Best Practices for a Successful Data Center Migration
As businesses grow and change, they inevitably reach a point where their existing information technology systems are unable to keep up with demand.
A data center migration, or relocation, is the process of moving assets from one environment to another. The move is often to a contemporary hybrid IT model including owned or colocation data center space and public or private cloud. The physical relocation of hardware is combined with virtual, software-defined workload migration.
When to Plan for a Data Center Migration
Businesses often start thinking about data center migration when they are planning to launch a new application or online service. Other common reasons include gaining access to a wider range of connectivity options or expanding data center capacity due to business growth, acquisition, or merger.
No matter your reasons for data center migration, it’s vital that you plan ahead. After all, there are numerous potential problems and challenges with this highly involved process. Overlooked security vulnerabilities, compatibility issues, data loss, and unexpected downtime are just a few. That’s why you must avoid rushed, ad-hoc deployments and transfers.
Instead, data center migration should be carefully planned to minimize risk and enjoy a faster return on investment. Other benefits of a successful migration include reduced energy costs, improved performance, and enhanced security.
Five Phases of a Successful Data Center Migration
To help companies start enjoying the benefits sooner, Silverback Data Center Solutions groups data center migration steps into five phases:
- Discovery
- Analysis
- Planning
- Execution
- Post-migration
Using this structure helps identify all assets and minimize risk during the migration process. It also enables easy deployment and configuration and helps future-proof your information technology systems.
#1 Discovery: Understanding your existing IT environment
The data center migration process begins with a full assessment of your current IT environment. Identifying which assets need to be moved is a vital starting point. This will help you determine an appropriate budget and timeframe and clearly define your migration goals.
The discovery phase addresses three main questions:
- What exactly are we migrating?
- Physical IT compute, storage, and network devices
- Server racks, PDU’s, and other hardware components
- Applications and specific data assets
- What is the budget and timeline?
- Technical and business costs
- Prorated across one to five years
- What are your desired outcomes?
- Bottom line and productivity impact
- Department-level success metrics
How you answer these questions depends entirely on your current environment and what you wish to achieve. For example, many migrations take place in an entirely virtual environment. This is when there are no hardware assets to think about.
Once you have outlined your priorities, it is time to move on to the best practices for this phase:
Consult with your stakeholders
Since a data center migration will likely affect many operational areas of the business, it is vital to work closely with your stakeholders of these operational areas. Each stakeholder must understand how and which business functions will be impacted during the migration. You should assign a capable project manager. Additionally, you need to build a cross-functional team to engage stakeholders and keep them informed about the process. At this point, you may also want to hire an external consultant to provide expertise to and augment your team.
Submit migration questionnaires
Each data center migration is unique, but there are several key questions to include in a survey to facilitate the discovery process. Here are some of the most important:
- Which existing network dependencies, processes, and interrelated applications might be impacted by the migration?
- Will you require any compliance certificates or attestations?
- What type of support will you need during and after migration?
- What are your availability requirements, as defined in your SLAs?
- Which security measures will be required to protect your data during migration?
Finally, if you migration includes physical data center assets, you will also need to determine your telecommunication, cabling, physical space, power, and cooling requirements.
Build your asset inventory
Building your asset inventory is essential to ensure you can keep track of all assets during the migration process. Conduct a data center audit that includes all physical and virtual assets, including data center hardware, applications, and storage assets. Also, your inventory should clearly identify network connectivity, how each asset interacts across your wider network. This will allow to create a network dependency summary.
#2 Analysis: Drafting your data center migration strategies
Now that you have a complete picture of your existing environment, the next phase is to determine your technical requirements. This stage involves creating a draft migration strategy, identifying a suitable migration method, and evaluating your infrastructure needs.
Draft your migration strategy
The first step is to define your target operating model. There are two main options, including building your own, or renting either an entire facility, or shared space in a colocation o. Building your own data center offers by far the greatest degree of control, but it also comes at huge capital expense. Renting or leasing your data center, by contrast, reduces the upfront capital expenses to the cost of the actual migration. Everything thereafter becomes an operational expense. The colocation option is something of a blend of the two, since it involves sharing the facility and infrastructure with other companies.
Regardless of the option you choose, there will be significant costs involved in any data center relocation project. How much it will cost and how much time it will take will depend on the type of migration. For example, a purely cloud to cloud migration will not have any physical components to worry about, so it might be cheaper and faster. If, however, you are migrating a physical data center, you will also need to think about the many physical components involved including network, telecommunications, equipment cabinets, and other infrastructure.
Identify your migration method
Data center migration is a broad term that represents the process of moving IT systems and workloads from one infrastructure to another. Some migrations are entirely software-defined. Others involve moving server racks and other physical systems.
The easiest migrations tend to be virtual to virtual, which involve moving virtual machines and data assets. More complex are migrations from physical to virtual or virtual to physical, since these involve retiring existing or purchasing new physical assets and moving your workloads to either a private or public cloud (or both). In the case of cloud-to-cloud migrations, you should also be aware that your existing vendor might charge data egress fees for moving your data out of their systems.
Evaluate your infrastructure needs
The last part of the analysis phase is to evaluate your technical infrastructure needs. Naturally, this process will be more complex for physical data center relocations. In these cases, you will need to think about physical space and infrastructure requirements, such as cabling, cooling, and power. It is crucial to consider optimal rack layouts to facilitate better cooling and energy use. For a physical data center relocation, you will also need accurate connectivity diagrams.
Next, you will need to think about the computing and storage requirements for your workloads and applications. This includes operating systems, hypervisors, and other important software required to run a data center. Finally, remember that every data center migration takes time and will inevitably cause some disruption. This is why it is important to plan ahead for upgrades and changes and determine a maximum period of allowable downtime.
#3 Planning: Creating a master plan for data center migration
Now that you understand your existing environment and where you want to be with your new data center, it is time to create the comprehensive data center migration plan itself. A solid plan will greatly reduce the risk of unexpected downtime and the potential revenue losses.
Outline your project plans
An outline of your master plan must include the complete asset inventory you compiled earlier. Count every application, hardware device, and data asset. Your project plan should include a migration checklist and roadmap that prioritizes assets and their respective migrations appropriately.
When migrating apps and data, there are three basic steps to think about. First, you need to extract the data from its source system. Next, you need to prepare it for migration, which may involve transforming it between different formats and file systems. Third, you will need to load the data into the new environment. Naturally, moving data between arrays is much easier than migrating between entirely different environments, such as public to private clouds.
When migrating hardware, be sure that all components are properly organized, packaged and expertly transported to minimize the risk of damage. If not secured during transit, problems like static discharge could cause serious damage leading to unexpected downtime. Ensure a chain of custody process is strictly followed to avoid any equipment or data loss.
Prepare a statement of work
For suppliers involved in the migration, prepare a statement of work with clearly identified roles and responsibilities. The statement of work should include a list of the data center migration steps. This involves documenting your deliverables, defining a timeline and performance period, outlining your evaluation criteria, and adding any relevant terms and conditions. Your team members must also understand their roles and responsibilities. All parties should understand their responsibilities, dependencies with other parties, and be accountable to one another. with specific reporting requirements.
Finalize your master plan
Finalizing your master plan involves allocating resources, planning your contingencies, and obtaining any external help and other resources you might need.
Finally, you should also consider hiring external resources to help with part of all of the data center relocation. This is especially important if your current team lacks the skills, experience, or time required to carry out the process in a safe, secure, and efficient manner.
#4 Execution: The data center migration event
Your migration may be broken into separate events or phases. The event begins with communication between all stakeholders and the final pre-migration testing and preparation. To coordinate your schedule, start with a roll call with the migration team and any backup resources 48 hours ahead of the move. Do this roll again around 4 hours before. This will help you ensure full resource availability through the event. Your communications plan should give your stakeholders multiple points of reference for the schedule. You can use physical posters, online dashboards, and email reminders. It is essential that you have real-time communication between everyone involved throughout the move.
Test your master plan
Before putting your master plan into action, it is important to conduct practice or trial runs. You should schedule a full trial run prior to the migration. To help with the testing and validation, you can use virtualization tools to carry out performance tests, functional tests, and other types of tests based on the type of workload.
Execute your master plan
Your project plan must be followed precisely to prevent any steps from being missed out or being followed out of sequence. As soon as you execute your master plan mark each task as complete. Using your project plan checklists to sign off on each task offers an easy way to keep track of things. For this, you can use a work-management platform like Smartsheet, Trello or Asana, which are web-based.
Facilitate knowledge transfer
Depending on the complexity of your data center relocation, you may need to facilitate the transfer of extra knowledge. For example, architectural decisions may need to be revisited as new information is provided. Technical designs may need to be modified. More resources might be required than previously expected. It is vital that your team is ready for such eventualities, not least because any complex data center relocation is likely to involve a few unknowns.
#5 Post-migration: Wrapping up your data center migration
Once you have executed every task in your data center relocation plan, the first thing to do is verify your migration. This involves a comprehensive follow-up assessment of all aspects of the new data center, including hardware, applications, data, and network infrastructure.
Finalize your documentation
Finalizing your data center move involves putting together the necessary documentation, which you will use to determine the success of the project. All assets must be tracked in their new environment. To make sure that everything has been taken into consideration, you should create an up-to-date inventory.
If your migration involves physical infrastructure and hardware, be sure to map your new server room layouts. Include complete data center cabling for both networking and power connections, preferably in a CMDB or DCIM solution. Other steps to take include creating a rack elevation documentation and availability heatmaps. Finally, you should use auto-discovery and REST APIs to keep the documentation automatically updated.
Execute your decommissioning plan
If you are relocating a portion of your workloads from a physical environment to private or public cloud, you will also need to execute a data center asset decommissioning plan. This will ensure that any retired data-bearing assets are properly wiped of potentially sensitive data. Further, retired assets may have significant market value. Sustainable practices should in mind in all cases. Be sure to validate and document all your decommissioning preconditions and package all retired hardware correctly.
You may also want to engage an IT asset disposition (ITAD) company at this point to ensure your assets are properly retired. Consider ITAD companies who specialize in data center IT equipment. Servers and other hardware that are no longer viable need to be recycled or repurposed in accordance with both external regulations and your company’s CSR policies. Data-bearing devices that are no longer needed must be securely wiped, degaussed, or physically destroyed.
Schedule follow-up sessions
The last step in the post-migration phase is to finalize the documentation with new ownership. You will schedule follow-up sessions with the teams and stakeholders to address any outstanding issues. Then, close the project. During the sessions, you should also address device contracts and life cycle management. Discuss any additional staff training or onboarding that might be required.
Final Words About Successful Data Center Migration
You can use the best practices we have covered in this guide as a starting point for developing a customized action plan specific to your company. After all, whether you are migrating just a couple of applications or two-hundred server racks of IT equipment across the country, every migration is different. With the right plan and the right help, your data center migration can help set your business up for future success – all while keeping risk to a minimum.
Silverback Data Center Solutions is an industry leader in data center migration, deployment, staffing, auditing, and decommissioning. We understand that your digital infrastructure is mission-critical, and that is exactly how we treat it. Contact us today to discuss your data center migration needs.
