June 21 – 23, 2009 Poughkeepsie, New York Converged Networking for Next Generation Enterprise Data...

20
June 21 – 23, 2009 Poughkeepsie, New York Converged Networking for Next Generation Enterprise Data Centers Dr. Casimer DeCusatis Distinguished Engineer IBM Poughkeepsie, NY [email protected] Enterprise Computing Community – ECC 2009

Transcript of June 21 – 23, 2009 Poughkeepsie, New York Converged Networking for Next Generation Enterprise Data...

June 21 – 23, 2009Poughkeepsie, New York

Converged Networking for Next Generation Enterprise Data Centers

Dr. Casimer DeCusatisDistinguished EngineerIBM Poughkeepsie, NY

[email protected]

Enterprise Computing Community – ECC 2009

June 21 – 23, 2009Poughkeepsie, New York

Trademarks

The following are trademarks of the International Business Machines Corporation in the United States and/or other countries.

The following are trademarks or registered trademarks of other companies.

* Registered trademarks of IBM Corporation

* All other products may be trademarks or registered trademarks of their respective companies.

Intel is a trademark of Intel Corporation in the United States, other countries, or both.

Java and all Java-related trademarks and logos are trademarks of Sun Microsystems, Inc., in the United States and other countries

Linux is a registered trademark of Linus Torvalds in the United States, other countries, or both.

UNIX is a registered trademark of The Open Group in the United States and other countries.

Microsoft, Windows and Windows NT are registered trademarks of Microsoft Corporation.

Red Hat, the Red Hat "Shadow Man" logo, and all Red Hat-based trademarks and logos are trademarks or registered trademarks of Red Hat, Inc., in the United States and other countries.

Notes: Performance is in Internal Throughput Rate (ITR) ratio based on measurements and projections using standard IBM benchmarks in a controlled environment. The actual throughput that any user will experience will vary depending upon considerations such as the amount of multiprogramming in the user's job stream, the I/O configuration, the storage configuration, and the workload processed. Therefore, no assurance can be given that an individual user will achieve throughput improvements equivalent to the performance ratios stated here. IBM hardware products are manufactured from new parts, or new and serviceable used parts. Regardless, our warranty terms apply.All customer examples cited or described in this presentation are presented as illustrations of the manner in which some customers have used IBM products and the results they may have achieved. Actual environmental costs and performance characteristics will vary depending on individual customer configurations and conditions.This publication was produced in the United States. IBM may not offer the products, services or features discussed in this document in other countries, and the information may be subject to change without notice. Consult your local IBM business contact for information on the product or services available in your area.All statements regarding IBM's future direction and intent are subject to change or withdrawal without notice, and represent goals and objectives only.Information about non-IBM products is obtained from the manufacturers of those products or their published announcements. IBM has not tested those products and cannot confirm the performance, compatibility, or any other claims related to non-IBM products. Questions on the capabilities of non-IBM products should be addressed to the suppliers of those products.Prices subject to change without notice. Contact your IBM representative or Business Partner for the most current pricing in your geography.

DB2* HyperSwap System z9*Cool Blue IBM* Tivoli*DRDA* IBM logo* WebSphere*DS8000 OMEGAMON* z9ESCON* Parallel Sysplex* zArchitecture*eServer ResourceLink z/OS*FICON* System p ` z/VM*FlashCopy* System Storage z/VSEGDPS* System x zSeries*HiperSockets System z

Enterprise Computing Community – ECC 2009

IBM STG3

Traditional Enterprise Environment

High Performance in a small package Relatively High Utilization of Processor Capacity, virtualization High network bandwidth (8-10 Gbit/s)

Ethernet LAN

Clustering (GDPS)

WAN / WDM

Storage

C. DeCusatis 0509

IBM STG4

. . .

. . .. . .. . .

1-8G SAN SwitchesFICON/Fibre Channel/iSCSI

1/10G Ethernet Switches LAN/NAS/appliances

access, core, aggregation

Clustering Switches(IB, other)

WAN / WDM

A Sprawl of Low Utilization Servers !

3 physical and logical networks (Storage, Ethernet, Clustering) + WAN/WDM

Wastes Energy, High CapEx, High OpEx

Hard to manage, impossible to optimize

Too many moving parts – poor reliability - Does Not Scale

The Problem with Today’s Data Center…IBM Dynamic Infrastructure Leads the Way to a Smarter Datacenter Network

Converged FabricTOR/Core

10/40G FCoCEE

Unified Management

LAN&Storage Re-use16G Storage

Virtual Storage

Integrated Clustering/Appliances

WAN / WDM

1 Network - Simplified and Converged

Virtualized, Integrated, and Easier to Manage

Lower Cost and Energy Efficient

Standards Based with Reuse of Existing Storage Investment & Room to Grow

Enterprise class reliability

Enables More Efficient Cloud Computing & New Functionality

IBM

C. DeCusatis 0509

IBM STG5

Simplification through convergence

Convergence without Rip & Replaceƒ Incremental migration ƒ Co-exist with legacy networks & I/O devicesƒ Ultimate goal is to simplify management

Shared Adapters for LAN/SAN/Clusterƒ Increased adapter utilization through virtualizationƒ Fewer adapters & switchesnReduced costs

ƒ Reduction in infrastructure costs

Reduced CapEx & OpEXƒ Power Savingsƒ Cooling Savingsƒ Space Savings

Converged FabricTOR/Core

10/40G FCoCEE

Unified

Management

LAN&Storage Re-use16G Storage

Virtual Storage

Integrated Clustering/AppliancesWAN /

WDM

C. DeCusatis 0509

June 21 – 23, 2009Poughkeepsie, New York

Evolution of Data Center Switching:reduced layers saves power, cost, latency (higher performance)

. . .

. . . . . . . . .. . . . . . . . . . . . . . . . . .

More powerful microprocessors drive virtualization

Server Virtualization and the resulting consolidation drive need for higher bandwidth

Higher bandwidth connections to network get attached to aggregation/core layer

Hypervisor switching, rack/row level integration consolidates access layer into server

3 layer 2 layer 1 layer

access

aggregation

core

IBM STGC. DeCusatis

7IBM Confidential

IBM Converged Network Roadmap

A family of IBM-Branded Converged Networking Platformsƒ Converged network adapters (CNAs) on IBM server platforms

ƒ Converged Fibre Channel over Ethernet switchesin partnership with other companies nTop of Rack (TOR) and Core Network Building Blocks

>10 Gbps Gen.8-10 Gbps Gen.Existing

Fibre Channel

Ethernet 1Gb1Gb

1,2,4 Gb1,2,4 Gb

Convergence opportunity

40Gb CEE40Gb CEE

16Gb16Gb

10Gb CEE10Gb CEE

ConvergedTop-of-Rack

8Gb8Gb

Infrastructure transitions

NAS, iSCSI

FCoCEE

8

Integrated Clustering/AppliancesWAN / WDM

FICON and FCP

Ethernet Clustering Switches(IB, other)

Standards Based with Reuse of Existing Storage Investment & Room to Grow

FC-aware switches

Converged FCoCEE

Switch ISL

FCoCEE

Evolution of Storage Switching

C. DeCusatis 0509

IBM STGC. DeCusatis

9IBM Confidential

IBM Ethernet Switch Roadmap

A family of IBM-branded Ethernet switches

ƒ Simplifies access to advanced network functions ƒ Advanced features in base products enable forward migration (MPLS)ƒ Offers customers a choice of networking platforms ƒ Standards based for low cost, reuse of existing network,

and future investment protectionƒ IBM tested & qualified to enterprise standardsƒ Installed & supported by IBM dynamic infrastructure servicesƒ Forms the basis for IBM internal cloud computing centers ƒ Positions IBM as an industry leader for integrated networking solutionsƒ Integrated with IBM building block solutions

Systems Management Views

Storage

Subsystems

Hypervisor

Server

VM VMVM

Hypervisor

Server

VM VM VMVM VM

SANSAN

VLAN

Separate Tools and Networks

EthernetEthernet

VLAN

1 Virtualization ManagementVSMHypervisor Management Tool

2 Storage ManagementStorage ManagerVendor ToolSVCBOFM

3 Network ManagementConfiguration ManagerVendor ToolBOFM

VSWITCHVNICVLAN

VM VM

Network Virtualization

FCoCEEFCoCEE

CEE/FCoCEE Server

Single Tool and Network

1 Converged Management

Network Performance Metrics

Network Fault Information

LUN LUN

Hypervisor SVC

CEE/FCoCEE

VSWITCHVNICVLAN

Multiple Management tools/viewsSingle Management view

11

Value added Integration of IBM Systems Director and Tivoli Service Management

Tivoli = Service Management

Integrated visibility, control & automation across heterogeneous business and

technology assetsSee the business

Govern and control the businessOptimize the business

IBM Systems Director = Platform Management

Detailed “care and feeding” of IBM hardwareTell me what I haveLet me install & configure itTell me if it’s workingLet me update it

• Server, Storage, Network Management

• Virtualization support• Launch-in-context integration

of Brocade Manager andTivoli (see below)

Systems Director Network Manager 2.1

• Enterprise Event & Network Management

IBM Tivoli OMNIbus & Network Manager 8.1

June 21 – 23, 2009Poughkeepsie, New York

What is FCoE ?

• Fibre Channel over Ethernet (FCoE)– T11.3 FC-BB-5 standard under development

• August adoption looks promising

• Encapsulate Fibre Channel frames in Ethernet – Desire to converge on a single wire for storage & networking– Leverage years of effort on Fibre Channel Standards– Compatibility with today’s installed FC infrastructure

Enterprise Computing Community – ECC 2009

IBM STG

Fibre Channel over Ethernet Overview FC over Ethernet

– Layers FC frames directly over Ethernet.

– Replacing lowest level of FC with Ethernet

– Requires FC equivalent no-drop behavior

…But… Ethernet needs enhancements, to:

– Provide no-drop behavior in face of congestion

– Manage traffic interferences

EthernetFrame

FCoEEncapsulation

FCPacket

FC-0 (Phy)

FC-1 (Encode/Decode)

FC-3 (Common Services)

FC-4 (FCP for SCSI)

FC-2 (Framing, Flow-Control)

Enet Phy

Enet MAC

FC-3 (Common Services)

FC-4 (e.g. FCP or FICON)

FC-2 (Framing, Flow-Control)

June 21 – 23, 2009Poughkeepsie, New York

CEE Components

• Priority-based Flow Control (PFC) – Improves on current 802.3x PAUSE mechanism by allowing link

level flow control to be applied separately on 8 traffic classes. This is necessary (esp. on FC traffic) to eliminate packet drops due to congestion.

• Enhanced Transmission Selection (ETS) – Supports bandwidth allocation to Priority Groups. Enables

bandwidth sharing between Priority Groups carrying bursty high offered loads. It allows other traffic classes to use the available bandwidth when the offered load in a traffic class doesn’t use its allocated bandwidth. Additionally, it allows strict priority for time-sensitive and management traffic requiring minimum latency

• Data Center Bridging Exchange (DCBX) Protocol – discovery/management framework for CEE-specific features and

used to exchange CEE capabilities between peers. It leverages IEEE 802.1AB Link Level Discovery Protocol for these exchanges.

Enterprise Computing Community – ECC 2009

IBM STG

Ethernet Convergence

Data Center Bridging (DCB) refers to a set of Ethernet enhancements currently being pursued in IEEE 802.1

– Per-priority Flow Control– Enhanced Transmission Selection– Discovery and Capability Exchange– Congestion Notification

Converged Enhanced Ethernet (CEE) is the name folks (EMC, HP, IBM and trade press) have given to the output DCB specifications for the following 3 functionswhich are viewed as required for FC convergence with Ethernet:

– Per-priority Flow Control– Enhanced Transmission Selection– Discovery and Capability Exchange

These are currently being developed by the CEE Authors Group as a “version 0” of the standards to be used as input to IEEE 802.1 (http://tech.groups.yahoo.com/group/cee-authors/)

DCB and CEE LAN Storage Clustering

Streaming Management

Traffic ExamplesDCETM - Cisco marketing term (does not refer to industry standard)

June 21 – 23, 2009Poughkeepsie, New York

What is Cloud Computing?A delivery method, a user experience, and a business model Cloud computing is an emerging style of IT delivery in which applications, data, and IT

resources are rapidly provisioned and provided as standardized offerings to users over the web in a flexible pricing model.

An infrastructure management and services delivery methodology Cloud computing is a way of managing large numbers of highly virtualized resources

such that, from a management perspective, they resemble a single large resource. This can then be used to deliver services with elastic scaling.

Monitor & ManageServices & Resources

CloudAdministrator

DatacenterInfrastructure

Service Catalog,ComponentLibrary

Service Consumers

Component Vendors/Software Publishers

Publish & UpdateComponents,Service Templates

IT Cloud

AccessServices

Enterprise Computing Community – ECC 2009

17 © 2008 IBM CorporationThe New Enterprise Data Center

Network Cloud TT9/21/07

Unified “Cloud” Management

Ensemble

SOA containers

Ensemble

SOA containers

Ensemble

SOA containers

Virtualized Servers, Storage,& Appliances

Unified Fabric Infrastructure

Dynamically Optimized For Business Needs Virtualized and shared Energy efficient and green Optimized service management Secure and resilient

Ensemble

Ensemble EnsembleEnsemble

Ensemble Ensemble

Extended DistanceConnectivity100 km or more

June 21 – 23, 2009Poughkeepsie, New York

Dense Wavelength Division Multiplexers (DWDMs): 32-64 channels or moreCoarse Wavelength Division Multiplexers (CWDM): 8-16 channels or less

WDM Overview

Transmitters Receivers

CombiningSignals

Separating Signals

Transmission Over FiberFICON

Fibre Channel

FCoCEE

FICON

Fibre Channel

Ethernet

FCoCEE

Ethernet

Enterprise Computing Community – ECC 2009

June 21 – 23, 2009Poughkeepsie, New York

Conclusions

• Data center design is undergoing an inflection point– Highly virtualized, scalable– Improved cost/performance – Energy efficiency– Cloud computing principles

• Networking has become a major part of this redesign effort– Convergence of SAN, LAN, and clustering fabrics– Introduction of FCoCEE– Virtualization over distance via WDM– Changing role of the enterprise mainframe and FICON

Enterprise Computing Community – ECC 2009

June 21 – 23, 2009Poughkeepsie, New York

Thank You!