The adoption of cloud infrastructure is accelerating; businesses are increasingly taking their IT infrastructure (data centers) to cloud, private and public. When properly implemented substantial cost savings, infrastructure rationalization, accelerated application deployment, rapid IT resources provisioning and IT operations agility are demonstrated.
However, contrary to what some of the greatest consulting brands pretend, the approaches to implement it remain ad hoc, incomplete, and sometimes misleading to client organizations; their primary objective is to deploy vendors’ cloud solutions (they call it cloud implementation) and train users to these cloud solutions (they call it change management). Let me say it straight, this simplistic approach to cloud infrastructure implementation does not allow businesses to enjoy the promised benefits to the fullest.
The fact of the matter is, cloud computing is neither a technology nor a software, it is primarily a service delivery model and to enjoy its benefits, cloud infrastructure must be implemented as an IaaS Service Delivery Model. The interesting conversation I had couple of weeks ago with a Top VMware IT Transformation Executive confirms that fact, there is no doubt that, the ITaaS strategy she has developed will provide VMware in the near future with a truly durable competitive advantage (Click the link to know more about VMware IT Transformation approach). It is honest, difficult to challenge and on top of that it guarantees tangible benefits.
Distancing itself from the massive marketing hypes surrounding cloud, the idea behind that cloud infrastructure implementation approach is simple: the benefits of cloud infrastructure result from the implementation of IaaS Service Delivery Models and NOT from the discrete deployment of cloud solutions and infrastructures.
Written for CIOs, CTOs, seasoned IT professionals willing to embrace a Cloud career and those looking for greater insights into cloud infrastructure, this article discusses traditional IT infrastructures e.g., data center transformation into high return Infrastructure-as-a-Service (IaaS) Service Deliver Models through three key recommendations of the integrated Cloud Computing Transformation Methodology (iCCTM). iCCTM is a DevOps approach to cloud transformation I l largely developed in my last book Cloud Computing: Advanced Business and IT Strategies. These key recommendations are:
- IT Executive Vision
- IaaS Service Delivery Model
- Organizational Changes Deployment
Recommendation #1 – Capture Business and Operational Concerns and Expectations and Derive from Them the IT Leadership Vision about Your IT Operations Operating Model and Virtual Infrastructure
Ignoring or overlooking the business drivers underlying IT infrastructure migrations to cloud is definitely one of the practices to avoid; in ninety-nine of cases they not only result in none of the expected benefits but also in useless cloud capabilities. In reality, the bottom line with understanding business drivers is to capture executive concerns and expectations examples reduction of IT expenses, acceleration of applications delivery and IT operations, the services and virtual resources likely to enable achievement of the objectives and translate all these elements into a high level IaaS Service Delivery model that reflects the executive vision.
Plan a series of workshops involving executive representatives from the business, application development, and IT operations around the following topics :
- Clarification of business and operational concerns
- Definition of primary and support services catalog
- Definition of the high level picture of the target IaaS service delivery model
To make these workshops as productive as possible use collaborative tools; they facilitate the definition of cloud strategy. The first one of these tools is the Cloud Strategy Map,the figure below (Source: Cloud Computing: Advanced Business and IT Strategies) illustrates a simplified version of the cloud strategy map as defined by the iCCTM methodology:
The Cloud Strategy Map offers a questionnaire that helps to clarify objectives, expected services as well as the virtual resources required to achieve objectives. It helps to get in a record time a complete picture of the executive vision. The strategy map in the above picture helped to collaborativeley articulate one of my client’s cloud strategy as follows :
“[Executive Concerns and Expectations] In order to fix the financial issues, 15% of CAPEX and OPEX each year since 2011, migrating our IT infrastructure to cloud will definitely help.
[Value Propositions and Objectives] Providing IT operations personel with four cloud services i.e., Application Deployment, Virtual Machine Self-Provisioning, Resources Auto-Scaling, and Automation will definitely help to substantially reduce infrastructure costs.
[Target Workload and and Expected Virtual Capacities] A cloud infrastructure supporting up to 5 times today’s workload (2,500 virtual machines) will do.
[Expected Organizational Capability] An agile and lean IT operations organization will help us not only accelerate our workflows but also improve our overall Time-to-Market. A DevOps pilot experiment will be set up for three months to identify the impacts, the gains, losses and risks versus IT operations services are delivered today.”
The second tool is the IaaS Service Delivery Model, it offers a never-before-seen perspective of cloud computing which sees it from three dimensions including the Operating Model, the Data Center, and the Cloud Platform within Virtual Computing Capacity (VCC). It is illustrated in exhibit below:
The core purpose of the IaaS Delivery Model is to offer a high level and comprehensive picture of the future cloud capability that covers not only infrastructure issues but also organizational and operational matters. It is intended to support all subsequent cloud architecture works ranging from assessment and design to implementation, testing and deployment.
IT Operations Operating Model Defined
The IT operations operating model expresses the idea that the cloud infrastructure by itself will never deliver expected benefits; it represents your IT operations department in the cloud setting including the users involved, the IaaS services they use to accelerate applications delivery, and the work flows between them.
In the figure above, the operating model details one of my client’s vision about the expected agile and lean organization, it says:
“[The IT Operations Organizational Structure] DevOps will help to adjust the operating model with the cloud service delivery policy.
[The IT Operations Players] AgreDevOps will help to put together five categories of users of our cloud services including the customer, the business owner, application developer, quality assurance, and IT operations and production folks.
[The Selected Best Practices and Associated Processes] The IT operations activities overall will be structured around four main processes i.e., Plan and Measure, Develop and Test, Release and Deploy, and Monitor and Optimize.
The Virtual Computing Capacity (VCC) Explained
The VCC is a high level picture of your virtual infrastructure architecture; it highlights the virtual resources required to meet your businesses objectives including computing resources (virtual machines and associated CPU and memory), storage expressed in terms of capacity, IOPS, and storage bandwidth, and Network resources expressed in terms of total network bandwidth.
What the the VCC in figure about says about my client’s vision is, “The virtual infrastructure will be primarily composed of composed of set of virtual machines (VM) built upon a resource pool built upon Data Center Fabric specifically designed to meet jjthe cloud’s architecture requirements and support the target workload of 3,500 VMs.”
My message at this point is, capturing executive concerns and expectations, taking advantage of collaborative tools such the Strategy Map and Service Delivery Model to translate them into a comprehensive executive vision is vital for building a cloud infrastructure that concretely helps to save cost durably, make your IT work flows flexible, and accelerate your IT operations processes.
Recommendation #2 – Architect Your IaaS Delivery Model Not Only from the VCC Perspective But from The VCC and IT Operations Operating Model Standpoints
Capturing the executive concerns and expectations and translating them into a high level IaaS Delivery Model is the first step of the cloud architecting effort; the high level IaaS Service Delivery Model has to be detailed to become the actionable deliverable that will support the IT Operations Operating Model and the Cloud Platform.
Contrary to what one can read and hear in several forums and webinars, architecting your cloud platform should not be limited to selecting the best cloud software, installing it, configuring it, and training IT operations users to using it. This incomplete perspective, sometimes backed by some of the biggest consulting brand fails to consider that cloud computing is primarily a three-dimension IT service delivery model including IT Operating Model, the Cloud Platform and in the case of private and hybrid cloud options the Data Center dimension.
The fact of the matter is, architecting the cloud isn’t only a matter of installing and configuring your cloud software, it is also about setting up operating models whereby IT operations users effectively take advantage of well-thought IaaS services to achieve expectations and more importantly, it is about making sure your data center is properly sized and designed to support the performance requirements of the cloud. Let’s address these issues one by one.
Adjusting the Data Center Architecture
If you have opted for a private or hybrid cloud option chances are you need to transform your data center because today’s fragmented, siloed and oversubscribed data center architectures are incompatible with the performance requirements of the cloud; the side effects of virtualization create network bottlenecks, they add traffic load density and increase networking links utilization to ultimately affect the overall data center network bandwidth. You will not have the choice but to take advantage of some of the Unified Data Center Fabrics e.g., Cisco Unified Computing System (UCS), Brocade Data Center Fabric Reference Architecture and Juniper Cloud-Ready Data Center Reference Architecture to simplify your data center architecture.
As you can imagine, transforming a data center is not easy; at its simplest this process takes:
- Estimating the target workload in terms of required number of Virtual Machines (VMs), CPU Cores per VM, amount of RAM per VM, amount disk storage per VM as well as storage and network bandwidths
- Identifying gaps and misconceptions versus projected workloads and versus the Unified Data Center Fabric requirements through a comprehensive assessment of current data center
- Clarifying simplification and consolidation areas from the computing, storage, networking, service, and management perspectives
- Designing the future Unified Data Center Architecture from the computing, storage, networking, services, and management perspectives.
At this point in the Data Center Transformation into IaaS Service Deliery Model process we definitely need a break. Part II of this article will be about:
- Leveraging the iCCTM Data Center Transformation Matrix to Estimate Cloud Target workload, Assessing Traditional Data Center against the Cloud Performance Requirements and more importantly against Unified Data Center Fabric Architecture requirements and Selecting the Industry’s top Data Center Fabrics
- Taking advantage of the DevOps as an Organizational Design Framework to implement IT Operations Operating Models in terms of Organizational Structure, Stakeholders, IT Operations / Production Services, Coordination and Decision-Making Mechanisms
- Designing a Cloud Platform architecture in Private, Hybrid and Public contexts
The information provided in this article are available and detailed in my recent book Cloud Computing: Advanced Business and IT Strategies. I honestly and sincerely recommend it; Parts III and IV of the book detail matters as key as:
- Taking advantage of iCCTM as a DevOps approach to Cloud Transformation projects
- IT Operating Models assessment, design, and deployment in a private, public, and hybrid cloud contexts
- Data Center assessment, design, and implementation for private and hybrid clouds
- Cloud platform architecture design, implementation and deployment in private and hybrid contexts
- Deployment of Organizational Changes
YOU KNOW WHAT? START 2015 WITH A BANG … BUY NOW CLOUD COMPUTING: ADVANCED BUSINESS AND IT STRATEGIES.
It is also available at Amazon, Barnes & Noble, iBookstore and in many other online channels.
Part III – The Integrated Cloud Computing Transition Methodology (iCCTM)
- Chapter 5 – The iCCTM Framework
- Chapter 6 – The Cloud Strategy Definition
- Chapter 7 – Definition of the IT Operating Model
- Chapter 8 – Defining the Service-Oriented Organization
- Chapter 9 – The Cloud-Enabled IT Capability Architecture Definition
- Chapter 10 – The Organizational Changes Deployment
Part IV – The Fundamental Case Study – Illustration of the iCCTM through a real-life case study
- Chapter 11 – The Apple Business Services (ABS) Case
- Chapter 12 – Defining ABS Cloud Strategy
- Chapter 13 – Restructuring ABS IT Operating Model
- Chapter 14 – Defining ABS Service-Oriented Organization
- Chapter 15 – Defining ABS Cloud Platform Architecture
- Chapter 16 – Deploying ABS Cloud and Organizational Changes
Merry Christmas and Happy New Year to All
Copyright © Philippe Abdoulaye 2014. All rights reserved.