You are on page 1of 71
HP Converged Infrastructure: Reference Architecture Solution Block Design Guide HP Converged Infrastructure delivers the

HP Converged Infrastructure: Reference Architecture Solution Block Design Guide

HP Converged Infrastructure delivers the Data Center of the Future

Table of contents

Introduction

3

The Business Value of Converged Infrastructure

5

The Opportunity

5

Getting Started

7

HP Converged Infrastructure Technology Foundation

8

Design Principles for the Data Center of the Future

8

Theory to Practice: Functional Description of HP Converged Infrastructure

9

Support for Industry Standards

19

Applying the HP Converged Infrastructure Maturity Model

19

HP Converged Infrastructure Reference Architectures

22

Overview

22

Generalized Reference Architecture

22

Virtual Resource Pools

23

FlexFabric

23

Matrix Operating Environment

24

ERP/CRM: SAP Business Suite 7, ERP 6.0 Use Case

28

Solution Architectural Principles

28

Solution Block Architectures

31

Specific HP Products Applied to the General Architecture

34

Messaging & Collaboration: Microsoft Exchange Use Case

39

Solution Architectural Principles

39

Solution Block Architectures

42

Specific HP Products Applied to the General Architecture

45

Virtualization: Generalized Use Case

47

Solution Architectural Principles

47

Solution Block Architectures

48

Specific HP Products Applied to the General Architecture

49

Cloud: Generalized Use Case

55

Definitions

55

Solution Architectural Principles

57

Solution Block Architectures

64

Glossary

69

For More Information

71

1
1

LIST OF FIGURES

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.

Figure 9.

Figure 10.

Figure 11.

Figure 12.

Figure 13.

Figure 14.

Figure 15.

Figure 16.

Figure 17.

Figure 18.

Figure 19.

Figure 20.

Figure 21.

Figure 22.

Figure 23.

Figure 24.

Figure 25.

Figure 26.

Figure 27.

Figure 28.

The Transformation to HP Converged Infrastructure

3

Improving Productivity Through HP Converged Infrastructure Maturity Model

6

HP Converged Infrastructure functional block diagram

11

HP Data Center Smart Grid functional block diagram

14

HP FlexFabric functional block diagram

17

HP Converged Infrastructure Functional Block Diagram

23

HP Converged Infrastructure Reference Configuration

25

SAP Enterprise Resource Planning 6.0 – functional description

28

Core SAP architecture description

29

Resource Consumption and Scaling Capabilities for SAP OLTP Activity

30

High availability SAP system

31

Small SAP Configuration

34

Medium size SAP system

35

High availability SAP system

37

Exchange 2010 enterprise topology

40

Exchange 2010 enterprise topology

45

HP Converged Infrastructure: virtualization – conceptual view

49

HP Virtualization Reference Configuration: Small to Medium Enterprise Build

50

HP Virtualization Reference Configuration: Large Enterprise Build

51

HP Virtualization Reference Configuration: Multi-Data Center Build

53

HP Software Private/Public Cloud Functional Concept Reference Architecture

58

HP Data Center Automation Center

60

HP Operations Manager i

61

Lifecycle Automation with HP Operations Orchestration

63

Network Lifecycle Automation

64

HP Private Cloud Infrastructure detailed architecture for HP BladeSystem Matrix

65

HP Private Cloud Reference Architecture

66

Cloud Example Configuration

68

LIST OF TABLES

 

Table 1.

HP Converged Infrastructure Elements

10

Table

2.

Network requirements

16

Table 3.

Scope and Stages of HP Converged Infrastructure Maturity Model

19

Table 4.

Scoping and mapping of stages of maturity to data center capability and HP products

20

Table 5.

CI MM stage and representative HP products

21

Table 6.

Elements of the HP Converged Infrastructure Reference Architecture

26

Table 7.

Service Definition Criteria for an SAP System

32

Table

8.

High

availability SAP system

33

Table

9.

Tiers

of service

41

Table 10.

Table 11.

Exchange 2010 server role resource requirements

43

Common HP Operations Orchestration usage scenarios

62

Revision History

Version. No.

Date

Comments

1.0

November 2007

HP Solution Blocks

2.0

September 2010

HP Converged Infrastructure

Introduction

Information Technology as a “service” has moved from concept to reality. Early adopters have already deployed major solutions, and it has become a standard objective for mainstream Information Technology (IT) architects and planners. HP has adopted the term “Converged Infrastructure” to describe how HP products and services can address this approach. This Reference Architecture guide provides a business and technical view of the adoption process.

Today’s IT stakes are larger than applications and platforms. They span the data center, from capacity to technology, processes, people, and governance. HP Converged Infrastructure opens the door to new approaches, and can enable IT management to defer or avoid costly data center expansions. For example:

Simplification: Collapse siloed, hierarchical, point to point infrastructure into an easily managed, energy-efficient, and re-usable set of resources.

Enabling growth: Efficiently deploy new applications and services, with optimum utilization across servers, storage, networking, and power.

On-demand delivery: Deliver applications and services through a common framework that can leverage on-premise, private cloud, and off-premise resources.

Employee productivity: Move human capital from operations to innovation by increasing automation of application, infrastructure, and facility management.

HP Converged Infrastructure enables organizations to achieve these goals while getting ahead of the growth curve and the cost curve. Figure 1 illustrates how HP Converged Infrastructure transforms the traditional data center into a unified, flexible resource for business.

data center into a unified, flexible resource for business. Figure 1. The Transformation to HP Converged

Figure 1.

The Transformation to HP Converged Infrastructure

HP Converged Infrastructure Today’s Datacenter Data Center of the Future Hierarchical Network Logical Switch
HP Converged Infrastructure
Today’s Datacenter
Data Center of the Future
Hierarchical
Network
Logical Switch
Discrete
Flat Network
Switches
1-1 Cabling
Server Pool
Virtual
Orchestrated
Resource
Automated
Pool
Provisioning
One-off
Provisioning
Storage Pool
Integrated Service
Management and
Security
Discrete Distributed Resources
Poorly Integrated Power
Optimized Power
Cooling Infrastructure
Cooling Infrastructure

Attaining this goal is achievable today, with this guide as the roadmap. Configurations based on HP BladeSystem c-Class, Superdome 2, HP StorageWorks, and HP Intelligent Resilient Framework provide a flat network and resource pool suitable for the enterprise. In the section “Converged Infrastructure Reference Architectures” these building blocks and others are assembled to address key enterprise workloads. This HP Converged Infrastructure architecture can be deployed with the approach taken by HP IT internally, and large HP customers:

1. Start from existing investments, whether from HP or others.

2. Extend those investments with standards-based components and building blocks from HP that have already been designed to support a “single pane of glass” management philosophy.

3. Pool those resources for allocation “as needed, when needed” adjusting to business requirements.

4. Converge the infrastructure while sticking to your priorities. HP already works with major vendors to verify that their application, operating system, virtualization and other products can be managed and optimized.

This sequence of this guide addresses the information needed by key IT stakeholders across the project lifecycle; starting with planning, then building, and finally running infrastructure.

“The Business Value of Converged Infrastructure” provides senior managers and planners with a look at how HP Converged Infrastructure solves real business problems today. It is also useful to IT management and staff, as it frames the business objectives that technology must address.

“HP Converged Infrastructure Technology Foundation” describes the core elements and architectural principles. It provides architects and designers with functional descriptions as well as a discussion of the HP Converged Infrastructure Maturity Model, and how the model accelerates IT results.

“HP Converged Infrastructure Reference Architectures” provides real instances of HP Converged Infrastructure. It starts with the “core” Converged Infrastructure architecture, and then builds with functional architecture, product mappings, and deployable, application-specific configurations. Deployment and operations specialists can use it as their initial reference when considering solution alternatives for upgrading or transforming data center capabilities.

Along with this comprehensive Converged Infrastructure model, HP offers tools, services, and workshops to support well-informed decisions. For example:

The HP Converged Infrastructure Maturity Model can help you assess the current and desired future state of your data center infrastructure, people, processes and governance. The result is a customized, action-oriented, high-level roadmap for structured improvement.

Data Center Transformation workshops can help define projects in the areas of consolidation, energy & space efficiency, automation, and business continuity & availability.

HP Application Modernization solutions can help visualize legacy application evolution. This creates opportunities for infrastructure renewal and determines the right legacy application transformation strategy for the business.

Proven Private Cloud and Enterprise Services that can help design, build, manage, and run the converged infrastructure.

HP Financial Services are available to transition from existing technology, acquire new solutions, manage those solutions through the lifecycle, and then retire technology. HP Financial Services can also show how HP Converged Infrastructure can reduce capital expenses.

Regardless of where an organization is in this process, HP Converged Infrastructure can be delivered as needed and at the required pace. It provides a solid foundation for real IT solutions, and an IT approach that extends the life of the data center, while enhancing agility and business value. Please consult your HP specialist or HP certified partner for assistance on your specific project.

The Business Value of Converged Infrastructure

The Opportunity

Infrastructure convergence has emerged as possibly the most important trend in decades – creating a mass ripple effect that has brought our industry to a major inflection point in how we approach IT. In today’s business landscape, everything is “accelerated” and “exponential” – across users, competitors, internal organizations, investors, and on and on. The value of IT or Data Center convergence has far reaching benefits across all these communities.

Your business customers expect IT costs and resources to be in-line with the business cost envelope while ensuring tighter compliance amidst tougher regulations. They also demand rapid IT results – not weeks or months, but days, hours, or even minutes. And they want all this with functionality, quality and security, and the option to change/add resources on demand.

Taken together, these incremental investments are forcing hard customer decisions on how to extend the life of the data center. The cost of any individual application pales against the cost of moving or expanding a data center. HP Converged Infrastructure delivers technical innovation by collapsing infrastructure and related costs and potentially avoiding or deferring on-premise facility investments.

Restructuring around services

In addition to addressing issues in the data center itself, IT needs to be able to freely and cost- effectively innovate for business advantage. It’s no longer about being an efficient world-class IT organization, rather about being a true partner in the business – responding quickly to business demands and bringing in new innovation for growth. IT becomes a service broker, building a service delivery channel that can match SLAs and the choice of services to the desired business outcomes for each customer.

Unfortunately, change does not usually happen easily. IT Sprawl is real: resources that are tangled up in legacy applications and architecture silos that demand dedicated infrastructure. Add in the new wave of applications – social networking, mobility, media, etc. - yet another set of unique and manually provisioned applications.

Because of IT sprawl, most organizations are spending upwards of 70% of the IT budget on operational costs while business innovation is throttled down to 30%. Even worse, about half of this 30% is allocated to application upgrade cycles. That leaves only about 15% of the budget for true innovation.

The bottom line: It has never been more strategic or critical to rethink and transform the data center. Innovation is an imperative for business and IT to survive and thrive.

Convergence delivers a services-oriented infrastructure so that IT organizations can reallocate staff resources from operations to innovation. HP Converged Infrastructure gives you rapid provisioning and flexible resourcing to match IT supply with fluctuating demand. It frees-up funds trapped in operations to reduce the burden of legacy and mission critical systems and reclaims the funds for innovation. Adopting the HP Converged Infrastructure Maturity Model as a roadmap to renewing data center operational skills can provide a huge execution advantage, as well as lead to improvements that extend the life of the data center. It covers a broad range of capabilities, as shown below.

Figure 2. Improving Productivity Through HP Converged Infrastructure Maturity Model Definitions CI MM - Stages

Figure 2.

Improving Productivity Through HP Converged Infrastructure Maturity Model

Definitions CI MM - Stages of Maturity Stage 5: • Automatic management and reallocation of
Definitions
CI MM - Stages of Maturity
Stage 5:
• Automatic management and reallocation of a converged infrastructure
to support services based on business process needs
Adaptively sourced
• IT becomes a trusted business innovation partner
Infrastructure
Stage 4:
• Infrastructure offered as a service
Automated Service
• Tiered service levels, supported by service-centric integrated IT processes
Oriented
• Rationalized technologies, architecture, management tools and processes
Stage 3:
• Cross functional IT infrastructure expert teams
Optimized
• Standard technologies, architectures, management tools and processes
Stage 2:
• Technology-based
Standardized
• Project-based decisions and dedicated infrastructure
Stage 1:
• Ad Hoc management tools and processes
Compartmentalized,
• Technology-based IT organization / cost center managed to budget
Legacy
Maturity
/ cost center managed to budget Legacy Maturity Where the savings come from HP Converged Infrastructure

Where the savings come from

HP Converged Infrastructure can help businesses overcome the confines of IT sprawl and shift resources from operations to innovation. HP Converged Infrastructure delivers the data center of the future, today. The foundation technologies described later in this guide have all been built with convergence in mind. They all reflect industry standards, leadership intellectual property, openness, and of course, customer choice. These result in fundamentally lower costs, and a lot less infrastructure. Here a few of the many examples of how HP solutions can simplify your data center and deliver the cost savings with which to fund innovation:

HP Intelligent Resilient Framework merges standalone switches to form a flat, unified fabric, eliminating hops and packets by the billions.

HP Virtual Connect FlexFabric reduces server edge infrastructure by 95%. Virtualized I/O reduces the need for large numbers of high-bandwidth connections while providing better performance.

HP Integrity servers provide 2-1 core count consolidation over older servers, and each core can support up to 20 virtual machines. Racks of test and development servers simply disappear.

HP Insight Dynamics lets you do drag and drop provisioning of complex, multi-tier applications. Cut your provisioning time and errors by a factor of 10.

HP Intelligent Power Distribution reports the power you actually use, device by device. Dynamic Power Capping can reduce infrastructure power cost per server up to 69%.

With HP Converged Infrastructure, you get a unified, service-ready, and shared infrastructure that can dynamically orchestrate and provision server, storage, and networking efficiently. As resources are allocated for specific workloads, they are also rapidly scalable and optimized for energy-efficiency, high availability and utilization. The bottom line is extending the life of the data center. This means delivering more business value, while using less infrastructure, and power, and investing more in innovation instead of operations.

Getting Started

The next sections will dive deeper into HP Converged Infrastructure technology, a “core” framework for using these technologies, and several solution use cases built upon this foundation. There are two broad scenarios for taking advantage of HP Converged Infrastructure.

A transformational approach that establishes a platform for consolidation, greenfield, or private cloud initiatives. HP BladeSystem Matrix supports HP-UX, Linux and Windows, and is an excellent solution for these types of projects. It provides a ready-to-go converged infrastructure that is quickly deployable to support your preferred hypervisors, operating systems, applications, and workloads.

A building block approach that is based on standards. This can gain ground on IT sprawl with servers, storage, network, management designed for convergence and greater investment protection.

For both of these approaches, HP offers preconfigured solutions such as HP BladeSystem Matrix, including HP BladeSystem Matrix with HP-UX, and HP Performance Optimized Datacenter (HP POD). This innovative solution is a fully equipped datacenter in a container, ready for your applications.

Consider how the HP Converged Infrastructure strategy and portfolio can meet and exceed your next set of objectives and requirements. Take advantage of these proven tactics for generating immediate returns, while laying a solid foundation for the future:

Identify the root cause of innovation gridlock and prioritize your projects that attack the top opportunities first.

Define projects that either build upon what you already have or evaluate alternatives for greenfield initiatives where the shared service model best serves the business.

Possibly create self-funding projects though financing options that allow transformation within your current budgets.

Apply the Converged Infrastructure Maturity Model assessment tools to help you identify ‘best fit’ convergence starter opportunities and a training path to build expertise.

Begin to build-out your shared services roadmap. This also could help profile the ideal projects and logical starting points.

Use vendors that can help architect solutions that can be changed quickly and easily – adding new functionality as needed. This validates today’s innovation doesn’t become tomorrow’s legacy.

Read on to better understand the new tools you can apply – or call your HP sales representative or authorized reseller to arrange a planning workshop.

HP Converged Infrastructure Technology Foundation

This section covers the basic design architecture of the converged infrastructure data center, based on HP developed models and available HP and partner technologies. This is followed by a description of the HP Converged Infrastructure approach. The functional shared service capabilities of HP Converged Infrastructure provide the foundation for architecting and evolving data center infrastructure based on best practices developed from decades of successful data center transformation and shared service deployments implemented by HP professionals.

Design Principles for the Data Center of the Future

The solution to IT sprawl is to break down the technology silos and bring all compute, storage, and networking resources together as a common pool. Resources can then be dynamically provisioned and shared by many applications and managed as a service. This approach brings together management tools, policies, and processes so that resources and applications are managed in a holistic, integrated manner. It also brings together power and cooling management capabilities so systems and facilities work together to extend the life of the data center.

The HP Converged Infrastructure has five overarching design principles that are important to IT and together enable the data center of the future. The infrastructure must be virtualized, resilient, orchestrated, open, and modular. Each is described briefly below and then detailed through the remainder of this section.

Virtualized

HP Converged Infrastructure benefits from the virtualization of resources: compute, storage, networking, applications, and operating systems. Virtualization provides an abstraction layer between the physical and the logical, making it easier and faster to reallocate resources to match the changing performance, throughput, and capacity needs of individual applications. This end-to-end virtualization improves IT flexibility and response to business requests, improving business speed and agility. Virtualization covers all key aspects of the infrastructure. This includes client desktops, networking, servers, memory, and storage.

Specific aspects of virtualization include:

Compute resources. Pools of modular server resources which can be virtualized using technologies such as VMware™ vSphere™, Microsoft™ Hyper-V™, and HP Integrity Virtual Machines.

Storage resources. Storage that is virtualized at the array, network, and file system level enables higher utilization of existing capacity, eliminates stranded capacity, enables tiering which lowers costs for new capacity, and provides a non-disruptive way to respond to data growth.

Network resources. Virtualized networks provide the ability to quickly provision service-specific classification and prioritization based on the physical and virtual workloads from the entrance point of the physical or virtual server edge extended throughout the network fabric to ensure consistent application end-user experience. This approach enables simplified, consistent service delivery and improved business agility. Virtualized configuration and management of the extended network to increase performance through higher utilization. Reduce the cost of network operations by reducing cabling-related provisioning costs and downtime.

Resilient

HP Converged Infrastructure integrates fault tolerant mission-critical technologies and high availability policies. Because diverse applications share virtualized resource pools, the infrastructure must have an operating environment that automates high-availability policies to meet SLAs and provides the right level of availability for each business application.

Security is also a key enabler of resiliency; the infrastructure must support policy-based security models, continuous intrusion prevention, and automated updating of security solutions. It must ensure the integrity and availability of applications, systems, and data with security solutions that are unified across physical and virtual server domains. This is essential to ensure threats can be blocked before they can proliferate over the network.

Resiliency is derived from the compute, storage and network resources and processes that provide failover, high availability and disaster recovery based on business requirements.

Orchestrated

HP Converged Infrastructure orchestrates the requests of the applications, data, and infrastructure, defining the policies and service levels through automated workflows, provisioning, and change management. Orchestration provides an application-aligned infrastructure that can be scaled up or down based on the needs of each application. Orchestration also provides centralized management of the resource pool and facilities, including billing, metering, and chargeback for consumption.

For example, orchestration should reduce the time and effort for deploying multiple instances of a single application. In addition, as the requirement for more resources or a new application is triggered, automated tools perform tasks that before could only be done by multiple administrators operating on their individual piece of the physical stack.

Open

HP Converged Infrastructure is based on industry standard technologies. This ensures customers can leverage existing investments as part of the consolidation and convergence process. Industry standards enable customers to adopt new technologies incrementally and at their own pace, and provide the technology openness they require to incorporate heterogeneous infrastructure.

Modular

HP Converged Infrastructure is built on modular design principles based on open standards, allowing for interoperability. The modular approach allows IT to integrate new technologies with existing investments, and provides infrastructure extensibility for the future.

The modular approach should remain open, so that special-purpose or existing infrastructure investments from multiple vendors can be managed under a common umbrella, and share resources with other elements of the pool.

Theory to Practice: Functional Description of HP Converged Infrastructure

HP Converged Infrastructure embodies the design criteria outlined above and integrates a rich set of technologies into a seamless, standards-based infrastructure. This approach enables IT organizations to deploy applications with minimal constraints on “best of breed” infrastructure technologies. The HP Converged Infrastructure architecture stack is illustrated below and includes three categories of products:

Core products that is essential to delivery of Converged Infrastructure services. These products provide the glue which enables rapid, virtualized, application deployment.

Supporting products that can be deployed in standalone configurations. These products have been designed to take full advantage of shared service capabilities.

Complementary products such as those from third party providers. HP Converged Infrastructure provides solid interoperability with, and management of, complementary products.

Table 1. HP Converged Infrastructure Elements Architecture Core Supporting Complementary Element Pillar

Table 1.

HP Converged Infrastructure Elements

Architecture

Core

Supporting

Complementary Element

Pillar

Element

Element

Virtual Resource Pools:

HP BladeSystem c-Class

HP ProLiant blades

Windows

Compute

HP BladeSystem Matrix (x86 and Integrity)

HP Integrity blades

Linux

HP Superdome 2

HP-UX

 

HP ProLiant servers

OpenVMS

HP NonStop Integrity BladeSystem

NonStop

Virtual Resource Pools:

HP SAN storage: P4000, EVA, XP

HP SAN Virtualization Services Platform

Third party storage

Storage

HP NAS storage: X9000

(SVSP)

HP EVA Cluster

HP StoreOnce

deduplication

HP Storage Essentials

HP Storage Networking

Data Center Smart Grid: Power management

HP Insight Control

iLO3 support

Third party facility

Data Center Environmental Edge

Sea of sensors

management systems

Intelligent Power Discovery

FlexFabric: Networking

HP Virtual Connect

HP TippingPoint

Third party networking products

A-series switches

Network Security

Virtual Connect Enterprise Manager

solutions

 

Matrix Operating

HP Insight Dynamics

HP Operations

VMware

Environment:

Orchestration

Management

Orchestration

Insight Dynamics – VSE

HP Insight Control

HP Systems Insight Manager

Microsoft Hyper-V

Citrix …

Insight Manager • Microsoft Hyper-V • Citrix … The following figure illustrates how the four pillars

The following figure illustrates how the four pillars of HP Converged Infrastructure map to the solution elements listed above in the context of a functional block diagram.

Figure 3. HP Converged Infrastructure functional block diagram FlexFabric • Virtual Connect [Flex-10] • Intelligent

Figure 3.

HP Converged Infrastructure functional block diagram

FlexFabric • Virtual Connect [Flex-10] • Intelligent Resilient Framewk • TippingPoint Security Matrix Operating
FlexFabric
• Virtual Connect [Flex-10]
• Intelligent Resilient Framewk
• TippingPoint Security
Matrix Operating
Environment
• Insight Dynamics
• Insight Control
OS
Hypervisor
Virtual Resource Pools
• SIM
App
• ProLiant servers
• Integrity servers
• BladeSystem
Compute &
Storage
• EVA & SVSP
Resources
• P4000
• X9000
Power & Capacity
Mgmt
Data Center Smart Grid
• Thermal Logic
• Insight Control
• Environmental Edge
Converged Networking
Orchestration
• Environmental Edge Converged Networking Orchestration The sections that follow detail functional capability,

The sections that follow detail functional capability, product mappings and advantages for shared services provided for each of the functional areas.

Resource Supply: HP Virtual Resource Pools

HP Virtual Resource Pools are derived from common modular infrastructure that enables the creation of shared capacity that can be combined, divided, and repurposed to match application demand more effectively. Using HP Converged Infrastructure, IT organizations can support business users through a common pool of virtualized resources that can are configured “on the fly” for different needs: e.g. core business, shared services, mission critical, cloud, and high-performance computing (HPC) applications. HP Virtual Resource Pools leverage the innovative capabilities of servers, storage, mission-critical systems, and scale-out technologies across the HP ProLiant, Integrity, Integrity NonStop, and StorageWorks portfolios. The resources can be provisioned to support third party virtualization environments such as VMware and Microsoft, as well as HP solutions such as HP Integrity Virtual Machines.

Business Challenge

While server virtualization has become widely adopted, there is still much efficiency to be gained by applying the lessons learned to other areas of the physical infrastructure. In addition, the adoption of virtual machine technology has begun to lead to ‘Virtual IT Sprawl’ where machine images are created indiscriminately and without a clear resource strategy.

Some specific server and storage challenges include:

Improve time to deployment and reduce wasted capacity.

Adhere to Service Level Agreement metrics such as response time, maintenance windows, and application availability.

Address power and cooling challenges as well as substantial facilities costs.

Reduce capital and operational costs.

Reduce the impact of data growth on storage investment for traditional and emerging data types.

Improve business continuity and availability for critical applications.

Simplify the deployment of storage for business applications.

Reduce wasted capacity and isolated, independent data repositories.

Considerations and best practices:

There are many best practices to consider when moving from isolated islands of compute and storage towards virtual pools of capacity. When evaluating server virtualization it is important to:

Make compute resources available as either virtual or physical resources.

Provide a mix of resources to maximize deployment efficiency AND meet business needs.

Leverage hypervisors to distribute workloads efficiently and rapidly.

Apply templates to ensure consistent deployment images for replicated workloads.

Some additional considerations apply when eliminating stranded storage capacity:

Utilize shared storage such as Network Attached Storage (NAS) or Storage Area Network (SAN) unless application specific requirements indicate otherwise.

Consider storage virtualization technologies to improve utilization and enable granular “pay as you grow” investments that are non-disruptive.

Leverage management tools to automate storage provisioning, reduce operational costs, and provide storage provisioning consistency.

Tier storage to improve ROI of storage investments.

HP Converged Infrastructure Solution

HP provides a comprehensive portfolio of server resources that leverage a range of hypervisor and operating system technologies.

HP ProLiant servers include blade and rack-mount 2, 4 and 8-socket x86 servers These support all major commercial hypervisor products, Microsoft Windows, and major commercial Linux distributions.

HP Integrity servers provide for a mission-critical Converged Infrastructure and include blade and rack mount form factors. HP Integrity server blades with HP-UX combine 2, 4, and 8 socket systems with Blade Link technology. HP Integrity Superdome 2 with HP-UX provide scale as you grow flexibly up to 64 processors. HP Integrity NonStop BladeSystem includes fault-tolerant NonStop software with the scalability expected in NonStop solutions.

The Integrity and ProLiant server blades share the same BladeSystem c-Class family of enclosures.

HP StorageWorks products deliver a range of storage solutions for midrange, enterprise, and mission- critical needs, and also provide strong integration with existing heterogeneous storage products. They provide a rich storage resource pool for HP Converged Infrastructure, such as these examples:

HP StorageWorks Enterprise Virtual Array (EVA) and XP disk arrays deliver virtualized Fibre Channel storage with highly efficient management tools.

HP StorageWorks SAN Virtualization Services Platform (SVSP) delivers network-based virtualized storage and provides a mechanism to combine HP and 3rd party storage into a single virtual pool.

HP StorageWorks P4000 products deliver virtualized iSCSI storage using a clustered scale-out architecture for improved reliability and non-disruptive growth.

HP StorageWorks X9000 delivers scale-out NAS file system virtualization and supports extremely large file namespaces exceeding 16 petabytes for large content depots.

Power and Facilities Management: HP Data Center Smart Grid

A data center is a highly complex ecosystem filled with IT hardware and racks, connected by miles of

wires and cables, with complex relationships between hardware and software. It intersects with another complex system, the data center facility itself, which delivers power, cooling and floor space

for IT. For reliability and availability, both infrastructure and the facility are usually over-provisioned.

A fixed energy and cooling buffer is usually allotted to make sure that critical peaks of IT usage do

not result in IT meltdowns. Plus, power and cooling is “fixed,” usually supplying a constant amount to

IT—no matter the data center status or workload.

Business Challenge

In most ways, power usage in current data centers is provisioned in an identical manner to compute

and storage: sized for peak, and typically largely underutilized. It is also difficult to track and

manage, leading to accidental shutdowns of critical resources.

Over-provisioning leads to several major issues:

Power distribution infrastructure costs are increased.

Data centers can be rated as out of capacity well before actual physical limits are reached.

Cooling is over-delivered, increasing cooling plant and utility costs.

Individual infrastructure components also contribute to excess power usage:

Inefficient power supplies.

Limited control over fans or power to processors, memory, storage, and network devices.

Considerations and best practices

Map and monitor power consumption to rack capacity.

Manage power consumption to reflect changing infrastructure power needs.

Use a TCO model to understand the cost of powering and cooling of servers, both existing and future systems.

Map, monitor and manage facility cooling to match infrastructure cooling needs.

When planning server deployments, don’t use device power consumption estimates. Understand actual usage where possible, which is not always reflected in nameplate specifications. HP equipment will tell you exactly how much power is being consumed and lets you control power use accordingly.

Tightly manage power delivery configurations to eliminate errors during configuration changes.

Understand data center thermal trends and airflow.

Deploy shutdown procedures that prioritize availability for mission-critical applications. Have a plan for infrastructure failure and maintenance planning.

HP Converged Infrastructure Solution

HP Data Center Smart Grid integrates environmental management, facilities management, and system management. It creates an intelligent, energy-aware environment across the infrastructure and facilities to optimize and adapt energy usage, to reclaim facility capacity, and to reduce energy costs. HP Data Center Smart Grid collects and communicates thousands of power and cooling measurements across systems and facilities in real time, giving your organization greater insight and control over energy use. This lets you support business growth by deploying more infrastructure within the same data center footprint, hosting more applications and making more effective use of your existing capacity and capital investments.

HP Data Center SMART GRID includes sensors, integrated controls, and planning and management tools at the chassis, rack and facility level. Key products and capabilities are shown in Figure 4.

Figure 4. HP Data Center Smart Grid functional block diagram ANALYSIS & PLANNING Facility-Level Records

Figure 4.

HP Data Center Smart Grid functional block diagram

ANALYSIS &

PLANNING

Facility-Level Records database Analysis and reporting Cooling optimization

Resource Pool - Level HP Smart Solver HP Capacity Advisor

POWER

MANAGEMENT

HP Environmental Edge

Wireless sensor grid Cooling policy management Integration with building systems

Facility Power and Cooling Rack Management Rack Rack I-PDU Chassis Chassis
Facility
Power and
Cooling
Rack
Management
Rack Rack
I-PDU
Chassis
Chassis

HP Insight Control

Intelligent Power Discovery Data Center Power Control Insight Control Power Mgmt

HP Thermal Logic

Dynamic Power Capping Common Slot Power Supply Sea of Sensors Thermal Logic Usage Reporting

6

Supply Sea of Sensors Thermal Logic Usage Reporting 6 Facility–level capabilities HP Data Center Environmental

Facility–level capabilities

HP Data Center Environmental Edge deploys a set of sensors within the data center to establish demand requirements. With instant, accurate measurements, you have the hard data you need to optimize and control your power and cooling efficiency. HP Environmental Edge integrates with leading facility management software to provide closed-loop management of cooling resources.

Data Center–level capabilities

HP Insight Control provides the umbrella software management facility. Products such as Data Center Power Control work within the framework of HP Insight Control to provide administrators with information on power consumption that can be mapped to application demand.

Data Center Power Control is a feature of HP Insight Control power management software. It allows data center administrators to define rules to handle power and cooling emergencies.

Intelligent Power Discovery (IPD) maps and reports power requirements for each data center device, thus providing the power management ‘bridge’ between data center and IT operations. IPD allows data center operations to tune cooling capability and manage the power grid.

Key capabilities include:

Associate servers to power circuits & improve power planning with precision that’s calibrated to real-time power demands, not estimates.

Show available power capacity to easily identify where to deploy new servers, and identify electrical and thermal overloads so you know which servers are at risk.

Integration with reporting tools enables administrators to visualize the data center thermal status, locate hotspots and identify at-risk services.

Find outlets that are providing power but are not correctly configured & identify erroneous data between the actual power state and manually entered or imported data.

Device–level capabilities

HP Thermal Logic technology is integrated with current ProLiant and Integrity server products, and will be incorporated into future releases of StorageWorks and HP Networking products. By reducing power usage on the compute infrastructure side, equivalent power can also be saved on the cooling side.

The HP Sea of Sensors is a collection of 32 smart sensors embedded at the device level to track thermal activity and enable automated power management. The sensors dynamically adjust system components such as fans, memory and processing to optimize system cooling and increase efficiency.

HP Power Capping and HP Dynamic Power Capping are implemented in system hardware and firmware and are therefore not dependent on the operating system or applications. Using the power monitoring and control mechanisms built into ProLiant servers, power capping is specifically designed to allow an administrator to limit, or cap, the power consumption of a server or group of servers. This provides increased flexibility in data center planning by allowing the administrator to manage parameters that are directly influenced by server power consumption. Power capping also allows the administrator to control server power consumption in emergency situations such as the loss of primary AC power.

HP Common Slot Power Supply gives customers the option to easily customize servers with the optimal power supply. By providing three off-the-shelf power supplies with different capacities, it enables customers to easily make changes at the chassis level to meet their changing resource needs.

Converged Network: HP FlexFabric

HP FlexFabric is a highly scalable data center fabric architecture and a technology layer in the HP Converged Infrastructure. This open architecture uses industry standards to provide high performance, secure, multiprotocol connectivity to storage and server resources using a single network fabric and wire-once simplicity.

FlexFabric combines intelligence at the server edge with an advanced orchestration and management layer to enable virtualization-aware networking, predictable performance, and rapid, secure, business-driven provisioning of data center to avoid problems of latency and network contention. The key products that deliver FlexFabric are positioned in a typical networking environment in Figure 5. Capabilities of specific products and enhancements from standard network hierarchies are discussed below.

Business Challenge

Faster time to application deployment.

Reduce complexity and lower network-related costs associated with delivery of applications.

Ensure network security.

Ensure performance and availability as workload is scaled.

Considerations and best practices

Converge and virtualize storage and server edge network connectivity.

Extend the size of the available network resource pool by implementing a ‘flat’ network design using large scale layer 2 domains.

Implement two-tier network design and switch virtualization technologies (e.g. Intelligent Resilient Framework – IRF) to improve performance, reduce latency, and reduce management complexity.

Optimize network resource to make effective use of high-throughput technologies.

Reduce the requirements for point to point wiring to support individual workloads and to reduce provisioning errors.

Deliver threat management that unifies security for virtual and physical workloads.

Provide support for multiple protocols (Fibre Channel, Fibre Channel over Ethernet, iSCSI, SAS) to simplify network infrastructure.

Dynamic bandwidth allocation on the fly.

The standard services provided by the network are illustrated in Table 2.

services provided by the network are illustrated in Table 2. Table 2. Network requirements Network Service

Table 2.

Network requirements

Network Service

Requirement

WAN Connectivity Layer

Routing to Global MPLS WAN, ISP’s

Application Optimization

Protocol Optimization of Selected Applications

Data Center Core Routing Layer

Routing to individual Cells and adjacent Data Center’s

Intrusion Prevention Systems

Intrusion Prevention for PCI Applications, network uptime and data protection

Distribution Layer

Resilient multi-path L2/L3 fabric (TRILL/VPLS/IRF)

Embedded Security

Physical and Virtual Firewall and IPS

Application Load Balancing

Application specific Load balancing services

Server Edge Virtualization/Switching

Edge connectivity for Servers and Storage

Enterprise Instrumentation

sFlow, consistent policy deployment throughout physical and virtual network fabric, detailed troubleshooting facilities embedded in the network fabric

I/O Consolidation

FCoE, DCB, iSCSI

fabric, detailed troubleshooting facilities embedded in the network fabric I/O Consolidation FCoE, DCB, iSCSI 16

HP Converged Infrastructure Solution

HP FlexFabric provides a portfolio of solutions that support traditional 3-tier and advanced, simplified 2-tier network topologies. Mapping of the previous table to HP 2-tier networking model is shown in below.

simplified 2-tier network topologies. Mapping of the previous table to HP 2-tier networking model is shown

Figure 5.

HP FlexFabric functional block diagram

Intelligent Management Center (IMC) HP Virtual Connect Enterprise Manager

Center (IMC) HP Virtual Connect Enterprise Manager Network Management Backbone Interconnect Server Edge
Network Management
Network
Management
Backbone Interconnect Server Edge Network
Backbone
Interconnect
Server Edge
Network
HP TippingPoint Network Security
HP TippingPoint
Network
Security
Server Edge Network HP TippingPoint Network Security Storage Network Backbone • HP A6600/A8800 • HP
Server Edge Network HP TippingPoint Network Security Storage Network Backbone • HP A6600/A8800 • HP
Storage
Storage

Network Backbone

HP A6600/A8800

HP A12500

Network Interconnect

HP A9500

HP A12500

Network Server Edge

HP A5820

HP Virtual Connect

FlexFabric/Flex-10

Servers
Servers
• HP Virtual Connect FlexFabric/Flex-10 Servers The specific network capabilities for this configuration are

The specific network capabilities for this configuration are described below.

Virtualize configuration and management of multiple network switch devices to enable simplified network design and extend network scale/utilization using Intelligent Resilient Framework (IRF) technology on HP A-series network switches.

Virtualize and enable per application, priority-based bandwidth control and wire-once connectivity migration with Virtual Connect Flex-10 & Virtual Connect FlexFabric for HP BladeSystem.

Implement highly-scalable core and aggregation switching platforms.

Implement consolidated, single-pane-of-glass, multi-vendor, multi-site network resource management using HP Intelligent Management Center (IMC)/ HP Network Management Console for automation and orchestration.

Future Enhancements

HP FlexFabric utilizes the latest industry standards, including higher speed Ethernet links, Virtual Ethernet Port Aggregation (VEPA), Fibre Channel over Ethernet (FCoE), and Converged Enhanced Ethernet (CEE). The CEE standard enables Ethernet to deliver a “lossless” transport technology with congestion management and flow control features needed in storage environments. HP is championing many of these standards in the IEEE and other organizations in order to give its customers a data center fabric that protects their technology investments. The use of proprietary approaches can cause organizational disruption and wholesale equipment replacement or lock-in during infrastructure extensions or upgrades.

Provisioning and Management: Matrix Operating Environment

Provisioning and management for HP Converged Infrastructure is delivered through software components of the HP Matrix Operating Environment (MOE). This enables management and automation of HP Converged Infrastructure to accelerate infrastructure delivery in a consistently repeatable way. This approach ensures efficient use of IT resources and staff time, and mitigates risks. In one example of HP Converged Infrastructure deployment, MOE is delivered as an integral part of HP BladeSystem Matrix via Insight Dynamics software, a keystone of HP’s broad offerings for infrastructure management.

Challenge

Traditional data centers are built around an aging, siloed architecture, which limits the ability to build in efficiencies and enhance service levels. In a traditional data center deployment, it can take weeks or months to implement new infrastructure and bring new services online. This challenge can be addressed using a shared services approach.

Shared services best practices

Automate application deployment and provisioning with standardized templates.

Provide governance and auditing to ensure resource accountability.

Provide a variety of platform and availability choices to ensure cost-appropriate service levels.

Provide hardware for deployments from a reusable and redistributable resource pool to ensure rapid access and maximize utilization.

Update operational processes to ensure maximum efficiency for operational staff.

Provide consistent reporting to satisfy management oversight.

HP solution

HP Matrix Operating Environment is a shared-services infrastructure management solution. HP Matrix Operating Environment delivers a common management platform for provisioning and adapting infrastructure to instantly respond to business demands. The HP approach optimizes and automates the management of the resource pools and the IT roles associated with provisioning and consuming resources from these pools, while operating in compliance with core business, security, and regulatory policies. HP Insight Dynamics and complementary products provide the core functionality of the Matrix Operating Environment.

Functional elements are listed below.

HP Insight Dynamics enables administrators to manage standalone physical servers, virtual machine hosts and guests, and physical hardware partitions on Integrity servers, all from a single management console. Insight Dynamics also enables both physical blade servers and virtual machines to be managed as logical servers so they can be easily moved and migrated within a Matrix environment. It delivers a full range of deployment, management, capacity planning, migration, movement, and disaster recovery capabilities when paired with supported HP server and storage technologies.

HP Insight Dynamics includes an embedded workflow automation engine based on HP Operations Orchestration that provides role-based design and provisioning. This enables integration with customer IT processes and extensible automation based on customer workflows.

The Designer within Insight Dynamics enables drag and drop development of infrastructure templates for all of the components of an application infrastructure. These templates are saved to form a service catalog.

The self-service portal within Insight Dynamics provides authorized users with a catalog of available applications for deployment. Authorization process is driven by the embedded workflow engine.

Virtual Connect Enterprise Manager takes input from templates to provision the connections between servers, storage, and networking resources for an application.

In addition to delivering applications and services, IT is responsible for managing and allocating costs. A shared services model where IT resources are centralized and shared among business units requires a cultural shift. Agreements need to be put in place between IT and the lines of business for the services that the business needs to have provisioned. HP Converged Infrastructure provides usage and service level reporting tools to ensure IT can demonstrate value add to business customers as well as track resource usage for charge-back.

Support for Industry Standards

HP has a tightly coupled partnering strategy to deliver a Converged Infrastructure in a deployment scenario that works for your unique needs and built on industry standard hardware and interconnects. HP has more than 180,000 channel partners worldwide, including major and emerging software and hardware vendors as well as system integrators who leverage our products and services. HP works closely with these partners to deliver integrated solutions based on open standards. We deliver solutions that work with your existing infrastructure and provide investment protection for the future.

The HP focus on standards and open partnerships enables IT organizations to move to HP Converged Infrastructure at their own pace and without the fear of being locked in. It enables them to leverage existing investments while reaping the benefits of pooled resources and orchestrated deployments. HP’s Converged Infrastructure provides the advantages that come with vendor consolidation and integrated engineering while still enabling choice and access to new technologies.

Applying the HP Converged Infrastructure Maturity Model

The technology integration described earlier in this section opens the door to major improvements in data center planning, deployment and operational processes. Infrastructure convergence reflects today’s data center economics, where hardware costs are no longer the only pivot point for IT decisions. Other factors, such as provisioning, power and cooling, and operational flexibility now play an important part in platform decisions. To address those new requirements, HP has created the Converged Infrastructure Maturity Model (CI MM) to help organizations put in place the skills and processes required to reap the benefits of HP Converged Infrastructure. HP provides a roadmap for Converged Infrastructure and a path to make steps from your current state to an automated and adaptively sourced infrastructure.

The value of applying CI MM comes from building capability in key aspects of the data center competence, as shown in the following table. By adopting HP CI MM and integrating the recommended skills and tools, each area of competence can be supported by the required level of maturity.

can be supported by the required level of maturity. Table 3. Scope and Stages of HP

Table 3.

Scope and Stages of HP Converged Infrastructure Maturity Model

Areas of Competence

Stages of Maturity

Technology and Architecture Management and Processes Culture and IT Staff Demand Supply and IT Governance

Stage 5 – Adaptively sourced infrastructure Stage 4 – Automated, Service-oriented Stage 3 – Optimized Stage 2 – Standardized Stage 1 – Compartmentalized or Legacy

– Automated, Service-oriented Stage 3 – Optimized Stage 2 – Standardized Stage 1 – Compartmentalized or

Within this guide, the discussion is limited to how CI MM can assist with the rollout of applications services within the HP Converged Infrastructure Reference Architecture. Specific application workloads, based on the Converged Infrastructure Reference Architecture, are defined in the next section. For a broader view of how to apply CI MM to improve data center capability, refer to the section at the end of this document titled “For More Information”.

Overview of the Converged Infrastructure Maturity Model

CI MM provides criteria-based measures of organizational capability. These criteria are based on

working with hundreds of organizations across a wide range of industries. HP’s goal is to assist companies to improve capability and achieve specific business objectives. HP has found that as IT organizations develop maturity relative to the model, their capability to deploy shared services

infrastructure increases. This increased competency then generates significant value to the business.

As an example, the table illustrates how HP Converged Infrastructure aligns to the maturity model.

HP Converged Infrastructure aligns to the maturity model. Table 4. Scoping and mapping of stages of

Table 4.

Scoping and mapping of stages of maturity to data center capability and HP products

CI MM Stage

Infrastructure Model

Organizational Capability

Converged Infrastructure Offerings

Stage 5 – Adaptively sourced infrastructure

Federated cloud

Automated management and reallocation of infrastructure

HP cloud offerings delivered via Converged Infrastructure

Stage 4 – Automated, Service-oriented

Private cloud

Infrastructure offered as a service

HP BladeSystem Matrix integrated with HP StorageWorks scale-out Storage, HPN A-Series

Stage 3 – Optimized

Standards-based, fully

Rationalized technologies, architecture, management tools and processes

HP BladeSystem c-Class with Virtual Connect, HP StorageWorks virtualization, HPN A-Series

virtualized

infrastructure

Stage 2 – Standardized

Integrated components, virtualization

Standardized technologies, drive deployment decisions

HP BladeSystem c-Class, HP StorageWorks disk arrays, HPN A-Series

Stage 1 – Compartmentalized or Legacy

Silos of distributed servers

Project-based decisions, ad hoc management tools and processes

HP Integrity or HP ProLiant Servers, HP StorageWorks DAS or Internal Storage, HPN A-Series and E-Series

DAS or Internal Storage, HPN A-Series and E-Series Organizations that have already deployed infrastructure

Organizations that have already deployed infrastructure based on the HP BladeSystem c-Class

already have IT staff with expertise at the “Level 2 – Standardized” maturity level. They can be leveraged to develop a broader set of Level 2 capability and also to peak their skills towards Level 3

or beyond.

Skills competency and maturity levels

HP Converged Infrastructure Maturity Model matches tools and skills at each maturity level, enabling IT organizations to incrementally grow capability and see the results reflected in improved business results. HP Converged Infrastructure is built from standards-based products, and so enables step by step improvement in capability. IT organizations can easily build on their existing investments, rather than having to start from scratch.

The following table illustrates how HP management tools enable organizations to achieve higher levels of maturity and capability.

Table 5. CI MM stage and representative HP products CI MM Stage Level of Automation

Table 5.

CI MM stage and representative HP products

CI MM Stage

Level of Automation

Supporting Converged

Infrastructure Products

Stage 5 – Adaptively sourced infrastructure

Automated management and reallocation of converged infrastructure

HP Cloud Service Automation, HP BTO Software

Stage 4 – Automated, Service- oriented

Infrastructure and applications offered as a service through self-service portal

HP Insight Dynamics, HP BTO Software, HP Cloud Service Automation for Matrix

Stage 3 – Optimized

Deploy template-based services and applications

HP Insight Dynamics

Stage 2 – Standardized

Rapidly reallocate resources to meet business needs

HP Insight Control

HP Virtual Connect Enterprise Manager, HPNA/NMMi

Stage 1 – Compartmentalized or Legacy

Scripted processes and silo-specific capabilities

HP Systems Insight Manager, HP Intelligent Management Center

HP Systems Insight Manager, HP Intelligent Management Center Operational advantages of high maturity levels At higher

Operational advantages of high maturity levels

At higher levels of maturity, resource pools become an elastic extension of business requirements. The large common pool of compute, storage and bandwidth means that the resources needed to meet new requirements can be provisioned quickly. Standardization of building block resources, coupled with simplified authorization processes mean less time and money are spent readying infrastructure for changing business needs. At the same time, resource reporting and analysis can ensure resource utilization, so that usage and cost allocation are transparent to business users.

Evolutionary vs. transformational approach

Adoption of converged infrastructure combines technology advances with improvements in organizational skills and processes. Each complements the other. Within existing environments, HP Converged Infrastructure naturally enables evolutionary adoption of tools and incremental levels of automation and efficiency. For new “greenfield” IT investments, HP Converged Infrastructure enables IT staff to deploy resources that span business needs, rather than simply individual silos.

The next section defines recommended architectures, and some specific configuration examples, for a broad range of common enterprise workloads. Because these workloads share a common infrastructure, capacity planning can be much more efficient. In the past, IT grew as a set of standalone silos, each planned and operated separately. With HP Converged Infrastructure, that larger set of requirements can be addressed up front, and efficiencies can be driven across the entire IT infrastructure.

HP Converged Infrastructure Reference Architectures

Overview

HP Converged Infrastructure enables IT organizations to deploy a standards-based, common infrastructure to implement new solutions, or evolve existing ones. This common Converged Infrastructure “core” can be flexibly deployed to meet business objectives and create an agile IT solution environment.

The sections that follow provide specific examples of how HP Converged Infrastructure can be used as a core architectural approach to supporting these common enterprise workloads:

ERP/CRM, using SAP ERP 6.0.

Messaging and Collaboration, using Microsoft Exchange.

Cloud computing, with a focus on enterprise private cloud.

Virtualization, covering x86 and HP-UX virtualization technologies and three use cases.

Additional use cases are under development.

These configurations demonstrate the flexibility of the basic HP Converged Infrastructure architecture. They are a guide for planners and architects to illustrate practical application of the architecture to mainstream enterprise workloads. They also illustrate how specific tools, such as orchestration templates, can be employed to simplify deployment. Please consult your HP sales representative for support on planning any particular solution or deployment for your organization.

Generalized Reference Architecture

HP Converged Infrastructure relies on a common set of components, associated with a comprehensive set of service definition, deployment, and management tools. Infrastructure components can be deployed, and redeployed, to meet business needs. These common components can be utilized all, or in part, to create a platform on which to build IT solutions. These components provide a “core” for solution deployments and enhance the maturity and agility level of the entire IT environment, as described earlier.

The agile data center provisions from pools of resources for services requested by the users from portals, APIs, or other interfaces. These virtual/physical resource pools include servers, storage, and networking which are allocated from automation resources over the converged network. These pools of resources will be orchestrated by the operations orchestration to deliver the services requested via “service catalogs” based on business SLAs. Depending on the SLA, the required scalability, availability, and local or remote disaster recovery will be enabled. Monitoring of these resources will occur at the infrastructure, application, and end user side. Capacity will be typically available on- demand and elastic to address workload spikes, such as during peak business events.

From a functional perspective, Figure 6 depicts how these data center operations are addressed by the core elements that form the HP Converged Infrastructure architecture.

Figure 6. HP Converged Infrastructure Functional Block Diagram SERVER POOL NAS Storage SAN FC Storage

Figure 6.

HP Converged Infrastructure Functional Block Diagram

SERVER POOL NAS Storage SAN FC Storage iSCSI Storage STORAGE POOL NETWORK INFRASTRUCTURE Hardware Elements

SERVER POOL

NAS Storage SAN FC Storage iSCSI Storage STORAGE POOL
NAS Storage
SAN
FC Storage
iSCSI Storage
STORAGE POOL
NAS Storage SAN FC Storage iSCSI Storage STORAGE POOL NETWORK INFRASTRUCTURE Hardware Elements Shared Service

NETWORK

INFRASTRUCTURE

Hardware

Elements

Shared Service Environmental Application Infrastructure Catalog Monitoring Monitoring Monitoring Server
Shared Service
Environmental
Application
Infrastructure
Catalog
Monitoring
Monitoring
Monitoring
Server
Automation
Hypervisors
Management
Storage
Provisioning
Manager
WAN
S
Fibre Channel
A
S
N
A
S
Storage
N
A
N
SAN/NAS
Resource Pool
Datacenter C
Gateways
Storage
Automation
Virtual Switches
Network
Automation
S
A
S
N
S
A
S
A
N
S
A
N
A
N
S
N
A
N
Management Software
Datacenter B
Control
Datacenter A
Converged Networking
Operations Orchestration
Datacenter A Converged Networking Operations Orchestration To facilitate the creation of Converged Infrastructure

To facilitate the creation of Converged Infrastructure across typical deployment solutions, an intelligent software layer “Matrix Operating Environment” draws from resource pools of server/storage/network Infrastructure and pre-defined shared-service application catalogs to dynamically link the right resources when provisioning application instantiations.

Virtual Resource Pools

Servers: Pools of Servers (typically in Blade enclosures such as the HP BladeSystem with x86 blades or HP Integrity) are maintained with HP Virtual Connect providing an IO abstraction for connectivity to networking and storage.

Storage: Disparate heterogeneous storage resources based on HP or other 3rd party disk arrays may be combined into seamless scalable pools of storage through the use of storage virtualization technologies from HP. Flexible storage access mechanisms may be used to connect to the storage including traditional Fibre Channel, iSCSI, Fibre Channel over Ethernet, and Serial Attached SCSI (SAS).

Network: A traditional hierarchical data center network built on heterogeneous network switches and routers are used to connect the application instantiations within the infrastructure resource pools to the WAN backbone. Technologies such as HP Networking IRF provide the ability for network backbone resources and capacity to be dynamically scaled through the grouping of switches together.

FlexFabric

The HP BladeSystem Virtual Connect FlexFabric module can dynamically allocate I/O bandwidth for storage and network connections. This provides a virtualized, efficiently utilized, and highly scalable data center fabric. This open architecture uses industry standards to provide high performance, secure, converged connectivity to storage and server resources. FlexFabric combines intelligence at

the server edge with an advanced orchestration and management layer to enable virtualization-aware networking, predictable performance, and rapid, secure, business-driven provisioning. This approach addresses traditional network performance bottlenecks such as latency and network contention.

Matrix Operating Environment

The key software building block of the Matrix Operating Environment is HP Insight Dynamics. Insight Dynamics delivers three key capabilities:

Provisioning: HP Insight Dynamics infrastructure orchestration capabilities can provision infrastructure in minutes to automatically activate physical and virtual servers, storage, and networking from pools of shared resources. Whether for a single virtual machine or infrastructure for a complex three-tier application, Insight Dynamics finds available resources, streamlines the approval process, and automatically provisions and configures what’s needed across infrastructure pools. It also provides the ability to create a shared service catalog of application templates and a portal to instantiate individual applications based on these templates.

Optimization: Insight Dynamics captures key data points about actual system utilization like power draw, CPU and network utilization every five minutes. This lets you quickly adjust and optimize your environment over its lifecycle so you can predictably make changes without time- consuming analysis. When combined with built-in re-balancing tools, this can eliminate weeks or months of tedious planning and implementation.

Protection: Insight Dynamics protects quality of service and offers continuity of services with a wide spectrum of high availability and recovery solutions, ranging from server-aware and application-aware availability solutions, to disaster recovery solutions for distances from campus to continental for both physical and virtual server environments.

Insight Dynamics builds on the essential server management delivered by HP Insight Control that unlocks the management capabilities built into HP ProLiant servers. Insight Control enables you to proactively manage server health - whether physical or virtual, deploy servers quickly, optimize power consumption, and control servers from anywhere. It also leverages HP Virtual Connect Enterprise Manager (VCEM), which centralizes connection management and workload mobility for HP BladeSystem servers that use Virtual Connect to access LANs, SANs and converged network infrastructures.

For Integrity servers, Insight Dynamics – VSE also includes advanced workload management that can automatically grow and shrink virtual servers as business needs change.

The Matrix Operating Environment is integrated with HP Business Technology Optimization (BTO) software, including HP Server Automation software for provisioning applications. By using the Matrix Operating Environment together with HP Server Automation, both infrastructure and applications can be provisioned in minutes.

A full HP Converged Infrastructure deployment can take advantage of all of these functional areas,

whether for a single application, or across an enterprise. A more limited use of the various elements

of the Converged Infrastructure functional core can still enhance IT capabilities for new or existing

applications.

From the functional description above, a deployment approach is derived and shown in Figure 7 below. This deployment architecture maps the Converged Infrastructure functional components to how they can be leveraged to build actual solution architectures. This “core” architecture can serve as a foundation for custom application deployments, whether new specific workloads such as those described further below, brand new applications, or modifications of existing ones.

This core physical deployment architecture forms the foundation for building the solution blocks that follow.

This core physical deployment architecture forms the foundation for building the solution blocks that follow.

Figure 7.

HP Converged Infrastructure Reference Configuration

Transport Transport (MPLS) (Internet) WAN Cloud – CI Cloud – CI CI Management Management CI
Transport
Transport
(MPLS)
(Internet)
WAN
Cloud – CI
Cloud – CI
CI Management
Management
CI Mgmt
Management
Software Severs
Software Severs
Software Servers
Internal
L2/L3
Internal
Internal
Core
Redundant
Internal
Core
Core
Core
Core
Network Services
Network Services
1/10
GigE / FCOE
Services
Management
FC
HP BladeSystem
• HP Cloud
1/10GE/FCoE
Service
Top of Rack
HPN A58x0
Automation
Virtual Connect
Flex 10
• HP MOE
Rack
Servers
SAN
iSCSI – FCoE Storage
NAS Storage
FC Storage
SAN iSCSI – FCoE Storage NAS Storage FC Storage The above core HP Converged Infrastructure core

The above core HP Converged Infrastructure core reference architecture “RA” provides the next level design. Specific reference configurations, described later, can be deployed as single or replicated deployments, based on an organization’s capabilities relative to the Converged Infrastructure Maturity Model. Maximum benefits accrue when deployed as service-oriented architecture, leveraging management automation via the HP BladeSystem Matrix Operating Environment (MOE) or HP Software to optimize TCO and CAPEX/OPEX.

Secondly, HP BladeSystem can be used to collapse the network layer in the Converged Infrastructure, reducing or eliminating the need to use extra hub routing to utilize the top of the rack switches, flattening the layer 2 levels. This is key to addressing virtualization: fewer hops, fewer cables or switch “NIC-SAN” ports as well as performance, power, and cooling.

Lastly, the core Converged Infrastructure reference architecture can connect to an existing heterogeneous environment and in-house or 3rd party automation software to take advantage of the virtual/physical pools of resources. These can include rack mount servers, hypervisors, Fibre Channel, iSCSI, FCoE, NAS Storage, and network devices.

This architecture is interchangeable between HP Converged Infrastructure components and heterogeneous components. The HP open, non-proprietary approach reduces costs and improves flexibility for customers as they upgrade their environments.

Elements of the configurations are defined in the table, below.

Elements of the configurations are defined in the table, below.

Table 6.

Elements of the HP Converged Infrastructure Reference Architecture

Symbol

 

Element

Description

Discrete servers HP ProLiant BL server blades, DL rack servers, ML tower servers, and HP

Discrete servers

HP ProLiant BL server blades, DL rack servers, ML tower servers, and HP ProLiant Scalable, HP Integrity, HP Integrity Superdome 2

Network services Firewall, Intrusion Prevention Services, Load Balancing

Network services

Firewall, Intrusion Prevention Services, Load Balancing

Network core Backbone (Enterprise WAN edge), Interconnect (Core/Distribution), Server Edge (BladeSystem- integrated, Top

Network core

Backbone (Enterprise WAN edge), Interconnect (Core/Distribution), Server Edge (BladeSystem- integrated, Top of Rack) - A12508, A9505, A5820, A8812, A6604, A6616, Virtual Connect

 
  Fibre Channel SAN Storage HP StorageWorks EVA, SVSP, XP. HP StorageWorks Storage Networking

Fibre Channel SAN Storage

HP StorageWorks EVA, SVSP, XP. HP StorageWorks Storage Networking

Network Attached HP StorageWorks X9000 Network Storage Systems
Network Attached HP StorageWorks X9000 Network Storage Systems
Network Attached HP StorageWorks X9000 Network Storage Systems

Network Attached

HP StorageWorks X9000 Network Storage Systems

 

(NAS) Storage

iSCSI SAN Storage HP StorageWorks P4000 G2 SAN, HP Networking A-Series FCoE capable devices

iSCSI SAN Storage

HP StorageWorks P4000 G2 SAN, HP Networking A-Series FCoE capable devices

BladeSystem HP Integrity server blades, HP ProLiant server blades, BladeSystem c-Class enclosures, HP Integrity NonStop

BladeSystem

HP Integrity server blades, HP ProLiant server blades, BladeSystem c-Class enclosures, HP Integrity NonStop BladeSystem

Top of Rack A5820 switches, and for BladeSystem environments, HP Virtual Connect Flex

Top of Rack

A5820 switches, and for BladeSystem environments, HP Virtual Connect Flex

10/FlexFabric

CI Management HP Insight Dynamics, HP Insight Control, Virtual Connect Enterprise Manager, HP Cloud Automated

CI Management

HP Insight Dynamics, HP Insight Control, Virtual Connect Enterprise Manager, HP Cloud Automated Software IaaS, PaaS, SaaS Small, Medium, Large Foot print

Symbol

Element

Description

WAN Wide Area Networking, including Multiprotocol Label Switching

WAN

Wide Area Networking, including Multiprotocol Label Switching

Shared service HP Infrastructure Operating Environment, HP Server Automation, HP IT Shared Service (ITSS) portfolio
Shared service HP Infrastructure Operating Environment, HP Server Automation, HP IT Shared Service (ITSS) portfolio

Shared service

HP Infrastructure Operating Environment, HP Server Automation, HP IT Shared Service (ITSS) portfolio

catalog

Environmental HP Environmental Edge

Environmental

HP Environmental Edge

Monitoring

Application HP Business Availability Center - HP Discovery and Dependency Mapping software for Infrastructure

Application

HP Business Availability Center - HP Discovery and Dependency Mapping software for Infrastructure

Monitoring

Infrastructure HP Business Service Management (BSM), HP Operations Manager software

Infrastructure

HP Business Service Management (BSM), HP Operations Manager software

Monitoring

Data center Thermal Logic, Data Center Environmental Edge, Performance Optimized Data center (POD)

Data center

Thermal Logic, Data Center Environmental Edge, Performance Optimized Data center (POD)

Environmental Edge, Performance Optimized Data center (POD) These elements provide a common approach to implementing

These elements provide a common approach to implementing solutions based on HP Converged Infrastructure. Sizing and other IT objectives must be considered for context, and the particular solution architecture adapted to meet IT needs and desired maturity level.

The specific application architectures and physical configurations defined in the following sections are all derived from the functional and component-level diagrams shown here. By providing a single infrastructure to support multiple workload types, HP Converged Infrastructure simplifies delivery of a shared service IT environment.

ERP/CRM: SAP Business Suite 7, ERP 6.0 Use Case

Solution Architectural Principles

Basic solution description and business need

SAP Business Suite 7™ is the newest application suite from SAP to help companies to optimize performance and reduce IT cost by synchronizing the release schedule of major SAP components (SAP ERP powered by NetWeaver™, CRM, PLM, SCM, and SRM, plus industry solutions and enhancement packages). The main benefits of the SAP Business Suite are to drive wider adoption of Service Oriented Architecture (SOA) design concepts, support for collaborative business processes, reduce the barriers for implementing new capabilities and functionalities, and provide a framework to deliver value more quickly. One of the popular components of SAP Business Suite is SAP ERP 6.0, which is shown in the figure below. The main benefits of this approach are the simplified workflow and seamless integration of the various business processes within an enterprise.

of the various business processes within an enterprise. Figure 8. SAP Enterprise Resource Planning 6.0 –

Figure 8.

SAP Enterprise Resource Planning 6.0 – functional description

Self Services Procurement Internet Sales Self Services Strategic Enterprise Management Supplier Relationship
Self Services Procurement
Internet Sales
Self Services
Strategic Enterprise Management
Supplier Relationship Management
And More…
Composite Applications
SAP ECC
SAP ECC Core
Switch Framework
Enterprise Extensions
Industry Extensions
SAP NetWeaver 7

SAP Environment Supported by HP Converged Infrastructure

Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
Enterprise Extensions Industry Extensions SAP NetWeaver 7 SAP Environment Supported by HP Converged Infrastructure
SAP Environment Supported by HP Converged Infrastructure The functionality within SAP ERP is split into modules

The functionality within SAP ERP is split into modules dedicated to the business functions in an enterprise:

Self-services

Analytics

Financials

Human capital management

Operations

Corporate services

SAP ERP is powered by SAP NetWeaver an innovative integration and application platform. SAP NetWeaver lays the foundation for new cross-functional business processes.

Basic SAP architectural principle and considerations

A typical SAP architecture consists of an application stack, central services, processes and a database instance. In the simplest form, all reside within a single operating environment, as shown below. In this case, if the application and database tier are run on a single Operating System, this is called an SAP two-tier architecture, or central-system architecture. The HP Converged Infrastructure model can bring benefits to such SAP central system architectures by easing deployments and offering additional resiliency features through server, storage, and network virtualization.

through server, storage, and network virtualization. Figure 9. Core SAP architecture description Operating

Figure 9.

Core SAP architecture description

Operating Environment

Application

SAP Processes

Dialog

Update

Batch

Spool

Gateway

SAP Central Services

Message

Enqueue

Database

Services • Message • Enqueue Database Sizing and scaling principles The recommendation of the

Sizing and scaling principles

The recommendation of the reference environment includes assumptions made by HP.

Assumptions

The SAP sizing methodology for HP platforms is based on performance data gathered from existing customers, tests performed in HP labs, customer tests in HP benchmark centers, and published industry benchmarks. Input is also taken from SAP Consulting and data utilized from sizing studies of actual customers in production.

The SAP sizing methodology assumes that the servers, mass storage units, databases, and reporting queries are well-tuned for the applications. A tuning effort is generally recommended before and after the system goes into production.

Verifying system performance with stress tests or operational performance tests is highly recommended in the months prior to going live with SAP—and after the system goes into production. Operational monitoring of infrastructure, as well as scheduled application reports such as SAP EarlyWatch, can provide feedback on consumption patterns and overall scaling of the SAP system.

While typical SAP system monitoring focuses primarily on compute resources, a converged infrastructure allows for more holistic application assessments and is inclusive of network and storage I/O.

a converged infrastructure allows for more holistic application assessments and is inclusive of network and storage

Figure 10. Resource Consumption and Scaling Capabilities for SAP OLTP Activity

Operating Environment

Application

SAP Processes

70 %
70
%

Dialog

Update

Batch

Spool

Gateway

SAP Central

10 %
10
%

Services

Message

Enqueue

Database

20 %
20
%
up out up up Out*
up
out
up
up
Out*
Database 20 % up out up up Out* Overall resource consumption in SAP OLTP environments is

Overall resource consumption in SAP OLTP environments is concentrated within the application tier. The diagram above illustrates load distribution between application tier, central services and database tier. From a system-scaling perspective, 70% of the resources are usually consumed by the application tier and about 20-30% by the database tier.

This understanding of scaling characteristics within the core SAP architecture is a key pre-requisite, when applying converged infrastructure principles. HP Converged Infrastructure provides a pool of compute resources and can adapt dynamically to the changes in workload by redeploying existing or additional compute resources in the SAP infrastructure.

While growth of an SAP two-tier system, or central system architecture, is accommodated through vertical scaling of services, an SAP three-tier system, or distributed system architecture, can allow for horizontal scaling. In the diagram above, both SAP application tier and database tier can scale up or out (with Oracle RAC), while SAP Central Services are limited to scaling up.

A converged infrastructure provides the broader application/service view while managing the differing scaling attributes of an SAP system. Since HP Converged Infrastructure is not limited to specific products, it is able to allocate the right-sized resources to specific tiers within SAP.

High Availability Principles and Capabilities

An SAP system consists of components that are either single or multi-instance capable. SAP work processes such as Dialog, Batch, Update, etc. can be installed redundantly across multiple operating environments. The database is typically single instance, although distributed databases such as Oracle RAC are supported.

SAP central services, which include the Enqueue and Message processes, are Single Points of Failure (SPOF) – if these services fail, the SAP application comes to a halt. SAP application servers are not

SPOF since availability is provided through distributed redundancy: SAP application servers can be deployed on many servers, if one fails there are potentially other instances to cover SAP user requests.

The database instance is also a SPOF, since it is a central component where SAP applications store, access, and modify data. These SPOF components are provided high availability through clustering or similar relocation of instance services. Whether provided through redundancy or relocation of processes, a converged infrastructure increases resiliency through SAP instance orchestration and by supporting native or virtualized HA solutions without compromising the SLA to the end-user.

An example of a typical highly available SAP system is shown below.

of a typical highly available SAP system is shown below. Figure 11. High availability SAP system

Figure 11. High availability SAP system

Operating Environment Operating Environment Operating Environment Application Application Application SAP Processes
Operating Environment
Operating Environment
Operating Environment
Application
Application
Application
SAP Processes
SAP Processes
SAP Processes
• Dialog
• Dialog
• Update
• Update
• Batch
• Batch
Operating Environment
Operating Environment
Application
Application
SAP Central
Failover Target
Services
• Message
Database
Failover Target
Target Services • Message Database Failover Target Solution Block Architectures Architectural Principles The HP

Solution Block Architectures

Architectural Principles

The HP tiered solutions program was designed to provide broad applicability across a wide range of customers and hardware infrastructure platforms (server, storage and network). However, this service classification framework can be applied within a single organization to meet a variety of departmental and geographical requirements. The service tiers definitions are based on Recovery Time Objectives (RTO) and Recovery Point Objectives (RPO), and have been drawn upon HP’s experience working with several small, medium and large customers including its own internal IT implementation.

The Shared Database Utility (SDBU) is a framework for provisioning dedicated and isolated database environments on a hosted platform. Shared Application Services Utility (SASU) is a framework for hosting multiple application resources from many different business units on a shared pool of server resources. These concepts originated at HP’s internal IT transformation projects and have been widely used at several customer sites. This reference architecture, referred to as Shared ERP Service Utility (SESU), has been built based on the concepts of Shared Database Utility (SDBU) and Shared Application Services Utility (SASU) applied against the Converged Infrastructure principles.

With the HP BladeSystem Matrix, you can take this service level concept a step further and actually create and publish service templates that clearly define the various SAP services that can be deployed in your organization.

Specific configurations

This document will use two different standards of defining services based on the recovery of failures as defined below:

“Incident Recovery” is the small-scale failure that does not activate the migration of the primary site production activity to the Secondary site. Incidents are considered to have a higher probability of occurrence, and the ERP service line offers higher levels of service to offset the higher probability of failure. This standard is summarized in the following table:

failure. This standard is summarized in the following table: Table 7. Service Definition Criteria for an

Table 7.

Service Definition Criteria for an SAP System

INCIDENT RECOVERY CLASS

RTO

RPO

BASIC

RTO < 24 Hours

RPO = 24 Hours

ENHANCED

RTO = 8 Hours

RPO < 4 Hours

PREMIUM

RTO = 2 Hours

RPO <1 Hour

SPECIALTY

Custom

Custom

2 Hours RPO <1 Hour SPECIALTY Custom Custom “Disaster Recovery” is the large scale failure that

“Disaster Recovery” is the large scale failure that requires the migration of the Primary site production activity to the Secondary site.

The business values can be approximated as follows:

DR Class 1: Critical to the business needs, affecting multiple regions and streams.

DR Class 2: Critical to the business needs, affecting single region or stream.

DR Class 3: Does not adversely disrupt business needs, impacts only a few users.

The choice of Service Level should align with and support the DR profile of the application being hosted on the system.

Adapting to Your IT Operations

Operating Principles

This SESU Reference Architecture, where the ERP services are tiered based on the guidelines in the next table, is based on the following Operating Principles:

All processing capacity is sold by the IT organization to the business units by the SAPS or TPM or an equivalent metric.

Standards are enforced to allow the efficient relocation of ERP applications.

Instrumentation captures performance metrics for each application.

Hardware operates at a prescribed capacity limit, based on the tier of service.

As prescribed tier limits are reached, applications are relocated to servers with available capacity.

Matching tier applications are migrated into available capacity based on prescribed application limits.

Mismatched tiers of service cannot co-exist on the same host.

Shared ERP Hosting

Shared ERP hosting allows multiple applications to coexist on a common server platform, which generally results in more effective utilization of Converged Infrastructure. The ERP Service Line offers shared service on all three standard Service Classes. This allows the Service Catalogue offering to provide a higher level of service at a lower average cost relative to standalone server solutions.

Software Versions and Hosting platforms

The consolidation of versions, with the support commitment from the vendors, provides a decreased diversity in the support infrastructure and allows a higher availability at a lower overall cost. The ERP Service Line must be in conversation with SAP, HP, Oracle, and Microsoft to discover methodologies of standardizing services for existing and future applications. Other services may be provisioned if there is sufficient demand. The platform should provide a stable operating platform for future development and deployment of systems. By reducing the variation and standardizing on a single platform, the procurement and support costs are much lower than in a distributed system.

SAN Infrastructure – Required & Optional Services

All ERP service offerings may use standard storage as offered in the service catalogue. Only Specialty service has the option of not using standard storage. The following table provides a summary of the standard offering in the Service Catalog.

a summary of the standard offering in the Service Catalog. Table 8. High availability SAP system

Table 8.

High availability SAP system

Tier 1 – Replicated RAID-1 Tier 2 – Replicated RAID-5

Tier 3 – Local Non-Replicated RAID-1

Tier 4 – Local Non-Replicated RAID 5

RAID-1 Tier 4 – Local Non-Replicated RAID 5 Replication is provided within the service at the

Replication is provided within the service at the quoted prices for storage. Replication may be Synchronous or Asynchronous based on the offering from the approved Data center. Replicated service to the DR site supports only the following set of combinations:

Production Tier 1 to DR Tier 1

Production Tier 2 to DR Tier 2

No other combinations are supported

Backup & Recovery Service

Backup & Recovery Service has three major component offerings to the client:

IP tape-based backup and recovery. This service is required for all classes of service and is billed to the customer based on the number of gigabytes stored. It is the slowest and least expensive.

Snapshot or “Business Continuity Volume” (BCV) recovery. This is the fastest and most expensive service offering and is available only on Tier 1 & 3 storage.

Nearline Storage. This service splits the gap between Tape and BCV for both cost and performance. Nearline or BCV storage may be required for certain classes of service based on the volume of data and the RTO of that service class. The decision is technical and not financial.

Specific HP Products Applied to the General Architecture

Small ERP

Typically these are in the range of 1,000 medium-weighted SD users. Modular compute is based on the ProLiant BL460c G6. In SAP central server form, this platform achieves SD benchmark performance in excess of 25,000 SAPS. In native or virtualized OE, this is sufficient capacity for a small ERP implementation.

this is sufficient capacity for a small ERP implementation. Figure 12. Small SAP Configuration Small SAP

Figure 12. Small SAP Configuration

Small SAP ERP – 10,000 SAPS

Linux/Windows: ProLiant BL460c G6 12 cores / 96GB StorageWorks P4000 G2 SAN 3-4 TB usable
Linux/Windows:
ProLiant BL460c G6
12 cores / 96GB
StorageWorks P4000 G2 SAN
3-4 TB usable Capacity
/ 96GB StorageWorks P4000 G2 SAN 3-4 TB usable Capacity Availability on native operating environment installations

Availability on native operating environment installations can be supported through existing clustering technologies on Linux or Windows. Virtualized OE installations provide availability features through HA feature sets (in both VMware and Hyper-V products), and require a minimum of two blades for production.

Two additional blades provide compute capabilities for quality assurance and development, and testing can be supported in either native or virtual operating environments.

Storage resources are served by a StorageWorks P4000 G2 SAN Solution using iSCSI connectivity. Capacity requirements can vary, however small productive ERP systems are typically well served with 2TB of storage requirement. Suggesting 6-8 TB of usable capacity is sufficient space for growth, and to account for two non-productive copies. iSCSI is also a suitable SAN protocol for implementation.

HP Converged Infrastructure “touch-points”

This system represents an example of HP’s Converged Infrastructure, where diverse operating systems, servers, storage, and networking, can be easily managed in an economical bladed environment at reduced levels of power and cooling, providing customers with an ideal platform for their mission- critical ERP applications.

This SAP configuration illustrates many of the features of an HP Converged Infrastructure. The network element of the Converged Infrastructure (called “FlexFabric”) is embodied in the use of Virtual

Connect. Moreover, the Virtual Connect Manager manages virtual I/O connections to physical blades, allowing for dynamic assignment of I/O resources to SAP instances.

The ProLiant BL460c G6 provides a modular approach in terms of compute resources. It can be provisioned to support both native and virtualized operating environments through Blade matrix management. Finally, the StorageWorks P4000 storage solution provides storage LUN virtualization along with the network converged iSCSI SAN protocol.

Medium ERP

This category is in the range of 5,000 medium-weighted SD users. Modular compute can be ProLiant BL460c G6 or Integrity BL870c i2. In cases of Linux, Windows, or HP-UX operating environments, the SAP system architecture is three-tier and OE homogeneous. With a minimum of six productive blades, a medium ERP requirement can be met with a scalable and highly available solution.

can be met with a scalable and highly available solution. Figure 13. Medium size SAP system

Figure 13. Medium size SAP system

Medium SAP ERP – 50,000 SAPS

Medium SAP ERP – 50,000 SAPS HP-UX: Integrity BL870c i2 16 cores / 96GB Linux/Windows: ProLiant
Medium SAP ERP – 50,000 SAPS HP-UX: Integrity BL870c i2 16 cores / 96GB Linux/Windows: ProLiant
Medium SAP ERP – 50,000 SAPS HP-UX: Integrity BL870c i2 16 cores / 96GB Linux/Windows: ProLiant
Medium SAP ERP – 50,000 SAPS HP-UX: Integrity BL870c i2 16 cores / 96GB Linux/Windows: ProLiant
HP-UX: Integrity BL870c i2 16 cores / 96GB
HP-UX:
Integrity BL870c i2
16 cores / 96GB
Linux/Windows: ProLiant BL460c G6 12 cores / 96GB
Linux/Windows:
ProLiant BL460c G6
12 cores / 96GB
i2 16 cores / 96GB Linux/Windows: ProLiant BL460c G6 12 cores / 96GB StorageWorks 4400 EVA
StorageWorks 4400 EVA 8-10 TB useable capacity
StorageWorks 4400
EVA
8-10 TB useable
capacity
/ 96GB StorageWorks 4400 EVA 8-10 TB useable capacity The productive SAP system distribution across a

The productive SAP system distribution across a minimum of six blades is as follows:

SAP database and central services run together on one blade with a failover target representing a second blade. Blades three through six represent an n+1 application tier where all remaining SAP application processes run.

Both platforms offer 40 to 50 percent additional scale-up capacity for the SAP database and central services. The n+1 distribution of the application tier provides 50 to 60 percent additional scale-out capacity within the minimum of four blades.

Both platforms support high-availability features in both native and virtualized operating environments. Availability on native operating environment installations can be supported through existing clustering technologies on Linux, Windows, and HP-UX. Virtualized OE installations are provided availability features through HA feature sets in VMware, Hyper-V, and Integrity VM

products. Virtualized servers within the application tier are recommended and offer an additional benefit of flexible movement of SAP application instances across physical hosts.

The number of non-productive SAP landscapes supporting a medium ERP system can vary. A minimum of two additional blades provide sufficient compute for quality assurance and development with operational testing in either native or virtual operating environments.

Storage resources are served by a StorageWorks 4400 EVA solution. Capacity requirements can vary, however medium productive ERP systems are typically well served with between 4 and 5 TB of storage. Suggesting 12 to 15 TB of usable capacity is sufficient space for growth and account for at least two non-productive copies. Fibre Channel and upcoming FCoE are suitable SAN protocols for this size of implementation.

HP Converged Infrastructure “touch-points”

This system represents an example of HP’s Converged Infrastructure, where diverse operating systems, servers, storage, and networking can be easily managed in an economical bladed environment at reduced levels of power and cooling, providing customers with an ideal platform for their mission- critical ERP applications.

This SAP configuration illustrates many of the features of an HP Converged Infrastructure. The blade enclosure architecture accommodates Integrity (Intel Itanium 2 processor) servers or ProLiant (Intel Xeon processor) servers.

The network element of the Converged Infrastructure (called “FlexFabric”) is embodied in the use of Virtual Connect, and Blade Link on Integrity platforms. Moreover, the Virtual Connect Manager manages virtual I/O connections to physical blades. This allows for dynamic assignment of I/O resources to SAP instances. In the case of the medium ERP Converged Infrastructure reference, these I/O resources include Fibre Channel SAN protocol and future FCoE, in addition to Ethernet.

Both the Integrity BL870c i2 and the ProLiant BL460c G6 provide a modular approach in terms of compute resources. They can be provisioned to support both native and virtualized operating environments through Blade matrix management. Lastly, the EVA4400 provides a virtual resource pool of highly efficient storage capacity.

Large ERP

This is typically in the range of 10,000 medium-weighted SD users. Modular compute for the SAP application is based on either the ProLiant BL460c G6 or the Integrity BL870c i2. For the database— the ProLiant DL980 G7 and the Integrity Superdome 2—are referenced and indicate the utmost scale- up capacity. In cases of Linux, Windows, or HP-UX operating environments, the SAP system architecture is three-tier and can be OE homogeneous or heterogeneous. With a minimum of ten to twelve productive servers, a large ERP requirement can be met with a scalable and highly available solution.

Figure 14. High availability SAP system Large SAP ERP – 100,000+ SAPS Database with HA

Figure 14. High availability SAP system

Large SAP ERP – 100,000+ SAPS

Database with HA Clustering or Cache coherency HP-UX: Integrity BL870c I2 - 16 Cores /
Database with HA
Clustering or Cache
coherency
HP-UX:
Integrity BL870c I2 - 16
Cores / 96GB
Linux/Windows:
ProLiant BL460c G6 - 12
Cores / 96 GB
HP-UX:
StorageWorks EVA & XP
Linux/ Windows:
Integrity Superdome 2
32 Cores / 512 GB
ProLiant DL980 G7
32 Cores / 512 GB
N+1 App tier with Central
services MA Solution
/ 512 GB N+1 App tier with Central services MA Solution The SAP database is hosted

The SAP database is hosted apart from the SAP central services and remaining application processes. It runs within a native operating environment and gains availability from clustering or cache coherent database solutions.

The SAP central services and remaining application processes can run within native or virtualized operating environments. High availability features are supported in both cases for protecting the SAP central services. Availability on native operating environment installations can be supported through existing clustering technologies on Linux, Windows, and HP-UX. Virtualized OE installations are provided availability features through HA feature sets in VMware, Hyper-V, or Integrity VM products.

Virtualized servers within the application tier are recommended, and offer an additional benefit of flexible movement of SAP application instances across physical hosts.

The number of non-productive SAP landscapes supporting a large ERP system can vary. A minimum of four additional blades provides sufficient compute capability for quality assurance and development with operational testing in either native or virtual operating environments.

Storage resources are served by StorageWorks EVA and XP solutions through a tiered data approach. Capacity requirements for large ERP vary far too greatly to form a reference value. However, StorageWorks EVA and XP solutions offer scalability and support tiered architecture of business data. Fibre Channel and upcoming FCoE are suitable SAN protocols for this size of implementation.

HP Converged Infrastructure “touch-points”

This system represents an example of HP’s Converged Infrastructure, where diverse operating systems, servers, storage, and networking, can be easily managed in an economical bladed environment at reduced levels of power and cooling, providing customers with an ideal platform for their mission- critical ERP applications.

This SAP configuration illustrates many of the features of an HP Converged Infrastructure. The blade enclosure architecture accommodates Integrity (IntelItanium 2 processor) servers or ProLiant (IntelXeonprocessor) servers.

The network element of the Converged Infrastructure (called “FlexFabric”) is embodied in the use of Virtual Connect, as well as Blade Link on Integrity platforms. Moreover, the Virtual Connect Manager manages virtual I/O connections to physical blades. This allows for dynamic assignment of I/O resources to SAP instances. In the case of the large ERP Converged Infrastructure reference, these I/O resources include Fibre Channel SAN protocol and future FCoE, in addition to Ethernet.

Both the Integrity BL870c i2 and the ProLiant BL460c G6 provide a modular approach in terms of compute resources. They can be provisioned to support both native and virtualized operating environments through BladeSystem management. Lastly, the EVA and XP disk arrays demonstrate storage virtualization at the array and SAN level. Together, they provide a virtual resource pool of storage capacity.

Messaging & Collaboration: Microsoft Exchange Use Case

Solution Architectural Principles

Basic solution description and business need

As the importance of email messaging has grown, Microsoft® Exchange has become a business, or even mission critical application, with strict service level agreements (SLAs) for Exchange service uptime. Microsoft® Exchange Server 2010 (Exchange 2010) introduces many new design features and configuration options to consider.

This section will focus on the themes of flexibility and reliability, and the design changes of Exchange 2010 to address them. Flexibility and reliability speak to the ability of organizations to deploy Exchange in a variety of ways and on a variety of server and storage platforms.

By implementing HP Converged Infrastructure, companies can change the way Exchange 2010 is deployed in the data center and move towards a service driven, standardized deployment model. Exchange 2010 servers can be quickly provisioned and deployed as business demand requires, using pre-built templates that define the resource requirements based on the roles deployed.

Basic Exchange architectural principles and considerations

The core architecture of Exchange has undergone many changes. In Exchange 2010, new features such as incremental deployment, mailbox database copies, and database availability groups (DAGs) have been added. These new features work in conjunction with other features such as shadow redundancy and transport dumpster, to provide a newly designed platform for high availability and site resilience.

Many organizations group their management tasks around a core set of server roles. Exchange 2010 maps Exchange Server management to this same approach. Server roles logically group the required features and components needed to perform specific functions. This enables administrators to easily choose which features are installed on an Exchange server. The requirement of a server role is that it can be run as an atomic unit of scalability.

A server role is composed of a group of features. Logically grouping features in server roles offers the following advantages:

Reduces attack surface on an Exchange server.

Allows you to install and configure an Exchange server the way you intend to use it.

Offers the ability to fully customize a server to support your business goals and needs.

Exchange 2010 includes the following server roles:

Mailbox Server: Hosts mailboxes and public folders.

Client Access Server: Hosts client access such as Outlook Anywhere (MAPI over HTTP), Outlook Web Access (OWA), ActiveSync service, Post Office Protocol version 3 (POP3), Internet Message Access Protocol version 4 (IMAP4), the Availability service, and the Autodiscover service.

Hub Transport Server: Routes mail within the Exchange organization.

Edge Transport Server: Sits at the perimeter of the topology and routes mail in to and out of the Exchange organization, provides protection against viruses and spam.

Unified Messaging Server: Connects a Private Branch eXchange (PBX) or VOIP telephone system to Exchange 2010.

The next figure illustrates an enterprise topology with each server role deployed.

The next figure illustrates an enterprise topology with each server role deployed.

Figure 15. Exchange 2010 enterprise topology

Enterprise Network

Phone system

(PBX or VOIP)

Enterprise Network Phone system (PBX or VOIP) Unified Messaging Voice mail & voice access External SMTP

Unified Messaging Voice mail & voice access

External SMTP servers
External SMTP
servers

Mobile phone

Web

browser

Outlook

Hub Transport Routing & Policy
Hub Transport
Routing & Policy
Mailbox Storage of mailbox items Client Access Client connectivity Web services
Mailbox
Storage of
mailbox items
Client Access
Client connectivity
Web services

Edge

Transport

Routing &

AV/AS

Web services Edge Transport Routing & AV/AS Outlook (local user) (remote user) Line of business
Web services Edge Transport Routing & AV/AS Outlook (local user) (remote user) Line of business

Outlook (local user)

services Edge Transport Routing & AV/AS Outlook (local user) (remote user) Line of business application
services Edge Transport Routing & AV/AS Outlook (local user) (remote user) Line of business application

(remote user)

Line of business application

(local user) (remote user) Line of business application High availability designs for Exchange have evolved over

High availability designs for Exchange have evolved over the years from a model based on Microsoft Windows® clustering (single copy clustering) with Exchange 2003 to an application replication model with Exchange 2007, including cluster continuous replication (CCR), local continuous replication (LCR), and standby continuous replication (SCR). Exchange 2010 builds on the model of application replication from Exchange 2007 with the introduction of Data Availability Group (DAG). Exchange 2010 integrates high availability into the core architecture of Exchange. Now customers of all sizes and segments can economically deploy a highly available messaging service in their organization.

The DAG model supports both failovers (unplanned) and switchovers (administrator initiated) at the individual database level, unlike Cluster Continuous Replication CCR in Exchange Server 2007, which required all databases configured on an entire server to be failed over. The DAG concept uses a multi-server database replication process which uses the new active manager component. The active manager monitors all of the active databases within a DAG, and in the event of a database failure, determines which of the database copies to make the active copy based on each database’s health and activation preference. Exchange 2010 supports combining the Client Access Server (CAS), Hub Transport Server (HT) and Mailbox Server roles on a single combined role server.

Sizing and scaling principles

With the improvements in deployment flexibility and the introduction of the DAG model for database resiliency, Exchange 2010 deployments can be more formalized around a building-block architecture that is tied into the concept of service tiers. HP has developed the tiered-solutions matrix for Exchange to provide guidance on building Exchange 2010 solutions to meet various Exchange service levels and business requirements.

Solutions range from basic mailbox services on HP’s entry-level server and storage platforms, to highly available configurations, built with HP’s enterprise server and storage platforms. The HP tiered solutions matrix matches server and storage hardware components and configurations along with the appropriate Exchange 2010 availability features for each tier. There are four tiers in the matrix:

Bronze, Silver, Gold, and Platinum. These tiers provide guidance in matching mailbox service levels with appropriate HP server and storage configurations.

Bronze Tier: The Bronze Tier is designed for deployments where cost is a driving factor. It provides basic services without the additional cost of high availability and is targeted towards smaller, cost- sensitive deployments. The Bronze Tier uses an all-in-one approach by combining each of the primary Exchange roles—Mailbox (MB), Hub Transport (HT), and the Client Access Server (CAS) on a single server.

Silver Tier: The Silver Tier is targeted to smaller environments supporting cost-sensitive users with limited IT budgets. For increased availability, the Silver Tier provides support for database replication using DAGs, but limits costs by combining the Exchange roles using the all-in-one approach with two servers.

Gold Tier: The Gold Tier is targeted toward a wide range of customer scenarios—from small to large organizations needing a high level of data protection and availability for their important users. This tier expands the servers and storage high availability options, along with dedicated Exchange server roles.

Platinum Tier: The highest level in the tiered solutions matrix is the Platinum Tier which is targeted to enterprise deployments requiring the highest levels of data protection for their most critical users. The Platinum Tier builds on the Gold Tier offering additional database copies and new storage options including RAID-less JBOD with enterprise server and storage hardware and Exchange’s high availability features.

The table lists the details of each service tier.

features. The table lists the details of each service tier. Table 9. Tiers of service  

Table 9.

Tiers of service

 

Tiers of

Database

Database

Combined

Disk types

Storage type

Combined

service

Availability

copies

Exchange

LOG and

Group

roles

Database

 

LUN

Bronze

No

1

MB/HT/CAS

SAS/MDL-SATA

RAID

No

Silver

Yes

2

MB/HT/CAS

SAS/MDL-SATA

RAID

No

Gold

Yes

2

Dedicated

FC/SAS/MDL-

RAID

Yes

 

SATA

Platinum

Yes

3

Dedicated

FC/ SAS/MDL-

RAID/ JBOD

YES/JBOD

 

SATA

FC/ SAS/MDL- RAID/ JBOD YES/JBOD   SATA Microsoft® Exchange Server 2010 is a complex application

Microsoft® Exchange Server 2010 is a complex application that requires extensive knowledge of deployment strategies, interdependencies, and server performance-related issues. To assist customers with proper server and storage sizing for their Exchange Server 2010 deployments, HP has developed the HP Sizer for Microsoft Exchange Server 2010. The algorithms used in this tool are based on testing of Exchange Server 2010 server roles and technologies, HP’s extensive experience with Microsoft Exchange, and intricate knowledge of server CPU, memory, and I/O subsystems.

The sizing algorithms in the HP Sizer are based on the recommendations from the following Microsoft TechNet articles:

To download the HP Sizer for Microsoft Exchange Server 2010, please visit HP's Microsoft Exchange Server solutions page or the ActiveAnswers Sizers page on hp.com.

Solution Block Architectures

Architectural Principles

The HP tiered solutions program for Microsoft ® Exchange Server 2010 was designed to provide broad applicability across a wide range of server and storage infrastructure platforms. However, this service classification framework can be applied within a single organization to meet various departmental and geographical requirements.

With HP Converged Infrastructure, this service-level concept can be taken a step further and used to create and publish service templates. These templates then define the various Exchange services that can be deployed in the organization. Typically for most organizations, Exchange is a static environment and servers are designed and installed with 2-3 years growth in mind. With HP Converged Infrastructure, additional Exchange resources can be added and provisioned within hours when needed—rather than deployed and paid for up-front—waiting for demand.

Specific Configurations

This Converged Infrastructure Reference Architecture for Exchange 2010 uses the Gold Tier service level described above and is designed to support 5,000 1GB mailboxes.

The following table highlights the Exchange building-block resource requirements for this HP Converged Infrastructure Reference Architecture.

Table 10. Exchange 2010 server role resource requirements Exchange # of Physical Core Minimum Networks

Table 10. Exchange 2010 server role resource requirements

Exchange

# of

Physical

Core

Minimum

Networks

Storage (per server)

Server Role

Servers

Server

count

Memory

 

or VM

(GB)

Client

3

VM or

4

8 GB

3 NICs

1 LUN

Access

Physical

Servers

-

2 Production

1 x 50 GB LUN

networks

(OS/Exchange binaries)

(NLB cluster)

-

1 Management

network

 

Hub

2

VM or

4

4 GB

3 NICs

2 LUNs

Transport

Physical

-

2 Production

1 x 50 GB LUN

Servers

networks (Teamed)

(OS/Exchange binaries)

-

1 Management

1

x 100 GB LUN

network

(HT queue DB and transaction logs)

Mailbox

2

Physical

8

32 GB

5 NICs

6 LUNs

Servers

-

2 Production

1 x 50 GB LUN

networks (Teamed)

(OS/Exchange binaries)

-

2 Replication

4

x 2000 GB LUN

networks (Teamed)

(Exchange DB/Log data

-

1 Management

storage)

network

x 2000 GB LUN (Restore)

1

storage) network x 2000 GB LUN (Restore) 1 For this Exchange 2010 service, a mixture of

For this Exchange 2010 service, a mixture of physical server and VM resources can be allocated. The Exchange CAS and HT server roles can be deployed on either virtual machines or physical servers. The Exchange mailbox role is provisioned on physical servers.

With Exchange 2010, an increased workload is placed on the Exchange 2010 CAS server, as this role now serves as the MAPI endpoint for Outlook client connections. With this increased workload, Microsoft recommends a 4:3 mailbox server to CAS processor core ratio. Each server is also configured with 8 GB of memory to satisfy the 2 GB per core recommendation. From a networking perspective, three NICs are defined for the CAS server role. One NIC serves as a dedicated management NIC (optional), while the other two are used on the production network to form the network load balancing (NLB) cluster for the three CAS Exchange servers.

For the Hub Transport (HT) server role, two servers with four processor cores will be over-provisioning against the recommended mailbox server to HT server core ratio of 5:1 (or 7:1 without antivirus). In this case, a single server can accommodate an Exchange workload based on a typical medium to heavy Exchange user profile. However, a minimum of two servers are required to meet the availability requirements of the Gold Tier service. To support the network requirements for the HT server role, three NICs are defined. One NIC serves as a dedicated management network (optional), while the other two NICs are teamed for production traffic.

For the two mailbox servers, the design recommendations to support 5,000 users are 8 cores and 32 GB of memory per server. During normal operation, each server hosts 2,500 users. However, each mailbox server is sized to support 5,000 active users in the event of a server failure. From a networking perspective, each mailbox server will require five NICs. One NIC is dedicated for

management purposes (optional), and the other four will form two NIC teams, one to support the production network and the other for the DAG replication and heartbeat traffic.

To support the storage requirements for 5,000 users with 1 GB mailboxes, 20 databases hosting 250 users per database will be configured. During normal operation, each mailbox server can host 10 active database copies (supporting 2,500 users) while maintaining a copy of each of the 10 active database copies on the alternative server. These databases are spread evenly across the four, 2 TB LUNs, with five database copies and associated log files configured on each LUN. In the event of a single server going offline, the mailbox servers are designed to handle the peak load of 5,000 users across 20 active database copies on a single server. Each server also requires a 2 TB LUN for database maintenance or restore operations.

Note the processor core ratios mentioned above do not take into account availability requirements. In this case, only one mailbox server with 8 cores is needed to host all 5,000 mailboxes. Using the 4:3 processor core ratio for mailbox to CAS role, only two CAS servers are required, however three are recommended to provide availability for that role. Likewise though only one HT server is required, 2 are suggested for high availability.

Adapting to Your IT Operations

This reference architecture should be adapted to meet your specific needs. The requirements of your organization can dramatically change the building-block requirements. Key factors include mailbox size, user profile, client access methods, retention policies, and availability/recovery requirements.

To support larger environments, you would simply deploy additional quantities of the specific building blocks as needed. For example, to support 20,000 mailboxes, up to eight mailbox servers are required (depending on the actual DAG design). Additional quantities of the other roles would also be required, but not necessarily four times the reference architecture, as redundancy is built into the base quantities. The exact number should be based on Microsoft processor core ratio recommendations and the number of mailbox servers. It may also be possible to scale up the original building blocks to support more users, rather than scaling out.

Specific HP Products Applied to the General Architecture

Gold Tier: 5,000 Mailboxes

The following diagram shows HP ProLiant BL460 G6 Server Blades and the EVA4400 as resources.

Gold Tier: 5,000 Mailboxes The following diagram shows HP ProLiant BL460 G6 Server Blades and the

Figure 16. Exchange 2010 enterprise topology

Windows 2008 SP2 Mailbox Servers ProLiant BL450c G6 Hub Transport Servers ProLiant BL460c G6 8
Windows 2008 SP2
Windows 2008 SP2
Windows 2008 SP2 Mailbox Servers ProLiant BL450c G6 Hub Transport Servers ProLiant BL460c G6 8 cores
Mailbox Servers ProLiant BL450c G6 Hub Transport Servers ProLiant BL460c G6 8 cores / 36
Mailbox Servers
ProLiant BL450c G6
Hub Transport Servers
ProLiant BL460c G6
8
cores / 36 GB
Client Access Servers
StorageWorks 4400
4
cores / 6 GB
5
NIC’s
ProLiant BL460c G6
EVA
3
NIC’s
1
x 50GB LUN (OS)
4
cores / 12 GB
Approx 11TB useable
1
x 50GB LUN (OS)
4
x 2TB LUN
3
NIC’s
capacity for Exchange
1
x 100GB LUN
(DB/logs)
1
x 50GB LUN (OS)
(Queue)
1 x 2TB LUN (Restore)
1 x 50GB LUN (OS) (Queue) 1 x 2TB LUN (Restore) HP Converged Infrastructure “touch-points” The

HP Converged Infrastructure “touch-points”

The Microsoft Exchange configuration illustrated and described above is based upon the core HP Converged Infrastructure reference architecture and utilizes many of its features.

The network element of the Converged Infrastructure is embodied in the use of Virtual Connect. The Virtual Connect Manager manages virtual I/O connections, both Ethernet and FC SAN, to physical blades. This allows for dynamic assignment of network and storage resources to Exchange instances.

The ProLiant BL460c G6 provides a modular approach in terms of compute resources. These servers can be provisioned to support both physical and virtual operating environments (see Virtualization use case). In this example, all roles are deployed on physical servers. Note that the memory specifications differ slightly from the previous table to match optimal configurations with IntelXeon5500 or 5600 series processors.

The HP Converged Infrastructure core approach is further illustrated with the StorageWorks EVA4400 storage solution provides storage LUN virtualization. Connectivity from the BL460c G6 to the EVA4400 is via Fibre Channel and Virtual Connect Fibre Channel. In the future, connectivity can be via FCoE and HP FlexFabric.

Note the HUB and CAS servers could be virtual machines rather than physical servers as depicted here. If virtualized, memory requirements would follow the guidelines in the following table and network redundancy would be provided at the host level, not within the virtual machines. To provide

full redundancy, the five virtual machines should be deployed across at least three physical host servers to achieve functionality in the event one host server is offline.

This solution can also be placed in the HP Converged management context described in Figures 6-7 and related discussion for infrastructure, application, environmental and other monitoring.

While a building block approach based on HP Converged Infrastructure is used for this particular Microsoft Exchange architecture, additional reference architectures based on HP Converged Infrastructure, including a particular solution block for BladeSystem Matrix and Microsoft® Exchange Server 2010 Reference Configuration are available. These can be found on the Microsoft Exchange

Virtualization: Generalized Use Case

Solution Architectural Principles

Basic solution description and business need

Virtualizing servers, storage and networking enables an IT organization to quickly address changing business demands. Virtualization also reduces costs by pooling resources - not just servers. Servers, storage, and networking are all resources that need to be optimized and used efficiently and effectively. Virtualization technology must support x86/x64 infrastructure for solutions like VMware vSphere, Citrix Xen, or Microsoft Hyper-V, but also 64-bit technologies like HP-UX, large scale Oracle, SAP, and others. A single virtualization umbrella is required to manage all of these.

In line with HP Converged Infrastructure principles, this optimization leads tactically to reduced power and cooling costs, and strategically to deferring or eliminating capital costs associated with over- provisioning. At the same time, it improves the productivity of the IT staff.

Finally, virtualization success addresses virtualizing the client. This means improving manageability and security of the desktop user, locally or remotely, in a way that is functional on desktops, laptops, tablets PCs, and others. When combined with HP's broad and proven thin client options, client virtualization reduces cost and risk at every level. User data resides in the data center where it is protected and backed up appropriately. New operating system deployments are reduced from months to days or even hours. User support becomes as simple as rebooting the client or if that client is defective, simply moving to another client and getting back to work with the same user settings and data the end user expects.

Basic Virtualization architecture principles and considerations

Virtualized

Virtualization technologies such as VMware vSphere and Microsoft Hyper-V are fundamental building blocks toward converged infrastructure where servers - physical or virtual - are logical resources to be deployed and consumed based on business demand. However, virtualization goes well beyond that layer - in order to provide a truly flexible environment, the network layer and storage resources must be flexible as well.

Resilient

Virtualized solutions must be highly available, robust and extensible. Examples of this:

Add a second BladeSystem c7000 for enclosure redundancy.

Move an enclosure to a redundant site and use SRM from VMware or Cluster Extension software (CLX) from HP to extend vSphere or Hyper-V virtualization farms across multiple sites for metro or continental cluster types of solutions.

Add a second EVA at a redundant site and use SRM or CLX to automate the high availability of storage resources in parallel with the HA of server resources.

Utilize HP StorageWorks SVSP or EVA Cluster to extend scalability beyond a single EVA and incorporate heterogeneous high availability across multiple sites.

Orchestrated

Essential to virtualization is the ability to deploy, configure, and consume compute, storage, and networking capacity based on real world business application demand in a fraction of the time normally required to assess needs and gaps, analyze outstanding solution, and then effect change for the business. IT responses to business needs can be reduced from weeks or months to hours or minutes. Increase application availability.

Open

Different server virtualization solutions can be utilized including Microsoft Hyper-V or VMware vSphere - SystemCenter or VirtualCenter. HP management tools and software integrate deeply with management technologies from Microsoft, VMware, or HP to provide a single pane of glass.

Modular

Scale up compute power by adding blades. Scale up and out storage I/O and capacity by adding EVAs and using HP's SVSP technology to tie it all together. Scale up networking I/O and availability by adding more FlexFabric mezzanine adapters and VC FF modules and then pulling them back to an HP networking core. Manage them through Insight Control and tight integration with other management tools to provide an outstanding virtualization environment.

Sizing and scaling principles

As workloads are virtualized on newer and more capable hardware and then managed very effectively, resources can be dynamically allocated based on ongoing analysis so that SLAs to the business can be met. In addition, downtime is reduced, business continuity becomes more easily attainable and provable, and security is improved - all resulting in greatly reduced business risk.

Consolidation, disaster recovery - especially for older legacy applications, along with test and development, are why companies traditionally look to virtualization. In the broader context of putting technology to work for your business, these are first steps toward automated, truly converged, infrastructure. Virtualization technologies such as VMware vSphere, Citrix XenServer, and Microsoft Hyper-V enable a dynamic environment where VMs can be moved and load balanced across the physical infrastructure. These capabilities are predicated on having a solid and redundant shared storage environment and a very predictable high performance network. Building an IT Infrastructure geared to delivering the SLAs to the business requires a tightly coupled set of infrastructures, management tools, and operating environments. These must be coupled with the policies, security, and network control required to dynamically allocate resources based on application need.

Solution Block Architectures

Architectural Principles

The best designed and delivered Virtualization platform today is also the data center of the future. This means a simple to build and manage operating environment for your business applications and services. This goes beyond just the basic IT infrastructure for virtualization to address the many layers of a successfully fully integrated and managed IT utility. Servers, storage, networking, virtualization software, and most critically, application aware management are all required. This yields an excellent IT operational control of not just the hardware, but also the virtualization layer, operating systems, and applications. IT can now provide the overlaying applications as a service to the business.

The following figure shows a conceptual view of the three main resource pools related to virtualization – servers, storage and networking. These are combined together with very strong management to enable a virtualization engine scalable across application requirement - whether that is a Windows, Linux or the very largest of HP-UX applications.

Figure 17. HP Converged Infrastructure: virtualization – conceptual view Virtualized Applications Non Virtualized

Figure 17. HP Converged Infrastructure: virtualization – conceptual view

Virtualized Applications Non Virtualized Applications UNIX App Oracle App App 1 App 3 Linux App
Virtualized Applications
Non Virtualized
Applications
UNIX App
Oracle App
App 1
App
3
Linux App
Windows App
App2
Physical & Virtualization Management Layer
HP Insight
BladeSystem
Control
c7000 Enclosure
HP BladeSystem c7000 Enclosure
Software
AMD/Intel or
(Server Resource Pool)
Insight
Integrity server
Dynamics
blades
Hypervisor
Hypervisor
Hypervisor
Hypervisor
Matrix
Storage Blades
Integrity VSE
P4000 VSA
BTO software
FlexFabric
FlexFabric
Virtual Connect
Virtual Connect
Network resource pool
1/10 GigE
FC
iSCSI/FCoE
Datacenter
Cores, Distribution, Edge,
Top of Rack
Ethernet
TippingPoint
Shared Storage
(Resource Pool)
ISCSI - FC
XP-EVA "FC",
P4000 -
"ISCSI", X9000
- NAS", SVSP
Core
SVF
Blade
Server
Blade
Server
Blade
Server
Blade
Server
Blade
Server
Blade
Server
Blade Server Blade Server Blade Server Blade Server Specific HP Products Applied to the General Architecture

Specific HP Products Applied to the General Architecture

Small to medium build configuration

A small or medium build can be cost effectively achieved by using all of the principles and practical implementation of the Converged Infrastructure model. This build employs HP BladeSystem x64 Blades running VMware vSphere, Microsoft Hyper-V, or Citrix Xen at the hypervisor level and the HP P4000 based iSCSI storage systems. CPU and memory capacity can be scaled up by adding additional Blade servers. Additional IOPS (I/Os per second) and storage capacity can be scaled up and out by adding additional P4000 nodes.

This configuration offers scale out CPU, memory, network and storage as needed for workloads. It delivers data center level high availability by design. Multi-site business continuity capability can be achieved by extending the present platform to a redundant site.

Figure 18. HP Virtualization Reference Configuration: Small to Medium Enterprise Build HP Networking Core A5820

Figure 18. HP Virtualization Reference Configuration: Small to Medium Enterprise Build

HP Networking Core A5820 (10GE) - IRF HP TippingPoint IPS 10 GB iSCSI
HP Networking Core
A5820 (10GE) - IRF
HP TippingPoint IPS
10 GB
iSCSI

Server, OS,

Network,

Storage, Hypervisor

Agnostic

HP Networking Virtual Connect FlexFabric
HP Networking
Virtual Connect
FlexFabric
Linux/Windows VMs Hyper-V – Xen – VMware ProLiant BL465 G7 24 cores / 96GB
Linux/Windows VMs
Hyper-V – Xen – VMware
ProLiant BL465 G7
24 cores / 96GB
StorageWorks P4000 G2 SAN 3-4 TB useable capacity
StorageWorks P4000 G2
SAN
3-4 TB useable capacity
96GB StorageWorks P4000 G2 SAN 3-4 TB useable capacity Architectural Assumptions This configuration is based on

Architectural Assumptions

This configuration is based on HP architectural best practices for this class of environment. It is based on the following assumptions:

VM typical - 1 processor, 4GB ram, shared LUNs (VMDK/CSV).

Minimum 3 nodes for N+1 redundancy.

iSCSI block level storage, N+1 storage node availability.

Most applications and software infrastructure virtualized, some applications and infrastructure may still run on physical hardware.

Virtualization aware network fabric.

Large enterprise build configuration

A larger medium or enterprise build will, in most cases, employ Fibre Channel storage for very high performance. It frequently includes enterprise level UNIX applications such as SAP, Oracle, Weblogic, and others. The ability to deliver the right platform to the right workload is the key to optimizing IT infrastructure investment. Applications that fit into the industry standard architecture such as Exchange, SQL, MySQL, etc can leverage the low cost and flexibility of the ProLiant platform. Within the same infrastructure applications such as scale up, UNIX-based, Oracle and SAP environments can be served by the highly scalable SMP Integrity platforms. Common networking, storage, and management enable a holistic approach to data center consolidation, virtualization, and management.

The configuration shown in Figure 19 is based on central management of all datacenter resources including both x64 and IA64 compute resources. These resources are sourced in the HP Converged Infrastructure elements through common server, storage, and networking, as well as resiliency, and management platform.

Key attributes include:

Scale out CPU, memory, network and storage as needed for workloads.

Data center level high availability by design.

Multi-site business continuity capable through the extension of the present platform to a redundant site.

the extension of the present platform to a redundant site. Figure 19. HP Virtualization Reference Configuration:

Figure 19. HP Virtualization Reference Configuration: Large Enterprise Build

Converged Infrastructure Reference Architecture

HP Networking Core A9500 – IRF TippingPoint IPS
HP Networking Core
A9500 – IRF
TippingPoint IPS
HP Networking Core A9500 – IRF TippingPoint IPS Server, OS, Network, Storage, Hypervisor Agnostic FC

Server, OS,

Network,

Storage, Hypervisor

Agnostic

IPS Server, OS, Network, Storage, Hypervisor Agnostic FC 10 GB HP Networking Virtual Connect FlexFabric HP-UX:
IPS Server, OS, Network, Storage, Hypervisor Agnostic FC 10 GB HP Networking Virtual Connect FlexFabric HP-UX:
IPS Server, OS, Network, Storage, Hypervisor Agnostic FC 10 GB HP Networking Virtual Connect FlexFabric HP-UX:
IPS Server, OS, Network, Storage, Hypervisor Agnostic FC 10 GB HP Networking Virtual Connect FlexFabric HP-UX:
FC
FC
10 GB HP Networking Virtual Connect FlexFabric HP-UX: Integrity BL870c i2 16 cores / 96GB
10 GB
HP Networking
Virtual Connect
FlexFabric
HP-UX: Integrity BL870c i2
16 cores / 96GB
Linux/Windows VMs
Hyper-V – Xen – VMware
ProLiant BL465 G7
24 cores / 96GB
StorageWorks 4400 EVA 8-10 TB useable capacity
StorageWorks 4400 EVA
8-10 TB useable capacity
/ 96GB StorageWorks 4400 EVA 8-10 TB useable capacity Architectural Assumptions This configuration is based on

Architectural Assumptions

This configuration is based on HP architectural best practices for this class of environment. It is based on the following assumptions:

VM typical size - 1-4 cores, 4-96 GB ram, shared LUNs (max x86 VM size = 8 cores).

Minimum 3 nodes for N+1 redundancy, HP Virtual Machines, Industry standard hypervisors.

Fibre Channel block level storage.

Mix of HP-UX, Microsoft, Linux applications and operating systems.

Combination of physical and virtual environments.

Virtualization aware network fabric.

Multi-data center build

The largest enterprise shops might require a deep and wide mix of platforms and technologies to meet their application requirements. Distributed data centers with multi-site high availability are the backbone of these environments. The ability to be able to serve each application with the resources it needs to meet the SLAs of the business is absolutely critical.

The next figure illustrates a typical configuration. The resource pool will include ProLiant and Integrity server blades and Superdome 2 as well as rackmount servers such as the HP ProLiant DL980. This allows seamless management of sub-CPU partitioned resources alongside highly scalable multi-CPU

platforms. Tiered storage based on cost, performance and data availability requirements is very typical. Technologies such as SVSP come into play here as all Fibre Channel attached storage resources can be unified and managed via the SVSP platform through its split path architecture. This enables storage techniques such as Thin Provisioning, snapshots, clones, synchronous replication across a multi-vendor storage pool. The HP approach to converged infrastructure preserves existing investments in storage technologies and can potentially save hundreds of thousands of dollars in array specific licensing.

HP's Insight Control software for platform management tied into HP's BTO software provides a highly- scalable management solution that meets the demanding business SLAs and requirements of this environment.

This configuration includes dedicated systems for high performance applications that can't fit into the constraints of a virtualized layer or require the highest levels of memory and I/O capabilities. Key attributes include:

Virtualized platforms for applications and infrastructure services. Scale out CPU, memory, network and storage as needed for workloads.

Data center level high availability by design.

Multi-site business continuity capable through the extension of the present platform to a redundant site.

Central management of all data center resources including both x64 and IA64 compute resources behind a common blade, networking, storage, and management platform.

Ability to support the very largest application requirements in the x64 as well as IA64 Windows, Linux and Enterprise Unix environments.

Figure 20. HP Virtualization Reference Configuration: Multi-Data Center Build HP Networking Core A12500 – IRF

Figure 20. HP Virtualization Reference Configuration: Multi-Data Center Build

HP Networking Core A12500 – IRF TippingPoint IPS SITE 1 10 GB
HP Networking Core
A12500 – IRF
TippingPoint IPS
SITE 1
10 GB
HP Networking Core A12500 – IRF TippingPoint IPS SITE 2 10 GB
HP Networking Core
A12500 – IRF
TippingPoint IPS
SITE 2
10 GB
FC Virtual Connect FlexFabric
FC
Virtual
Connect
FlexFabric
FC Virtual Connect FlexFabric
FC
Virtual
Connect
FlexFabric
HP-UX: Integrity BL870c i2 Integrity Superdome 2 Linux/Windows: ProLiant BL465 G7 DL980 G7 Storage StorageWorks
HP-UX:
Integrity BL870c i2
Integrity Superdome 2
Linux/Windows:
ProLiant BL465 G7
DL980 G7
Storage
StorageWorks
EVA and XP
SVSP
BL465 G7 DL980 G7 Storage StorageWorks EVA and XP SVSP Architectural Assumptions This configuration is based

Architectural Assumptions

This configuration is based on HP architectural best practices for this class of environment. It is based on the following assumptions:

VM typical size - 2-8 cores, 48-384 GB ram, shared LUNs (8 cores max).

Multiple data centers for business continuity and DR.

HP Integrity Virtual Machines (hypervisor for HP-UX), industry standard hypervisors (ex. VMware vSphere, Microsoft Hyper-V, Citrix Xen).

Electrical isolation and logical partitioning available on Integrity servers through nPARs and vPARs.

Fibre Channel pooled storage (tiered), including “always on” capability.

Mix of Unix, Microsoft, Linux applications and operating systems.

Virtualization aware network fabric.

HP Converged Infrastructure “touch-points”

HP Converged Infrastructure includes infrastructure virtualization, and application-specific virtualization is delivered for specific configurations. In all of the above virtualization solution cases, the HP Converged Infrastructure principles, as well as the core reference architecture, are employed to build a flexible, and resilient virtualized server, storage and networking environment. These use cases are managed within the broader context discussed with Figures 6 and 7. Core components are utilized as described below.

Converged Network / FlexFabric FlexFabric extends network intelligence away from the core switches and closer to the application, simplifying complex hierarchical network infrastructures.

For example, a 100MB connection can be created for management, a 2GB connection for iSCSI, 2GB for Live Migration or v-Motion, etc. Using and teaming 1GB NICs results in a waste of

bandwidth for connections that require very little bandwidth ( such as management) and often under provisioning for components that need more but are limited by available NICs and ports. Virtual Connect FlexFabric allocates the bandwidth needed and delivers high availability through teaming and other means.

Provisioning / Matrix Operating Environment As IT organizations progress to high levels of resource optimization and control, shared-services engines that quickly and easily provision while managing application environments become important. This environment is built from HP infrastructure, mission-critical, and business service software portfolios to meet the needs of heterogeneous enterprise IT environments across multiple delivery models such as on-site, outsourced, and over the cloud. For the virtualization layer, it also allows multiple hypervisors to be used and virtual machines moved not only across servers, but also different hypervisor technologies.

Capacity Management / Smart Grid The virtualization solutions described can all be placed in context of the HP Converged Infrastructure intelligent, energy-aware environment. IT and facilities can optimize and adapt energy use, reclaim trapped facility capacity, and reduce energy costs. This enables more efficient use of existing capacity and capital investments, enabling organizations to support business growth with more applications and services within the same energy footprint.

Businesses can accurately visualize and dynamically control data center energy use and environmental impact across the entire data center to increase power density, efficiency, and capacity use. The Data Center Smart Grid extends Thermal Logic technology from IT systems to environmental monitoring and control across the facility, and eventually to third-party IT and data center systems.

Server and Storage Resource / Virtual Resource Pool Virtual Resource Pools are created from virtualized compute, storage, I/O, and networking systems that can be combined, divided, and repurposed to match any application demand quickly and efficiently. When combined with a virtual machine layer, this enables businesses to support a shared- service model with a common pool of adaptive, virtualized resources based on standards that can be optimized for all types of enterprise, cloud, and high-performance computing (HPC) applications. This enables you to meet the dynamic requirements of a HP Converged Infrastructure and shared-services model by increasing the scalability, flexibility, and utilization of resources.

Cloud: Generalized Use Case

HP Converged Infrastructure supports cloud deployments. The use case configurations provided here focus on “private cloud” to complement the other use cases, which are also enterprise focused. Other cloud use cases will be documented separately. Definitions in the following subsections on cloud computing definitions, cloud characteristics, cloud portfolio architecture models, and cloud deployment models are quoted from the “The NIST Definition of Cloud Computing”, US National

Definitions

Cloud computing definition

Cloud computing is a model for enabling convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released as a service with minimal management effort or service provider interaction.

Other definitions that might be encountered when searching on the internet are:

The cloud is the next stage in the evolution of the Internet – the means through which everything will be delivered as a service – from computing power to business processes to personal interactions. Wherever, however, and whenever you need it.

Cloud computing is an environment where highly scalable and elastic services can be easily consumed over the network (typically the internet) through a low-touch, self-service, pay-per-use business model.

The approach for achieving multi-tenancy has a pivotal role in defining the software infrastructure of the cloud. There are two design centers for achieving multi-tenancy:

Multi-tenant application: where multiple tenants can be supported within a single instance of the application. This cloud design center relies on a highly integrated design with high availability architected in the application. Examples of multi-tenant applications are Salesforce.com and HP Snapfish.

Multi-tenant infrastructure: where multiple tenants are hosted across multiple isolated instances of the application. Like HP’s Converged Infrastructure, this cloud design center relies on automation, virtualization, and workload management to provide an infrastructure container that can be flexibly created, scaled, and modified, and in which the user can run a full range of existing applications.

Cloud characteristics

Not every IT function that is packaged or delivered as a service is automatically a cloud service. Cloud services differentiate from other (shared) services through specific architectural characteristics. Below are five key criteria for cloud services:

On-demand self-service: A consumer can unilaterally provision computing capabilities, such as server time and network storage, as needed automatically without requiring human interaction with each service’s provider.

Broad network access: Capabilities are available over the network and accessed through standard mechanisms that promote use by heterogeneous thin client platforms (e.g. mobile phones, laptops, and PDAs).

Resource pooling: The provider’s computing resources are pooled to serve all consumers using a multi-tenant model, with different physical and virtual resources dynamically assigned and reassigned according to consumer demand. The customer generally has no control or knowledge over the exact location of the provided resources but may be able to specify location at a higher

level of abstraction (e.g., country, state, or data center). Examples of resources include storage, processing, memory, network bandwidth, and virtual machines.

Rapid elasticity: Capacities and capabilities can be rapidly and elastically provisioned to quickly scale up and rapidly released to quickly scale down. To the consumer, the capacities offered by cloud services often appear to be infinite and can be purchased in any quantity at any time, most often without planning or notice ahead of time.

Measured service: Cloud systems automatically control and optimize resource use by leveraging a metering capability at some level of abstraction appropriate to the type of service (e.g., storage, processing, bandwidth, and active user accounts). Resource usage is monitored, controlled, and reported providing transparency for both the provider and consumer of the utilized service.

Cloud portfolio architecture models

The cloud portfolio architecture model is essentially the way cloud service providers segment and structure their cloud service portfolio in a layered fashion. Cloud service providers may decide to play in one of more of these service segments.

Cloud Software as a Service (SaaS): The capability provided to the consumer is to use the provider’s applications running on a cloud infrastructure and accessible from various client devices through a thin client interface such as a Web browser (e.g., web-based email). The consumer does not manage or control the underlying cloud infrastructure, network, servers, operating systems, storage, or even individual application capabilities, with the possible exception of limited user-specific application configuration settings.

Cloud Platform as a Service (PaaS): The capability provided to the consumer is to deploy onto the cloud infrastructure consumer-created applications using programming languages and tools supported by the provider (e.g., java, python, .Net). The consumer does not manage or control the underlying cloud infrastructure, network, servers, operating systems, or storage, but the consumer has control over the deployed applications and possibly application hosting environment configurations.

Cloud Infrastructure as a Service (IaaS): The capability provided to the consumer is to rent processing, storage, networks, and other fundamental computing resources where the consumer is able to deploy and run arbitrary software, which can include operating systems and applications. The consumer does not manage or control the underlying cloud infrastructure but has control over operating systems, storage, deployed applications, and possibly select networking components (e.g., firewalls, load balancers).

Cloud deployment models

There are multiple forms of cloud computing:

Private cloud: The cloud infrastructure is owned or leased by a single organization “members only”, delivered over a private network and is operated solely for that organization, either by that organization (internal private cloud) or by an outsourcer (external private cloud). Private clouds can provide many of the benefits of cloud computing without lots of disadvantages (security, reliability, trust). An internal private cloud combines the operating and organizational model of IT shared services with attributes of a cloud based service architecture.

Community cloud: The cloud infrastructure is shared by several organizations and supports a specific community that has shared concerns (e.g., mission, security requirements, policy, and compliance considerations). For example, different universities could make equipment available to a community cloud to do technical computing for students in those universities.

Public cloud: The cloud infrastructure is owned by an organization selling cloud services to the general public, a large industry group or global class.

Hybrid cloud: The cloud infrastructure is a composition of two or more clouds (internal, community, or public) that remain unique entities but are bound together by standardized or proprietary technology that enables data and application portability (e.g., cloud bursting).

Solution Architectural Principles

Basic solution description and business need

Implementing virtualized infrastructure brings a new set of challenges.

Proliferation of virtual machines and platforms. This includes issues related to virtual sprawl and lack of tools and resources to manage security and compliance of virtualized machines. Software license compliance related to virtual infrastructure adds another layer of complexity. Hardware investments may slow down, but are not eliminated; adding virtual machines requires ongoing addition of processing, memory, storage, and networking infrastructure.

Staffing and people concerns - Just like any other technologies that purport to provide efficiencies and cost savings, virtualization faces resistance from IT staff concerned about downsizing. In addition, virtualization adds yet another complex new technology for IT professionals to learn and master, while they struggle to keep up with managing their existing infrastructure.

Process issues - Planning to manage physical and virtual infrastructure in an integrated manner, while maintaining service performance and quality, is a challenge that most organizations face. In addition, compliance of virtualized infrastructure must be managed just like the physical infrastructure requiring management tools that provide a seamless view of the entire environment.

The figure below builds on aligning Supply and Delivery with the Demand and takes a more business Centric approach to leverage of the Cloud and Services being delivered on the Cloud plus addressing the above new challenges.

Figure 21. HP Software Private/Public Cloud Functional Concept Reference Architecture Service Catalog and Portal Layer

Figure 21. HP Software Private/Public Cloud Functional Concept Reference Architecture

Service Catalog and Portal Layer Service Health Service Billing Business Service Offering Enterprise Service
Service Catalog and Portal Layer
Service Health
Service
Billing
Business
Service
Offering
Enterprise Service Portals
Availability
Portfolio
Catalog
Center
Mgmt
Service
Service Usage
Service
Service
Systinet
Service Desk
Manager or
MOAB, IUM
Request Mgmt
Access
Service Manager
Systinet
Revenue
Order Mgmt
Charging
User Mgmt
Settlement
Service Portal
MOAB, IUM
Customers Env
ITFM
Backend
Service Configuration and
Activation
Delivery Assurance
Operations Orchestration, MOAB
Business Availability Center, Operations
Manager, SiteScope, MOAB
Resource Provisioning and Configuration
Server Automation, Network
Automation, Storage Essentials, Insight
Orchestration
Resource Management
MOAB, Insight Dynamics, IUM
CMDB, Discovery
Resources (Infrastructure, Platform, Software, Information or Business)
DemandDeliverySupply
Service Governance
PPM, Service Manager
Service Management
CMDB, Discovery, Service Manager
POWER & SERVERS STORAGE NETWORK COOLING
POWER &
SERVERS
STORAGE
NETWORK
COOLING

SOFTWARE

INFORMATION

CLOUD

CMDB, Discovery, Service Manager POWER & SERVERS STORAGE NETWORK COOLING S O F T W A
COOLING S O F T W A R E INFORMATION CLOUD • Demand Definition: This layer

Demand Definition: This layer represents all aspects of capturing and meeting demand from your customers (internal or external), captured and made available as configure-to-order – that is to say highly standardized and simplified services that can be readily delivered at the press of a button in minutes/hours not days/weeks/months. The demand layer encompasses the lifecycle of the relationship between the consumers and service instances.

Delivery Definition: This layer represents all aspects of assembling and managing the underlying resources into a form that can be offered as a useful/appropriate service to consumers in an automated fashion. The delivery layer encompasses the lifecycle of both the service offering and the service instance and binds the resources to the appropriate service instances.

Supply Definition: This layer represents the building blocks (resources) that a service is comprised of – which could be combination of compute/storage/network, or an existing IaaS/PaaS – internally or externally to the company. The supply layer encompasses the lifecycle of resources – which may be assets of the company or sourced as services from external entities.

Service Governance Definition: This represents key aspects of service strategy and service design (ITIL v3) – managing the portfolio, suppliers, partners, markets, customers, finance, enterprise architectures and policies, demand, cost models, investments, service lifecycle (inception to retired), security requirements, business continuity, etc.

Service Management Definition: This represents key aspects of service transition, service operations and continual service improvement (ITIL v3) – change, configuration and release management, incident/problem management, event management, service level monitoring, continuity/capacity/security/performance management.

Another way to think about the distinction between Service Governance and Service Management is time frame – governance is about the big picture over the entire lifecycle of services from strategy

generation through to design. Service management is largely about operating the service environment: 1) getting things into production, 2) making sure it is meeting SLAs, 3) fixing it when it is not, and 4)providing feedback into the lifecycle for improvements.

Key point: Building a cloud offering is not performed in isolation – the full range of service management and governance principles and products apply and the cloud offering must integrate with these.

Sizing and scaling principles

A private cloud built on HP’s Converged Infrastructure can address on several organizational or business requirements. The most common requirements addressed by an HP Converged Infrastructure private cloud are: compliance with open standards, reducing cost, operational efficiency, applications resiliency and workload elasticity.

Scalability requirements can range from 10K to 100Ks of servers and can cover petabytes of storage. Services are designed to accommodate multiple tenants, to be flexible in terms of capacity and number of tenants, and are typically configured and accessed by users through a self-service portal. By contrast, traditional enterprise computing is focused on single-tenant architectures, sized for a defined workload and number of users.

HP Converged Infrastructure “touch-points”

Data center automation solution components The HP Data Center Automation “DCA” Center automates the management of every part of the cloud infrastructure, including servers, network devices, and other assets. It is specifically designed to meet the needs of large, heterogeneous, geographically distributed environments. There are several applications that power the HP Data Center Automation Center, as shown below. The automation products will be discussed here.

Figure 22. HP Data Center Automation Center Operations Orchestration Network Server Storage Automation

Figure 22. HP Data Center Automation Center

Operations Orchestration

Network

Server

Storage

Automation

Automation

Essentials

BSM

Monitoring events and alerts

Automation Essentials BSM Monitoring events and alerts ITSM Help desk/ change requests BSA Essentials Community |
Automation Essentials BSM Monitoring events and alerts ITSM Help desk/ change requests BSA Essentials Community |
Automation Essentials BSM Monitoring events and alerts ITSM Help desk/ change requests BSA Essentials Community |
Automation Essentials BSM Monitoring events and alerts ITSM Help desk/ change requests BSA Essentials Community |
Automation Essentials BSM Monitoring events and alerts ITSM Help desk/ change requests BSA Essentials Community |
Automation Essentials BSM Monitoring events and alerts ITSM Help desk/ change requests BSA Essentials Community |
Automation Essentials BSM Monitoring events and alerts ITSM Help desk/ change requests BSA Essentials Community |

ITSM

Help desk/

change requestsEssentials BSM Monitoring events and alerts ITSM Help desk/ BSA Essentials Community | Subscription Services |Executive

BSA Essentials

Community | Subscription Services |Executive Dashboard and Reporting

Release Control

CMDB

|Executive Dashboard and Reporting Release Control CMDB HP Operations Orchestration (OO) software helps automate

HP Operations Orchestration (OO) software helps automate standard IT tasks and integrates critical management systems in the data center. By executing and reporting on out of-the-box workflows, HP Operations Orchestration enables standardization of IT processes, increase in quality, reduction of errors and cost savings.

HP Operations Orchestration helps in the process by creating workflows that tie together diverse tools, processes and domains so the risk of failure is significantly reduced. In addition, any risks arising from the unavailability of key personnel or groups are reduced due to the process information being captured in the workflows.

HP Server Automation software automates across the complete operational lifecycle of servers and software; including bare-metal or Virtual Machine provisioning, patch management, software deployment, configuration management, code deployment and rollback, and audit and compliance. It manages MicrosoftWindows, Solaris, Linux (Red Hat and SuSE), HP-UX, and AIX systems.

HP Network Automation software delivers network changes and configuration management through real-time visibility. This solution automates and controls network compliance