ABSTRACT The amount of risk mitigated during any data center migration is directly proportionate to the thoroughness with which the discovery of the environment is performed. In this whitepaper, the Alithea Consulting Group examines the key strategic points that must be integrated into any data center relocation initiative including, proper planning, well-defined logistics, careful target site preparation and flawless execution. Authored By: Kris Domich, Managing Principal Kris.Domich@alitheacg.com 2010 CORPORATE RIDGE DR, 7 TH FLOOR, MCLEAN, VA 22102 +1.703.434.3040 ALITHEACG.COM 2014 ALITHEA CONSULTING GROUP LLC
INTRODUCTION The average data center relocation is a monumental task and one of the most complex and challenging endeavors for any organization. With mission-critical services, applications, and data at risk, any failure in key processes can have detrimental repercussions. Yet keeping the business running while the migration is in full swing requires extensive planning, detailed logistics, well-choreographed activities and comprehensive program management. Relocating a data center impacts every aspect of the business and can have a long project lifecycle. It can be an opportunity to drive costs out of the environment, refine the data center strategy and build a foundation for long-term growth. To minimize the risks of unexpected downtime, interruption to revenue, and customer satisfaction losses, the relocation project must align business and IT objectives in a comprehensive data center relocation strategy. Key business factors driving the decision to go forward with data center relocation include: Adoption of a cloud-based consumption model Requirement to upgrade data center facilities due to current space, power and cooling limitations The need to identify more cost-effective data center facilities Improvement in redundancy, survivability, and business continuity measures Consolidations of physical data centers as a result of mergers and acquisitions Project Management Establish the Project Team It is paramount to develop a multi-disciplinary team that includes internal and external stakeholders from IT operations, facilities management, business unit, and 3 rd party vendors. Once this team is identified, a dedicated program manager should be assigned to coordinate all activities from planning, design, migration and post execution. Recruiting stakeholders at the inception of the project eases buy-in from all parties. Garter Research confirms that it is imperative to have a multifunctional project team assembled prior to initiating the data center relocation*. Leveraging the experience of these functional participants and the knowledge behind the complexity and interdependencies among IT, the business, and the facilities will ensure a successful data center move. Moving a data center requires buy-in and participation by all facets of an organization. Ultimately, the business must decide how much, if any downtime can be tolerated during the move. The technical procedures used to facilitate the move must satisfy those business requirements and leadership from the topdown must ensure that all participants are working toward a common goal. It is not uncommon for a data center relocation program to bring parts of the organization together that don t normally collaborate at such an intimate level. Gartner: A Checklist for Data Center Relocation, David J. Cappuccio, John R. Phelps, Sept 25, 2009 Page 2 of 8
Discovery Understanding what you have and how it all fits together Upon establishing a project team consisting of stakeholders, technical personnel, and 3rd party vendors, an exhaustive discovery of the environment is required. The migration team should collect all available information related to infrastructure, interfaces, and the data flows of each application. This can be performed through a variety of manual and automated interrogation methods. The end result must be a comprehensive understating of application and infrastructure relationships classified into manageable units referred to as bundles or affinity groups. In order to successfully create these relationships, the next step is to determine the interdependencies between the infrastructure components and applications. This is commonly achieved through interviews and workshops, ultimately gaining signoff before proceeding to the development of migration strategies and plans. By leveraging a combination of automated and manual collection approaches, risks associated with migration plans based on misinformation are minimized. Once the system/application inventory is mapped and validated, migration bundles should be developed based on a variety of criteria intended to minimize both the risk and cost of the moves. Recommended migration events and methods should be reviewed and approved by a migration steering committee or equivalent governance body prior to developing the master migration road map and plan. Page 3 of 8
Logistics - Relocation Approach Selecting the proper relocation approach requires an evaluation of business strategies, requirements, cost constraints and acceptable business risks. Once the business directions are identified, the relocation team can develop the best relocation methodology or employ a combination of approaches including: 1. Lift and Shift: This approach involves taking a successful backup of a system, powering it down, moving it and powering it back up. The lift and shift approach is perhaps the simplest but does carry the highest risk of the three methods described in this paper. This is due to the fact that other than the backup taken before power-down, only one instance of the systems is maintained and is subject to perils in transit. That said, this method is typically used for non-critical systems that can tolerate complete outages for brief periods of time. 2. Swing: Intended to move the "identity" of the server to temporary hardware without disrupting IT and business operations. This approach requires setting up temporary systems at the target site and replicating data to those systems in order to shift an application or service to the target site quickly and then powering down and relocating the equipment from the source site. The temporary equipment is retired once the service or application is again running on its original equipment. This method is commonly used when the time it takes to physically relocate a system exceeds the organization s tolerance for downtime of the application or service. Organizations that have dual or multi-data center environments often initiate a site failover of applications and services and move the original equipment to the new site upon successful startup of the applications and services in the secondary data center. The cost for a swing move can average higher than a lift and shift. This is because one must factor in the cost of the temporary equipment and the engineering time that is required to configure that equipment for use. 3. Logical Migration: Used for existing virtual machines or as an opportunity to migrate physical systems to virtual platforms. Instead of physically relocating assets, a logical move entails bundling applications and servers into logical move groups that would minimize risk and provide an optimized schedule. With the increased mainstream adoption of virtualization today representing a significant portion of many organizations, logical moves are much more common place than they were five or more years ago. This method has a moderate to high degree of complexity as far as planning but reduces the overall risk of most migrations. This is mainly due to the fact that servers are treated like files. They can be copied and tested while the original server is still up and running, which minimizes and nearly eliminates the need for any downtime. There are tools available to make this method easier and safer than in years past. 4. Hybrid Migration: A hybrid migration is most common and involves two or more of the three methods described. It is commonly used due to the fact that most organizations are comprised of both physical and logical infrastructure, and are not 100% virtualized. An exception would be a migration of cloud-based logical assets between cloud providers. Page 4 of 8
Target Site Preparation - Detailed From-To Mapping Prior to relocating data center assets, it is important to perform an inventory of all virtual and physical servers and determine how each will be handled (migrate, move, virtualize, upgrade, retire, etc.). Also, determining the interdependencies and generating dependency maps between applications, servers, storage, etc. is recommended. By completing a thorough asset inventory, IT stakeholders develop a clear understanding of the assets to move, their maintenance windows, downtime tolerances and interdependencies between applications, servers and storage, etc. Most relocation initiatives will require the data center to be moved in a series of events commonly referred to as waves. Each wave may contain one or more bundles which are logical/physical groups of applications and infrastructure that must move in unison in order to preserve the integrity of a service or dataset. In order to accomplish this without causing collateral damage during a given migration event (i.e. disrupting nonrelated services), it is critical to understand how the entire environment is sewn together. This will ensure that it can be systematically dismantled and reconstituted with no unplanned disruptions in service. During this asset bundling exercise, IT stakeholders must define the logical components of each service and application and map those components back to the physical devices. They then establish a logical sequence of events of how the assets will be moved. In doing so, the team addresses all environmental, logistical and readiness factors and devises a physical space plan to ensure an efficient and uninterrupted migration. It is also important to pay close attention to assets that have reached end-of-life (EOL), are being virtualized or have been decommissioned as you may not want to include them in the relocation. Target Site Preparation Pre-Deployment of Racks and Pre-Patched Cables Generally server racks are not designed to be moved fully populated although there are exceptions especially with disk arrays and large, integrated system frames. Outside of the exceptions, most racks are not designed for the horizontal stresses experienced when attempting to maneuver the rack over long distances or up/down ramps. It is recommended that the systems be removed from the racks prior to physically relocating them whenever possible for safety and to limit the risk of peril. It is important to keep in mind that a bundle may be composed of systems from multiple racks at the source data center. This means that there may not be a free rack available to move with the systems of a given bundle. For this and other reasons, most data center migrations will require some new racks to be pre-deployed at the target site in order to receive incoming systems from the source data center. Some organizations may elect to leave all racks behind at the source and deploy a uniform system of new racks at the target site. This offers many advantages that may outweigh the costs such as standardization, quick integration of environmental sensors or rack-level security, and a uniform aesthetic in the target site. When considering the requirements for the target site, be mindful of the opportunity to make improvements and correct any shortcomings that developed over time to sustain growth at the source site. One area in Page 5 of 8
particular is the cable plant (OSI Layer 1); a data center relocation gives an organization the option to hit the reset button on how they deploy and manage Layer 1. Companies must pay special attention to pre-patching each rack pre-terminated cables that are accurately labeled at both ends. Use a standardized color scheme for the various types of cables (i.e. PROD/NON-PROD, Secure/Non-Secure, KVM, Heartbeat, OOB Management, etc). Having each rack pre-patched will save a considerable amount of time and minimize troubleshooting headaches on move days when time is precious. In fact the most common reason for delays that we have seen during migration events is cabling that is absent or poorly installed. It is easy to underestimate the time required to safely remove cables and accurately account for the termination points. 4 out of 5 data centers we move have cables that are either not labeled or have labels that cannot be trusted. Execution Scheduling and Logistics When choosing a target date or range of dates for the relocation, determine the allowable downtime for business-critical systems, and prioritize the equipment and applications that must be up and running first. Scheduling requires tight coordination between business units and application owners and must be governed by the organization's tolerance for application/service downtime. Keep in mind that some platform vendors obligate you to use their services for powering down and reinitializing a system. This potential fee-for-service arrangement (recertification) is required so that there is no lapse in warranty and maintenance. Factor that cost into your plan when preparing your budget. During data center relocations, not only does the equipment need to be packaged properly, it must be handled, secured and shielded from electrostatic and electromagnetic damage. Clients often worry about who holds the liability for potential damage to their valuable IT assets. Any party that is facilitating the transport of data center equipment must provide full replacement value insurance for theft, damage, or loss. Plus, this insurance must be applied to each conveyance (each vehicle or vessel transporting equipment) as opposed to tying the insurance to the entire migration itself, which may be comprised of multiple events and conveyances. Data Center Relocations of the Future Research shows that 2014 and beyond will continue to see an increase in managed services and colocation as well as multi-tenant wholesale data centers. Many businesses looking to achieve their business and IT goals will elect to offload administrative responsibilities to colocation and/or managed service providers, enabling a sharpened focus on core business priorities and accelerating its time to revenue. The evolution of modular data centers (MDC) is expected to grow on the trajectory scale with a CAGR (Compounded Annual Growth Rate) 35%-45% through 2015, according to Frost and Sullivan*. As the ever increasing need for data processing continues to increase, so will costs associated with maintaining traditional data centers. Given this trend, IT executives will examine the option of MDCs and benefit from their flexibility, reduced build time and lower OPEX and CAPEX. * Modular data centres transcending traditional data centres, Published: 18 Jan 2013, By Gautham Gnanajothi Page 6 of 8
Cloud-to-cloud (C2C) migrations are also gaining in popularity as clients are seeking out more efficient, scalable and cost effective cloud providers. Having the ability to move easily (without first transferring data) between providers is an important consideration. More importantly, securing data being moved between cloud providers is a differentiator*. Clients are expecting that the cloud provider they are switching to is secure and compliant during the entire time that their data is being maintained by the provider. Alithea Consulting Group is well suited to support your cloud migration projects by leveraging our block-level any-toany data migration appliances. Our team of certified experts has collectively overseen more than a hundred C2C migrations representing several thousand server images and hundreds of petabytes of storage. Conclusion Data center relocation affects every IT asset servers, storage, applications, and the network that supports them. Comprehensive planning is critical to minimizing business disruption. The Alithea Consulting Group works hand-in-hand with its clients to define the most appropriate relocation approach given your timeline, budget and business requirements. Our approach is flexible enough to scale based on changing requirements, even when those changes occur mid-project. Our service pays particular attention to your data center uptime requirements to ensure an efficient relocation without any unplanned outages. Alithea Consulting Group can assist in helping select a project manager and multifunctional stakeholders for your relocation project. Obtaining sponsorship from representatives from multiple areas of the business prior to the move is critical to ensure that appropriate resources, unique disciplines and commitments are available throughout the project. * Tech Target: Secure cloud to cloud migration essential to poaching cloud customers, Gina Narcisi, Published: 20 Jun 2013 Page 7 of 8
Our experienced team will identify the proper methodology for your data center relocation, and detail the advantages and risks that require consideration for each approach. Together with our clients, we embrace best practice methods such as application rationalization, bundling and move groups and also factor disaster recovery into the relocation plan. As a matter of course, we recommend taking an inventory of all assets, reviewing the interactions between each component and considering equipment SLAs and warranties. This provides a baseline of the equipment you have on hand, gives an inventory list to compare to once the relocation is complete and ultimately protects your investment. Relocating a data center is a complex, sensitive job that requires special equipment, proven safe techniques, and highly-trained IT professionals for seamless execution. The Alithea Consulting Group s team of skilled IT professionals executes a data center migration with the appropriate level of rigor and control, using proven methods and expertise to ensure that the business can continue uninterrupted. We provide qualified and insured transportation, and professional rigging and packing to avoid damage during transportation. In addition to our vast expertise in data center relocation projects, Alithea Consulting Group draws on the extensive and practical experience of hundreds of cloud migrations and cloud platform deployments to ensure a smooth and effective transition when moving enterprise IT environments to the cloud. Working hand-inglove with our clients, through exhaustive planning, logistics identification, flawless and timely execution, Alithea Consulting Group provides peace of mind when it comes to executing complex data center relocations. About the Author: Kris Domich is a globally recognized Data Center Thought Leader with more than 22 years of experience in data center design, relocation, management, and operations. Kris repertoire spans continents with significant experience in Europe, Middle East, and Asia, in addition to the US. Recognized globally for a strong business acumen and profound technical knowledge, Kris has served as a direct advisor to the executive leadership of numerous Fortune 100 Companies. Kris is a regular speaker at international trade shows and other speaking events, and regularly publishes articles on data center trends. Kris is committed to excellence and delivering unparalleled customer service to all clients of the Alithea Consulting Group. Page 8 of 8