Cisco Unified Computing System · Virtualization Step2 10 GE VM VM VM VM VM VM VM VM Hypervisor s...

Cisco Unified Computing System

Maciej Bocian

mbocian@cisco.com

Architecture Sales Manager

Data Center and Virtualization, Central Europe

CCIE#7785

Unified Computing Continues Data Center 3.0 Strategy

Automation Utility MarketConsolidation Virtualization

Data Center Networking

Unified Fabric

Unified Computing

Private Clouds

Inter-Cloud

•Many under utilized servers•Cable sprawl•High power, cooling costs•High CAPEX•For every $1 spent on server capex ~$5 spent on opex

Server Virtualization - Key DC Trend

Hypervisor

•Cable sprawl•power, cooling costs•Less number of access layerEthernet ports

Access

Server

Fibre-

ChannelEthernet

SAN BSAN ALAN

4 x 1GE

Virtualization Step1

Hypervisor

Access

Server

Fibre-

ChannelEthernet

SAN BSAN ALAN

•GE to 10GE in access layer•Less interfaces –reduced Cable sprawl•Savings from power and cooling

Hypervisor

Access

ServerUnified IO

SAN BSAN ALAN

•Unified I/O - LAN & SAN consolidation•Reduce NICs, HBAs,•Reduce cabling•More Savings from power and cooling•Lower capex

10 GE/FCOE

Cisco confidential and proprietary

Unified Fabric

• Wire once, low latency

FC and Ethernet

• Virtualization aware • Dramatic reduction in

adapters, switches,

pass thru modules

Industry Standard Servers

• Blade Form Factor• Intel Xeon Processor 5500

series.

• More than double the

memory capacity of

competing systems

Virtualized Services

• Fine-grained control, portability, and visibility of network, compute, and storage attributes

• Increased Processor Efficiency with Hypervisor Bypass

Up to 30% fewer components, switches, cabling, and management modules to purchase, manage, power, and cool

Up to 30% lower memory and SW licensing costs via Cisco Extended Memory Technology

Up to 10% better processor performance via Cisco Hypervisor Bypass Technology

Automated Provisioning

• Embedded single point of management and provisioning

• Visibility and control across technology silos

• Ongoing management and compliance

Up to 90% greater administrator efficiency, with faster changes and fewer incidents

Process Automation (ITIL)

tions a

upport

Scalable Unified Fabric that delivers up to 320 server nodes in a single system

The Cisco Unified Computing System is designed to dramatically reduce datacenter total cost of ownership while

simultaneously increasing IT agility and responsiveness.

Server evolution

Server Deployment: Rackmounts

First generation

Rack-optimized

Top of Rack or End of Row switches

Cables

Benefits

Space utilization

Highly flexible

Weakness

Cabling

Serviceability

Power efficiency

Server

Second generation• Blade servers

• Integrated switches

• Fixed backplane

Benefits• Space utilization

• Cable aggregation

• Power efficient

Weakness• I/O flexibility

• Aggregate management

• Large chassis needed to

amortize switch/mgmt costs

Server Deployment: Blades

Rack Rack

Server

ServerS

Infrastructure• Servers & Switches

• Physical & Virtual

Challenges• Many points of management

• Consistent policies

• Diagnostics

• Training

• Security

Management ManagementManagement Management

Server Deployment Scale

Software Switch Software Switch Software Switch

Simplifying the Data CenterMgmt Server

Simplifying the Data CenterMgmt Server Mgmt Server

A cohesive solution

Simplifying the Data CenterMgmt Server

A cohesive solution

Embed management

Simplifying the Data Center

A cohesive solution

Embed management

Unify fabrics

Mgmt Server

A cohesive solution

Embed management

Unify fabrics

Optimize virtualization

Mgmt Server

A cohesive solution

Embed management

Unify fabrics

Optimize virtualization

Remove unnecessary

Switches

Adapters

Management modules

Less than 1/2 the support infrastructure for agiven application

Mgmt Server

Scalable compute platform

Integrated virtualization

Natural aggregation point: Network

Unified embedded management

Embedded on the network controller

Wire once: I/O on demand

LAN, SAN, IPC

Efficient Scale

Cisco network & services scale

Fewer servers with more memory

Lower cost

Fewer servers, switches, adapters, cables

Lower power consumption

UCS Components

UCS ManagerEmbedded– manages entire system

UCS Fabric Interconnect20 Port 10Gb FCoE40 Port 10Gb FCoE

UCS Fabric ExtenderRemote line card

UCS Blade Server ChassisFlexible bay configurations

UCS Blade ServerIndustry-standard architecture

UCS Virtual AdaptersChoice of multiple adapters

UCS Building Blocks

Blade Chassis

6RU Chassis

Blades and Power supplies plug-in from front

Blades

Power & cooling budget allows leading edge processor performance and memory capacity

Combinations of half slot and full slot blades

Up to 8 Half slot blades

Up to 4 Full slot blades

Power Supplies

4x 2,500W hot-plug Power Supplies

90% efficient

N+N redundancy (grid redundant)

No zoning

4 single phase 220V, IEC320-C19

UCS B200 M1 Blade

Blade Attributes:–2 x Intel Xeon 5500 Series Processors–12 x DIMM slots - up to 96GB RAM–2 x optional SAS hot-plug hard drives–RAID 0, 1, 0+1–1 x 10Gb dual port mezzanine adapter–Remote and local access to keyboard, video, mouse, serial–Integrated with UCS Manager

UCS B200 M1 overview

PCIe x16Mezz connector

IOH(I/O Hub)

Memory

ServiceProcessor

LocalHard drives

SAS controllerICH

(I/O CotrollerHUB)

UCS B250 M1 Blade

Blade Attributes:–2 x Intel Xeon 5500 Series Processors–48 x DIMM slots - up to 384GB RAM–2 x optional SAS hot-plug hard drives–RAID 0, 1–2 x 10Gb dual port mezzanine adapter–Remote and local access to keyboard, video, mouse, serial–Integrated with UCS Manager

-- Available Fall --

Network + Compute Virtualization

Single Integrated System

Mgmt SAN ALAN

5 x 8 = 40 5 x 8 = 40 5 x 8 = 40 5 x 8 = 40 5 x 8 = 40 5 x 8 = 40 5 x 8 = 40 5 x 8 = 40 320 Total

Mezzanine Cards (Server I/O Options)

Intel 82598 (Oplin)

Qlogic ISP2432 or Emulex Zephyer

Instantiates Fibre Channel, Ethernet/LAN, virtual NICs

(in conjunction with NIV support from Central Switch )

Oplin (Intel 82598)

3rd party based

MENLO based

MEZZ_CARD

IO_CHANNEL

(single lane used)

(1) x4 PCIe

PALO based

MEZZ_CARD

(1) x16 PCIe

MENLO PALO

3rd party based

MEZZ_CARD

(1) x8 PCIe

LAN NIC

(ex. Oplin)

IO_CHANNEL

(single lane used)

IO_CHANNEL

(single lane used)

(1) x8 PCIe

MENLO based

MEZZ_CARD

IO_CHANNEL

(single lane used)

(1) x4 PCIe

PALO based

MEZZ_CARD

(1) x16 PCIe

MENLO PALO

3rd party based

MEZZ_CARD

(1) x8 PCIe

LAN NIC

(ex. Oplin)

IO_CHANNEL

(single lane used)

IO_CHANNEL

(single lane used)

(1) x8 PCIe

Overall System (Front)

Top of Rack Switch

Chassis

Overall System (Rear)

Uplinks

Memory Expansion

02/10/2009 Cisco Inc., Company Confidential - NDA Required 2727

Savings

Higher server consolidation & larger VM density

Reduces CPU, power/cooling, and SW licensing costs

Competition - max 18 & high density with proprietary tech

Memory Expansion

Cisco Extended Memory solution

Intel’s Nehalem memory controller is designed for future DIMM technology

32GB DIMM

4 x 8GB DIMM

32GB DIMM

Cisco Extended Memory technology makes four DIMMs look like one!

Does not exist

Optimizing Memory with the Xeon 5500

Xeon 5500 Xeon 5500

Xeon 5500Xeon 5500

Classic

Cisco UCS With Memory Extension

• 12 DIMMs• Max 96GB

• Higher Performance

• Lower PerformanceOr

• Higher Performance

Nehalem Memory Architecture

8GB DIMM

6 DIMMs

8GB DIMM

6 DIMMs

96 GB max with 8 GB DIMMs

Nehalem-EP

Processor

Slot 16

Slot 17

Slot 18

Slot 19

Slot 20

Slot 21

Channel 0

(green)

Channel 1

(blue)

Channel 2

(red) 8GB

Slot 22

Slot 23

8GB Slot 8

Slot 9

Slot 10

Slot 11

Slot 12

Slot 13

Slot 14

Slot 15

8GB Slot 0

Slot 1

Slot 2

Slot 3

Slot 4

Slot 5

Slot 6

Slot 7

Expanded Memory Blade

Nehalem-EP

Processor

Slot 16

Slot 17

Slot 18

Slot 19

Slot 20

Slot 21

Channel 0

(green)

Channel 1

(blue)

Channel 2

(red) 8GB

Slot 22

Slot 23

8GB Slot 8

Slot 9

Slot 10

Slot 11

Slot 12

Slot 13

Slot 14

Slot 15

8GB Slot 0

Slot 1

Slot 2

Slot 3

Slot 4

Slot 5

Slot 6

Slot 7

8GB DIMM

24 DIMMs

8GB DIMM

24 DIMMs

384 GB max with 8 GB DIMMs

Virtualized Adapter

What is SR-IOV about?

Single Root IO Virtualization (SR-IOV) allows “virtualizing” the 10 GigE link (via the PCI-Express bus) into multiple “virtual links”.

SR-IOV is a PCI-Sig standard

In other words you can create multiple “vmnics” each with its own bandwidth allocation

Server

Virtual Switch

pNIC: 10 Gbps

This is what SR-IOV enables

This could be Nexus 1000v

10 Gigabit Ethernet

Adapter Offerings

“Free” SAN Access for

any Ethernet equipped host

Software FCoE

Existing Driver Stacks

10GbE/FCoE

PCIe Bus

FC10GbE

Compatibility

VM I/O Virtualization and Consolidation

PCIe x16

10GbE/FCoE

Virtualization

Virtualization Adapter

True wire once architecture – highly dynamic

Network policy and visibility brought to VMs

Hypervisor bypass support – increases performance

Reduce NIC and mezz card infrastucture

Virtual

Machine

Virtual

Machine

Virtual

Machine

Virtual

Machine

Virtual

Machine

Virtual

Machine

Switch

HypervisorHypervisor

Today’s Server Cisco with Palo

Virtualization Adapter

Virtualized adapter designed for both single-OS and VM-based deployments

Provides mobility, isolation, and management from the network

Secure

Transparent to hosts

Cut-through architecture

High Performance

2x 10Gb

Low latency

High BW IPC support

128 vNICs

Ethernet, FC or SCSI

500K IOPS

Initiator and Target mode

California Palo Adapter

PCIe x16

10GE/FCoE

MAC 0 MAC 1

Defineable

Management and service profiles

Server

Identity (UUID)

Adapters

Number

Type: FC, Ethernet

Identity

Characteristics

Firmware/BIOS

Revisions

Configuration settings

Network

Uplinks

LAN settings

etc…

Firmware

Revisions

Storage• Optional Disk usage

• SAN settings

• LUNs

• Persistent Binding

• SAN settings

• vSAN• Firmware

• Revisions

Configuration Points

Blade identities can be duplicated, automatically moved

and deployed, and failed-over to another blade

Firmware and bios included – competition does not do

“Stateless” environment

Significant process/labor savings

Service Profile• Encapsulation of HW state – MAC, WWN, Firmware, BIOS

Service Profiles

Blade identities can be duplicated, automatically moved

and deployed, and failed-over to another blade

Firmware and bios included – competition does not do

“Stateless” environment

Significant process/labor savings

Service Profiles

Service Profile Service Profile

Service ProfileService Profile

Integrated Stateless Computing

Attributes no longer tied to physical hardware

Not just identity

Seamless server mobility

Within interconnect domain

Dynamic Provisioning

Complete infrastructure repurposing

Integrated with 3rd part tools

SAN LAN

Chassis-1/Blade-5

Chassis-9/Blade-2

Server Name: LS-AUUID: 56 4d cd 3f 59 5b 61…

MAC : 08:00:69:02:01:FCWWN: 5080020000075740

Boot Order: SAN, LAN

LANSAN B

Unified Management

hassis

Two Failure Domains

Separate fabrics

Central supervisor, forwarding logic

Distributed Fabric Extenders

Traffic isolation

Oversubscription

Infrastructure Management

Centralize chassis management

Intrinsic system management

Single management domain

Scalable architecture

10GE/FCoE

ChassisManagement

Unified Management

UCS Manager

Single point of device management

Adapters, blades, chassis, LAN & SAN connectivity

Embedded manager

GUI & CLI

Standard APIs for systems management

XML, SMASH-CLP, WSMAN, IPMI, SNMP

SDK for commercial & custom implementations

Designed for multi-tenancy

RBAC, organizations, pools & policies

XML API

Custom Portal

Systems ManagementSoftware

Standard APIs

View 1 View 2

UCS Graphical interface

CONTENT PANENAVIGATION PANE

Top directory map tells you where you are in tree

Multi-Tenancy Model

NetworkManagement

Company

HR Finance

Facilities

Policies

PoliciesServerServer

Server

ServerServer

Server

ServerServer

Server

ServerServer

Server

Compute Blade

Compute Blade Fa

Compute Blade

Compute BladeCompute Blade

Compute BladeCompute Blade Fa

Compute Blade

Compute Blade Fa

Compute Blade

Compute Blade Fa

Compute Blade

Compute Blade Fa

Compute Blade

Compute Blade Fa

Compute Blade

C-Series Rack mounted servers

Compute

Innovations•Extended Memory•Virtualized Adapter•Hypervisor bypass

•Unified Management

Customer Choices for Innovation

Fabric

Innovations•Unified Fabric

•Fabric Extender•VN-Link

Innovations•Unified management

•Unified Fabric •Extended Memory•Fabric Extender

•Virtualized Adapter•Hypervisor bypass

•VN-Link

Unified

Work in any data center environment

Existing mgmt tools

Build Your Own

SMASH CLP, SNMP, IPMI

Options to Deploy Innovation

Integrated

Existing mgmt tools plusUCS Manager

Integrated mgmt tools

Available in CY2010

Key Customer Requirements for Server Infrastructure

Scalable Infrastructure Energy Efficient

Open / Standards-based

Optimized for Virtualization

Easy to Manage

Consolidated Infrastructure

Cisco Solution

Cisco UCS Architecture is Form-Factor NeutralCustomer Has Choice

Whether blade or rack form-factor, Cisco UCS customers benefit from

Consolidated & Unified Infrastructure

Unified Management & Dynamic Provisioning

Virtualization Optimization

Memory extension technology

Blade & Rack serversRack serversBlade servers

Cisco UCS Rackserver Features & BenefitsUCS offers Significant Value

Intel Nehalem EP Processors

RAID, Redundant power & fan

Front accessible HDDs & fans

Memory Extension Technology

Unified Mgmt

VN-LINK Technology

Performance

Availability

Serviceability

Scalability

Manageability

Virtualization Ready

Key Cisco Differentiators

Cisco UCS C-Series Rack-Mount Servers

Item CPU Size Memory Disks Adaptor

UCS C250 M1

(memory intensive)

Intel Nehalem

48 DIMM

384 GB

SAS/SATA

Drives

5 PCIe

UCS C210 M1Intel Nehalem

12 DIMM

SAS/SATA

Drives

5 PCIe

UCS C200 M1Intel Nehalem

12 DIMM

SAS/SATA

Drives

2 PCIe

UCS C200 M1

UCS C210 M1

UCS C250 M1

Cisco UCS C250-M1 Overview2RU Virtualization & Database Server

Form-Factor

2 Rack Units

Processor

2 x Nehalem EP

Memory

Memory Extension Technology

48 DDR3 DIMM slots

Up to 384 GB memory

5 x PCIe slots

4 x 10/100/1000 LOM ports

Up to 8 SFF SAS/SATA Drives

Availability

Redundant Power Supplies & Fans

Serviceability

Front accessible FANs & Disks

Management

2 x 10/100 Mgmt ports

Cisco Lights-out Mgmt

Target Applications: Virtualization, Database and

other Memory Intensive Applications

2RU , 2-socket Server for Enterprise Data Centers

Cisco UCS C210-M1 Overview2RU General Purpose Server

Form-Factor

2 Rack Units

Processor

2 x Nehalem EP

Memory

12 DDR3 DIMM slots

Up to 96 GB memory

5 x PCIe slots

2 x 10/100/1000 LOM ports

Up to 16 SFF SAS/SATA Drives

Availability

Serviceability

Front accessible Disks

Management

1 x 10/100 Mgmt port

Target Applications: Web, App and Infrastructure Applications with High IO

capabilities

2RU , 2-socket Server for DC & Branch/Remote Offices

Cisco UCS C200-M1 Overview1RU General Purpose Server

Form-Factor

1 Rack Unit

Processor

2 x Nehalem EP

Memory

12 DDR3 DIMM slots

Up to 96 GB memory

2 x PCIe slots

2 x 10/100/1000 LOM ports

4 x 3.5” SAS/SATA Drives

Availability

Serviceability

Front accessible Disks

Management

1 x 10/100 Mgmt port

Target Applications: Web, App and Infrastructure

Applications

1RU , 2-socket Server for DC & Branch/Remote Offices

Summary

Announcing Unified Computing System

A single system that unifiesCompute: Industry standard x86

Network: Unified fabric

Virtualization: Control, scale, performance

Storage Access: Wire once for SAN, NAS, iSCSI

Embedded management Increase scalability without added complexity

Dynamic resource provisioning

Ability to integrate with broad partner ecosystem

Energy efficient Fewer servers, switches, adapters, cables

Lower power and cooling requirements

Increase compute efficiency by removing I/O and memory bottlenecks

Sample Case Study Large Enterprise

320 Servers

Time to provision new applications: days to weeks

$21M spent on CapEx

$800K spent on powerand cooling (3 year)

3,520 Cables

31 Racks

Legacy System

320 Servers

Time to provision new applications: minutes

$12M spent on CapEx

$650K spent on power and cooling (3 year)

480 Cables

12 Racks

Unified Computing System

Savings

Business Benefits Unified Computing System

Reduces total cost of ownership

CAPEX: Up to 20% reduction on average

OPEX: Up to 30% reduction on average

Cooling and power efficient

Increases business agility

Provision applications in minutes instead of days

Automation reduces service outages

Just-in-time resource provisioning

Investment protection

Industry standards-based

Co-exist with existing data center infrastructure

Leverage existing management applications via API

Initial Target Customers

Virtualization

IT organizations with large server and data center

virtualization initiatives

Large Data Set

Applications requiring large memory such as business

intelligence, financial and engineering modeling

Service Provider

Hosted solutions, multi-tenancy

Cisco Unified Computing SystemTechnology Partners

Adapter OS Hypervisor Application Mgmt Storage

10 GbE

WindowsServer

SUSELinux

Hyper-V

Oracle VM

SQL Server

Oracle RACOracle DB

BusinessSuite

vCenter

SystemCenter

Smarts

BladeLogic

CLARiion/Symmetrix

SystemCenter FAS

Virtualization has created a market transitionA unique opportunity to create a more agile and efficient infrastructure where resources can dynamically move within the data center

Cisco is uniquely positionedto lead the transformation to the next generationdata center with a new architectural approach

The Unified Computing Systemis a next-generation data center platform thatunites compute, network, storage access,and virtualization into a cohesive system

The Unified Computing System unleashesthe power of virtualization Helping customers to reduce TCO, increasebusiness agility and improve energy efficiency

SummaryThe Right Solution at the Right Time

Registrujte se za Cisco Networkers 25-28. januar 2010. Barselona28-31. mart 2010. Bahrein

Cisco Unified Computing System · Virtualization Step2 10 GE VM VM VM VM VM VM VM VM Hypervisor s...

Documents

SAN DIEGO UNIFIED PORT DISTRICT ACT

Oracle VM 3 Oracle VMのSANを使用したディザスタ・リカバ …...ホワイト・ペーパー – Oracle VMのSANを使用したディザスタ・リカバリ・ソリューションへの統合

San Diego Unified SchoolSan Diego Unified School District Board ...prod031.sandi.net/board/reports/2007/1211/7c_powerpoint.pdf · San Diego Unified SchoolSan Diego Unified School

SAN DIEGO UNIFIED PORT DISTRICT CODE Published Pursuant … · 2020-04-01 · SAN DIEGO UNIFIED PORT DISTRICT . Published Pursuant to San Diego Unified Port District Ordinance 19

SAN FRANCISCO UNIFIED SCHOOL DISTRICT RECORDS … · SAN FRANCISCO UNIFIED SCHOOL DISTRICT RECORDS 1854-2005 ... The San Francisco Unified School District Records were transferred

Hitachi Unified Storage VM Block Module Hitachi Volume ... · Hitachi Unified Storage VM Block Module Hitachi Volume Shredder User Guide ii ... You can shred all ... VM Block Module

SAN JUAN UNIFIED SCHOOL DISTRICT PROCUREMENT …

San Francisco Unified & Hamilton Family Center

Hitachi Unified Storage VM Block Module · 2019-04-23 · Hitachi Unified Storage VM Block Module Encryption License Key User Guide Preface This guide describes and provides instructions

Hitachi Unified Storage VM Block Module Hitachi SNMP · PDF fileHitachi Unified Storage VM Block Module Hitachi SNMP Agent User Guide vi Preface Intended audience This document is

Hitachi Unified Storage VM All Flash System Comparative Guide: Choose the Best All Flash Unified Storage Solution for Your Virtualization Needs

Hitachi Unified Storage VM Block Module Hitachi Thin Image ... · Product Version Contents Getting Help FASTFIND LINKS Hitachi Unified Storage VM Block Module Hitachi Thin Image User

San Diego Unified School District

San Juan Unified School District / Homepage...San Juan Unified School District San Juan Unified School District Risk Management 3738 Walnut Avenue, Carmichael, CA 95608 P. O. Box 477,

SAN MARCOS UNIFIED SCHOOL DISTRICT · PDF filePre-Employment Processing Appointment to complete the process. Welcome to the San Marcos Unified School District! ... San Marcos Unified

San Diego Unified Port District San Diego Unified Port District Development of an Enterprise GIS system

San Francisco Unified School District

Hitachi Unified Storage VM Best Practices Guide for HNAS ... · The Hitachi Unified Storage VM (HUS VM) storage systems can be configured in numerous ways. When attaching a Hitachi

Fast Facts - San Diego Unified

Hitachi Unified Storage VM Block Module · Hitachi Unified Storage VM Block Module ... Cluster system architecture ... HUS VM microcode