Data Warehouse applications:<br /><ul><li>Designed for core and memory-intensive appsrequiring massive local storage and I/O scaling
    59. 59. Greater savings on per-socket/core application and hypervisor licensing
    60. 60. Silicon and system-level RAS features for mission-critical applications
    61. 61. Industry unique: double the memory of any other server in this class</li></ul>Compute:<br /><ul><li>2S Intel Westmere EX</li></ul>Memory:<br /><ul><li>64 DIMM Slots
    62. 62. 1TB of memory using 16GBDIMMs</li></ul>Local Storage:<br /><ul><li>16 SSD or HDD
    63. 63. Up to 9.6 TB</li></ul>I/O:<br /><ul><li>6 x PCIe
    64. 64. 2 x 10GbE SFP+ LOMs, 4 x 1GbE LOMs</li></li></ul><li>UCSM C-Series Integration<br />UCS<br />Manager<br />Customer benefits<br /><ul><li>Unified Management across entire UCS portfolio
    65. 65. Advanced capabilities extended to rack servers</li></ul>Feature details<br />UCS 6100<br />UCS 6100<br /><ul><li>C-Series servers connectivity</li></ul>Management connectivity through FEX to FI<br />Data connectivity directly to Fabric Interconnect <br /><ul><li>Stateless computing</li></ul>Service profiles extended to C-series<br />Migration among compatible B & C series servers<br /><ul><li>All UCSM management services</li></ul>Automated discovery<br />Fault and monitoring<br />Firmware updates<br />Nexus 2248<br />Nexus 2248<br />Mgmt Connection<br />Data Connection<br />
    66. 66. Legacy FabricUnderutilized Resources, Stranded Bandwidth<br />Server Life Cycle<br />Network<br />Boot<br />Production<br />VMotion<br />Vmotion<br />2 Gb<br />Back<br />2 Gb<br />LAN<br />Front<br />2 Gb<br />SAN<br />8 Gb<br />SAN<br />14 Gb<br />3 Gb<br />2 Gb<br />4 Gb<br />8 Cables<br />Total<br />Used<br />Used<br />Used<br />
    67. 67. Unified FabricLower cost: Half the switches, ¼ the adapters Higher Capacity, Performance & Flexibility <br />Network<br />Server Life Cycle<br />Production<br />Boot<br />VMotion<br />Vmotion<br />2 Gb<br />Back<br />2 Gb<br />LAN<br />Front<br />2 Gb<br />20 Gb<br />Unified<br />Fabric<br />SAN<br />8 Gb<br />SAN<br />14 Gb<br />3 Gb<br />2 Gb<br />4 Gb<br />20 Gb<br />8 Cables<br />2 Cables<br />Total<br />Used<br />Used<br />Used<br />
    68. 68. Unified FabricLower cost: Half the switches, ¼ the adapters Higher Capacity, Performance & Flexibility <br />Network<br />Server Life Cycle<br />Production<br />Boot<br />VMotion<br />Vmotion<br />Back<br />Front<br />20 Gb<br />Unified<br />Fabric<br />SAN<br />20 Gb<br />3 Gb<br />2 Gb<br />4 Gb<br />10 Gb<br />20 Gb<br />10 Gb<br />2 Cables<br />Total<br />Used<br />Used<br />Used<br />
    69. 69. Extending Convergence Beyond Access LayerIndustry’s First Director-Class Multi-hop FCoE<br />SAN Core<br />FC<br />Nexus 7K and MDS 9500: <br />High-Performance, Resilient Switches<br />for LAN/SAN<br />FCoE<br />Common Core <br />FCoE<br />Hosts<br />Access <br />StorageTargets<br />FCoE<br />FCoE<br />FCoE<br />Benefits<br /><ul><li>Wire once flexibility
    70. 70. Up to 45% access layer CapEx savings*
    71. 71. 492% ROI for converged networks*
    72. 72. Seamlessly integrate converged networks with existing MDS FC SANs</li></ul>Features<br /><ul><li>Nexus 7000 FCoE:: Industry's highest performance Director-Class SAN platform
    73. 73. MDS 9500 FCoE support: Preserves investments in Fibre Channel SANs
    74. 74. Multi-hop FCoE support: Spans Nexus 5000, Nexus 7000, and MDS 9500.</li></ul>Delivering Highly Scalable End-to-End Convergence<br />* Based on IDC’s UF Beyond the Access Layer White Paper <br />
    75. 75. Cisco Unified Fabric Switching Portfolio<br />LAN/SAN<br />SAN<br />LAN<br />New!<br />New!<br />Nexus 7000<br />MDS 9500<br />Nexus 5000<br />Nexus 4000<br />Scalability<br />MDS 9200<br />Nexus 2000<br />Nexus 1010<br />Nexus 3000<br />MDS 9100<br />Nexus 1000V<br />Cisco NX-OS: One OS from the Hypervisor to the Data Center Core<br />Convergence<br />VM-Aware<br />Networking<br />10 GbE<br />switching<br />Fabric<br />Extensibility<br />Cloud Mobility<br />
    76. 76. New Nexus 3000 and 5500 Switches<br />Shipping now!<br />AvailableQ2CY11<br />Nexus 5548UP and 5596UP:<br />Unified Ports and L3 Option<br /><ul><li>High density 1RU/2RU ToR Switches
    77. 77. All ports 10GE/1GE/FCoE/8G FC
    78. 78. Increased FEX scale (upto 24)
    79. 79. IEEE 802.1Qbh support
    80. 80. Cisco FabricPath(coming soon)</li></ul>Nexus 3000:<br />Purpose-build for High-Frequency Trading<br /><ul><li>High density 1RU 10GbE/1GbE/(40 GbE Ready)
    81. 81. Ultra-low latency (~1usec port to port )
    82. 82. L2/L3 – Unicast & Multicast
    83. 83. Wire-rate, non-blocking</li></ul>Proven NX-OS with Full-Featured L3<br />Wire-Rate High Density and Low Latency<br />
    84. 84. Cisco Nexus 1000V<br /><ul><li>Industry’s most advanced software switch for VMware vSphere
    85. 85. Built on Cisco NX-OS
    86. 86. Compatible with all switches
    87. 87. Compatible with all servers on the VMware Hardware Compatibility List
    88. 88. Winner of VMWorld Best in Show 2008 and Cisco Most Innovative Product of 2009</li></ul>VM<br />VM<br />VM<br />VM<br /> vSphere<br />1000V<br />VEM<br />Server<br />1000VVSM<br />Physical Switches<br />
    89. 89. Cisco Nexus 1000V<br />Policy-Based VM Connectivity<br />Mobility of Network and Security Properties<br />Non-Disruptive Operational Model<br />VM<br />VM<br />VM<br />VM<br />VM<br />VM<br />VM<br />VM<br /> vSphere<br /> vSphere<br />Nexus<br />1000V<br />Nexus<br />1000V<br />Server<br />Server<br />VMware vCenter<br />Physical Switches<br />Nexus 1000VVSM<br />
    90. 90. App<br />App<br />App<br />App<br />App<br />App<br />App<br />App<br />App<br />App<br />App<br />App<br />OS<br />OS<br />OS<br />OS<br />OS<br />OS<br />OS<br />OS<br />OS<br />OS<br />OS<br />OS<br />Phase 1 VirtualizationConsolidated Environment<br />Hypervisor<br />Hypervisor<br />Hypervisor<br />
    91. 91. App<br />App<br />App<br />App<br />App<br />App<br />App<br />App<br />App<br />App<br />App<br />App<br />OS<br />OS<br />OS<br />OS<br />OS<br />OS<br />OS<br />OS<br />OS<br />OS<br />OS<br />OS<br />vSwitch<br />vSwitch<br />vSwitch<br />Phase 1 VirtualizationMore Virtual Infrastructure Needed<br />Hypervisor<br />Hypervisor<br />Hypervisor<br />
    92. 92. App<br />App<br />App<br />Net State<br />Net State<br />OS<br />OS<br />OS<br />Net State<br />vSwitch<br />vSwitch<br />vSwitch<br />Phase 2 Virtualization: MobilityNetwork Policies Tied to vSwitch – Not Mobile<br />X<br />Hypervisor<br />Hypervisor<br />Hypervisor<br />
    93. 93. App<br />App<br />App<br />Net State<br />Net State<br />OS<br />OS<br />OS<br />Net State<br />vSwitch<br />vSwitch<br />vSwitch<br />Cisco VN-LinkNetwork Policies Tied to Virtual Machine<br />Hypervisor<br />Hypervisor<br />Hypervisor<br />
    94. 94. App<br />App<br />App<br />Net State<br />Net State<br />OS<br />OS<br />OS<br />Net State<br />vSwitch<br />vSwitch<br />vSwitch<br />vNetwork Distributed Switch<br />Cisco VN-LinkNetwork Policies Tied to Virtual Machine<br />VN-Link Domain<br />Hypervisor<br />Hypervisor<br />Hypervisor<br />
    95. 95. Adapter FEX and VM-FEXExpanding Cisco Fabric Extender Offerings<br />VM FEX<br /><ul><li>IEEE 802.1Qbh
    96. 96. Ecosystem of multiple NIC vendors </li></ul>Features:<br /><ul><li>Adapter FEX split a physical NIC into multiple logical NICs
    97. 97. VM-FEX extends Adapter FEX technology to virtual machine</li></ul>Hypervisor<br />vNIC<br />vNIC<br />Benefits:<br /><ul><li>Single point of management
    98. 98. Increased 10 G bandwidth utilization – less power, cabling adaptors with adapter FEX
    99. 99. Dynamic network & security policy mobility during VM migration with VM-FEX</li></ul>vEth<br />vEth<br />Adapter FEX<br />
    100. 100. Location ID/Separation Protocol (LISP) Enabling Global Workload Mobility<br />AvailableQ2CY11<br />LISP decouples host IP identity from host location enabling global IP address portability<br />IP address and session move with VM<br />IP address A<br />IP address A<br />Benefits:<br /><ul><li>Seamless Global Workload Mobility across cloud
    101. 101. Network scalability, secure Multi-tenancy
    102. 102. Simplified IPv6 transition</li></ul>Feature:<br /><ul><li>IP address Portability
    103. 103. On-Demand Route lookup
    104. 104. IPv4/ IPv6 encapsulation</li></ul>Cisco NX-OS: Delivering Location Independence with OTV and LISP<br />
    105. 105. The Cisco AdvantageWorkload Mobility between data centers<br />Layer-2 extension using OTV, <br />Typical Use case :<br /><ul><li>Work Load migration
    106. 106. Disaster recovery </li></ul>Data Center 1<br />Data Center 2<br />Data Center<br />Interconnect<br />OTV<br />VMwarevCenter<br /> vSphere<br /> vSphere<br />Nexus<br />1000V<br />Nexus<br />1000V<br />IP/MPLS Core with GETVPN<br />
    107. 107. Bringing it all together: Any Application. Any Location. Any Scale.<br />2<br />1<br />3<br />Toronto<br />Fabric Compute, Convergence and Scale<br />Workload Mobility<br /><ul><li>UCS
    108. 108. FabricPath/TRILL, VPC
    109. 109. OTV LAN Extensions
    110. 110. FCIP, I/O Acceleration
    111. 111. FCoE, Unified Ports
    112. 112. LISP
    113. 113. OTV
    114. 114. VN-Link
    115. 115. Port Profiles
    116. 116. VM-FEX
    117. 117. VDC, VRFs</li></ul>New York<br />Fabric Extensibility<br />VM-Aware Fabric<br />
    118. 118. Cisco’s Architectural Differentiators<br /><ul><li>Unified FabricBy using FCoE and VN-Link to dramatically reduce network adapters, switches, and cabling, uncompromised I/O performance and flexibility are utilized to provide a platform that can truly meet the connectivity demands of virtualizing any compute environment .
    119. 119. Leading Industry Performance, Scale and MobilityUCS providies the latest in technologies enabling customers to realize the full benefits of virtualization regardless of location, application or scale.
    120. 120. Feature-rich Service ProfilesUCS Manager implements role and policy-based service profiles and templates that provide substantially more features for managing server identities and controlling server deployment.
    121. 121. Enterprise-class Management ArchitectureUCS management architecture provides an enterprise-class foundation for system management
    122. 122. Virtualization OptimizationsExtended memory, Virtual Interface Cards and network integration with hypervisors are three major virtualization optimizations offered with UCS.</li></li></ul><li>Q & A<br />#CNSF2011<br />
    123. 123. For conference presentations visit: <br /><br />Please take a moment to complete the <br />Networkers Conference Event Evaluation Form<br />#CNSF2011<br />
    124. 124. #CNSF2011<br />