Datacenter transformation and migration strategies

This blog is not meant to be a manual for data-center transformation. It serves to be a guide in assisting to make the best decisions for your environment when considering a data-center transformation (moving from physical to virtual and fabricating the road-map to hybrid cloud).

Each Enterprise is unique and therefore should be the Architect of its Hybrid IT Journey.

Why Choose Data-center Transformation?

• To Reduce Cost
• To Provide Infrastructure Control.
• To meet Regulatory Requirements & Governance
• To meet Future Needs of IT

Best Approach for Datacenter Transformation:

This 3 phased approach starts with Assessment, which will determine the strategy and road-map for design and migration. The illustration below depicts the suggested phased approach.

• The assessment phase should be focused towards understanding business priorities, identifying boundaries, and creating a migration plan that would work across the DCs. As an outcome of assessment, applications should be categorized on basis complexity and impact to the business. This would help create a migration plan. During the assessment one needs to capture data to understand virtualization readiness and risks mitigation plan to ensure minimal impact to end-users.
• Through the Design phase one should deliver a comprehensive logical architectural design for virtualization environment that addresses the project requirements. In this phase the desired environment must be designed and built and also tested through pilot migration.
• Once the design is validated the migration phase should began based on the overall migration plan.

Phase I: Assessment

As described below, a four step approach can be defined for carrying out the assessment.

Detailed Activities:

• Capture detailed information about Application and Infrastructure
• Validate the captured data with data provided
• Validate the asset-to-application bundle mapping
• Capture details of software installed
• Capture each physical component and map it to its application bundle
• Validate the interdependencies between the application bundles
• Capture the network information from the DC
• Final acceptance sign-off from Application owner and IT team on the details captured
• Determine criticality of applications
• Application Interdependencies
• Label each application to be migrated, moved, virtualised, upgrade or retire/redevelop with the app team
• Understand if any configuration changes need to be done or moved as-is
• Understand maximum downtime available for each application

Assumptions:

• Availability of up-to-date CMDB
• Share the list of for each application bundle
• Provide access to Workload environment
• Historical performance stats for applications are available
• Application team has data regarding the criticality, complexity and risk of the applications

Phase II: Design

• The design should ensure the quality and availability across the DC environment and manage costs more efficiently.
• Aim to accelerate business breakthroughs by operationalizing technology innovations that scale to deliver material business impact.
• Improve time-to-project completion, operational reliability and efficiency, and user satisfaction
• Create executable IT strategies and roadmaps that support customers’ business goals.
• Simplify and operate more reliable and efficient IT environments by assessing operational readiness, building for performance and high availability, and applying interoperability and enterprise management expertise.
• Reduce costs and maximize the return on your investments.

Detailed Activities:

• Creating Target Reference Architecture
• Classify the infrastructure to move groups based on Application Name, Line of business, Application category. Target RTO, Target RPO, Component type, Application manager, Point of contact etc.
• Create Application move based on Assessment report
• Consider high availability and disaster recovery scenarios
• Selection of the migration methods for various applications
• Creation of Test Plan and Fail-back plan

Assumptions:

• Customer IT and App team have access to all data needed to categorize applications
• Customer maintains physical and logical architecture documents of existing infrastructure
• Customer maintains network diagrams

Phase III: Migrate

• In this phase one must accelerate large scale virtualization deployments using best practices, methodologies & tools to achieve target ROI faster and lower costs.
• Deliver an in depth reviews of your virtualized environment to optimize your virtual infrastructure with concrete and actionable recommendations.
• Review the current virtualized environment’s design and provide a gap analysis against virtualization best practices.

Detailed Activities:

• Define Project management plan like Resource, communication, risk management, change, scope, schedule.
• Define Move Groups Schedule and Program Schedule in view of tasks, duration, resource availability, application regions, dependencies and constraints.
• Define Day-of-Move Plans and define "move plan" schedule and logistics with respect to application criticality, resource availability, infrastructure readiness.
• Plan and schedule governance meetings and review of movement schedule and also the discussion of risks and upcoming tasks.
• Define Test and acceptance criteria matrix, Plan and define test of core and shared infrastructure, Plan and define Application Provisioning test matrix and Plan for Program Level system integration test.
• Define data migration plan.

Assumptions:

• Customer follows standard project management practices which can be adapted for this program.
• All stakeholders including third party vendors will be involved in the planning exercise.
• Once the move plans are baselined, the schedule can be changed through only a Change Approval Board.
• Standard tools (MS Project, CA clarity) can be utilized to develop project schedules.

Conversion Type:

• Physical-to-Virtual (P2V) : Describes the process of decoupling and migrating a physical server's operating system, applications, and data from a physical server to a virtual machine guest hosted on a virtualized platform.
• Physical to Image (P2I) : Moving an operating system from a physical hardware platform to a file image.
• Virtual to Physical (V2P) : The absolutely reverse of P2V, Virtual to Physical is a conversion of a virtualized operating system to be directly installed on or migrated to a physical platform. The tool will need to update the operating system with the hardware drivers of the target hardware.
• Physical to Physical (P2P) : Moving an operating system from one hardware platform to another. The hardware may be the same type or completely different. The tool will need to update the operating system with the appropriate hardware drivers for the target machine and remove the unnecessary hardware drivers.
• Virtual to Image (V2I) : Moving an operating system to an image file stored on a system.
• Virtual to Virtual (V2V) : Moving an operating system from one virtual machine to another virtual machine. Generally, the virtualization host has tools for doing this natively and those should be considered if moving from the same vendor's virtualization host to another. If there is a need to move between different vendor versions of virtualization, the migration tool might act as an aid in this process where the native tools might not support such a move.
• Image to Physical (I2P) : Move of an operating system, stored in an image file, to run in physical hardware.
• Image to Virtual (I2V) : Moving an operating system, stored in an image file, to run in a virtual machine.

Tools Overview:

There are "N" number of tools available in market. Here is the top 5 majorly used migration tools. The latest version of these tools might be supporting few more additional features as well.

• PlateSpin Migrate:

1. It is a physical/virtual conversion tool that delivers the fastest and most efficient P2V (and in fact anywhere-to-anywhere) migrations, because of NetIQ's proprietary block-based transfer protocols and Server Sync technology.
2. It decouples workloads from their underlying server infrastructure and automatically copies and moves them over the network. The flexibility to move and re-balance workloads in any direction between physical and virtual hosts - physical-to-virtual, virtual-to-physical, virtual-to-virtual, physical-to-physical, in and out of imaging formats and so on - ensures optimal datacenter efficiency.
3. It Perform anywhere-to-anywhere migrations to and from physical servers, virtual hosts, and image archives, with broad support for 32- and 64-bit Windows and Linux operating systems. With PlateSpin Migrate you can standardize on a single product for all of your workload migration needs.
4. Task based wizards & Drag-and-Drop Interface helps to Drag and drop workloads across physical and virtual hosts. It also Resume the failed migrations/replications.
5. Reduce risk with automated planning, migration and live testing. Improve the speed and quality of datacenter initiatives.

• Cristie Clone Manager:

• It allows users to create exact copies of running machines between, and within, different physical, virtual and cloud environments. Users can optimize resource allocation, manage their server infrastructure, and utilize disaster recovery capabilities.
• It supports migration to and from physical machines, vSphere and HyperV hypervisors, vCloud Director, vRA, Microsoft Azure, SoftLayer and AWS.
• It provides a lightweight console which allows machines to be cloned by simply dragging and dropping a source machine (shown on the left hand side of the console) to a target machine or environment (on the right). There is an option to install an agent on source machines which is used to gather information about CPU, memory and disk usage so as to optimize the target machine configuration. However, a key feature of CloneManager is that it is not necessary to install an agent or any software on the source machine.
• It is used extensively by cloud and managed services providers who wish to on-board their customers’ machines from their existing DCs as well as by end users to provide complete mobility for their machines.
• It also allows customers to create a clone of a running system. CloneManager provides
  1. The ability to create a live clone of a running system.
  2. The cloned system may be targeted to physical dissimilar hardware.
  3. The cloned system may also be targeted at VMware vCenter, vSphere, vCloud, Hyper-V or Azure virtual machine environments (Hypervisor or cloud).
  4. A cloned target system maybe also used in warm standby mode. This is where a short synchronization process is run to bring a previously cloned system up-to-date prior to booting the clone into operational mode. This process is called Clone Sync.

• Double-Take Move:

• It Move is a comprehensive workload migration solution. It allows you to move an entire server, known as a source, by mirroring an image of that source to another server, known as the target.
• The source and target servers can be physical or virtual. The image of the source contains the server's system state (operating system, server configuration, and installation applications) and all of the source server’s data.
• You can also migrate just a source's data, in which case the target's system state (operating system, server configuration, and installed applications) will be used with the source's data.
• It Move uses patented data replication technology that allows users to continue accessing and changing data during the migration. As changes are made on the source, replication keeps the image of the source stored on the target up-to-date.
• It Move replicates, in real-time, only the file changes, not the entire file, allowing you to more efficiently use resources. When you are ready to cut over to the new server, Double-Take Move applies the source system state and after a reboot, the source is available and running on what was the target server hardware.
  1. Real-Time Data Movement.
  2. Migrate any combination of Physical or Virtual Servers (X2X).
  3. P2V and V2V auto provisioning (Hyper-V and VMware ESX).
  4. Manages the times around bandwidth use via an intelligent scheduling interface. Controls how much bandwidth migration operations can consume via Flexible Bandwidth Limiting and Intelligent Compression.
  5. Migrates physical or virtual servers that have a different number of drives, drive sizes, CPUs or memory than the original source.
  6. Migrate within DCs or across locations for DR datacenter migrations and consolidations.
  7. Entire server or just data migrations (SAN storage migrations).

• vRanger Pro:

• It offers comprehensive data protection that supports VMware vSphere. Delivering fast and reliable backup and recovery, vRanger Pro captures the entire image of a VM and offers restore at the image, file or application level (When using Quest Recovery Manager for Exchange).
• All backup jobs are executed without interrupting service--that is, while the source machine is running. An intelligent resource scheduler maximizes throughput by leveraging all available ESX hosts and controlling the number of simultaneous jobs, reducing the backup window. To save time and disk storage, vRanger Pro supports multiple space savings mechanisms such as Change Block Tracking (CBT), incremental backup, differential backup, and Active Block Mapping (ABM). Finally, vRanger Pro allows you to connect to multiple vCenter Servers to easily protect larger or distributed environments.
• vRanger Pro can be used in either a network based mode (using a Direct-To-Target configuration for ESX backups) or in a Fiber mode (LAN Free). The installation requirements differ for each mode of installation. Bear in mind that actual requirements vary depending upon a number of factors; the values below represent a good starting point for an optimal experience during your evaluation
• Features and Benefits:
  1. Performs entire image or partial differential backup
  2. Integrates with VCB to offload backups from the host
  3. Offers full VMware VirtualCenter integration
  4. Follows virtual machines when moved to new hosts
  5. Includes VSS driver to provide "transnationally consistent" backup images
  6. Compresses files so they can be sent over LANs/WANs to remote locations
  7. Contains P2V disaster recovery for image-level backups of physical machines
  8. Includes database repository to capture trending metrics
  9. Restores a virtual machine or files from within a virtual machine
  10. Offers improved compression for faster backups.
  11. Provides Startup wizard for even easier implementation.
• Primary Limitations:-
  1. It only protects the VMware virtual environment (with the recent announcement to protect Hyper-V). In a heterogeneous environment of physical and virtual servers, vRanger is inadequate.
  2. vRanger’s license is dependent on the number of processor cores being utilized, which tends to be expensive for growing business and for predictable cost estimation based upon growth.
  3. vRanger tends to have a lower ROI (Return on Investment) and higher TCO (Total Cost of Ownership) due to higher operational expenses of putting together, managing and monitoring servers, storage, networks, operating systems and data protection software.
  4. The non-integrated nature of vRanger leads to finger-pointing among the server, storage, networks, operating systems and data protection software vendors when a problem occurs.
• VMware vCenter Converter:

• VMware vCenter Converter lets users quickly, easily and affordably convert Microsoft Windows and Linux-based physical machines and third-party image formats to VMware virtual machines.
• It also converts virtual machines between VMware platforms. Automate and simplify physical to virtual machine conversions as well as conversions between virtual machine formats with VMware vCenter Converter. Use the intuitive wizard-driven interface of VMware vCenter Converter to transform your physical machines to virtual machines.
• VMware vCenter Converter can run on a wide variety of hardware, and supports most commonly used versions of the Microsoft Windows and Linux operating systems.
• With this robust enterprise class free migration tool we can:
  1. Convert Microsoft Windows and Linux-based physical machines and third-party image formats to VMware virtual machines.
  2. Complete multiple conversions simultaneously with a centralized management console. Multiple simultaneous conversions enable large-scale virtualization implementations.
  3. Minimize the number of steps to conversion with easy-to-use wizards.
  4. Hot cloning makes conversions non-disruptive - with no source server downtime or reboot. However the hot conversion may fail if the source server has massive live and concurrent IOPS.
  5. Sector-based copying enhances cloning and conversion speed
  6. Support for cold cloning (conversion that requires server downtime and reboot) in addition to hot cloning.
  7. Quickly and reliably convert local and remote physical machines into virtual machines without any disruption or downtime.
  8. Convert other virtual machine formats such as Microsoft Hyper-V, Microsoft Virtual PC and Microsoft Virtual Server or backup images of physical machines such as Symantec Backup Exec System Recovery or Norton Ghost to VMware virtual machines.

Tool Selection Pointers:

Choosing the appropriate tool for server migration will depend on several factors including -

1. Source and target environment Compatibility (hardware and software)
2. Cross Platform support
3. Availability
4. Performance
5. Requirement
6. Vendor Support
7. Licensing Cost

• Converting a Physical server into a Virtual server (P2V) or Virtual server into a new VM (V2V) is generally a straightforward process and most of the migration tools perform this process well.
• For typical P2V and V2V Migration, VMware converter gets the job done and is a great free tool when only a few systems need to be migrated. However, when working in an enterprise environment there are some fundamental limitations that make it less suitable for a migration task.
• A few key points are:
  1. Lack of discovery of candidates in centralized management console - Needs to be manually populated by the migration engineer.
  2. Lack of integrated testing functionality - No pre-test of the source to target for operation issues such as access issues, OS issue or Network issues.
  3. Lack of functionality for performing staged migrations - Can only go one direction in the migration (X2V).
  4. No real optimization to improve performance for latencylow bandwidth networks.
  5. VMware dedicated tool - Works only with VMware ESX and ESXi as the end point.
  6. Sometimes the hot conversion may fail if the source server has massive live and concurrent IOPS.

Methodology for Datacenter Transformation:

• This workflow is very much self explanatory.

Migration Strategy For Application:

• This workflow is very much self-explanatory.

Top Five Things to Watch Out For in a Data Center Migration:

• Hidden Complexity - You might not aware all of the back-end attachments to the primary applications you are going to be moving. There might be few legacy applications sitting in your current data center that are very much older. So start a detailed inventory with your business partners to track everything down and make sure you have an owner. All the information you discover needs to find a way into a CMDB, instead of maintaining in a spreadsheet on someone’s notebook.
• Post-Migration Testing - Capture the time performance end-to-end on a specific set of transactions on key applications. Document those tests, then repeat them after the migration. This process is a challenging.
• Application Delivery Optimization - If you use load balancers you will have to revsit their configurations and understand how you are going to manage the migrations. This may require some additional investment for duplicate hardware you weren’t expecting. Look for changes you can make in the current configurations to build better modularity. That requires a different plan for the migration to be worked out.
• Migration Breaks Regular Work Schedules - You must proactively inform your end users and support teams that some of them will be putting in overtime to do the QA needed to support the migration. We have to face it, you will be inhibited on; when you can move certain applications because of application owner freezes and critical process times. The migration scheduling alone will require months of planning. It is never too early to start, but expect overruns in overtime.
• What’s Hidden Isn’t Usually the Wealth - Somewhere in all those applications and back-end databases you will find some hardwired IP address or domain names. Not only should you start re-hunting them out at once.

What to Concentrate on During Datacenter Transformation:

• Hidden Complexity
• Post-Migration Testing of Workloads
• Effective Communication to Application Owners
• Application Delivery Optimization (if it use load-balancers)

Benefits of Datacenter Transformation:

• Improved Efficiency and Scalability
• Improved Business Response
• Capacity on Demand
• Reduced Business Challenges

How it can Work?

• Set strategic objectives and execute with measurable outcomes
• Ask innovative questions and provide actionable answers to stake holders
• Imagine new possibilities and deliver unrivalled insight
• Embrace challenges and tackle complexity with rigor

Reference Materials:

• VMware vCenter Converter Standalone Documentation
• PlateSpin Migrate
• Critie CloneManager
• vRanger Pro
• Double-Take Move

Definitely, there are many more elements that comprise the data center transformation, but if you focus on the above suggested things, you will make great steps to assuring success and pleased customers across the globe.

Do you think this approach can add few cents to your datacenter transformation strategy planning?