One of the monumentally challenging tasks that might be presented to IT personnel in any organization is the migration of datacenter resources from one datacenter to another. The ability of IT operations in any organization being able to plan the move, test the move of resources, and execute the move with little to no downtime is an impressive feat to pull off. However, it can be done with the right strategy, planning, testing, and execution. Datacenter migrations are becoming more common especially with organizations moving resources to the public cloud from privately owned datacenters. However, some may go the opposite direction as well. When thinking about migrating datacenter resources – what are the reasons for doing this? What are the challenges involved with successfully performing a datacenter migration? What steps might be involved for organizations looking at a datacenter migration?
Datacenter Migration Reasons
With the complexity and expense often involved, why would an organization move from one datacenter to another? Generally speaking, most commonly the determination is made from a business perspective. The following are reasons that are typically behind datacenter migrations:
- Business mergers or Business acquisition – Companies may be merged together or one company may buy another company. This may call for datacenters to be consolidated or moved
- Downsizing, rightsizing, or scaling out – Due to changes in business either growth or decline in business may require companies to change datacenter locations or migrate resources. Additionally, new datacenter locations may be added if scaling out
- Moving to the public cloud – Despite the tremendous advantages in many aspects of moving to the public cloud, there is no “magic button” that can be flipped to make it happen without proper planning and testing
Planning for Datacenter Migration
With all the complexity and technical considerations that need to be made when migrating between datacenters, organizations must have established goals and a plan of action. Several questions must be asked to established the goals of the migration:
- What problem is an organization trying to solve?
- Will the migration be a total migration of resources or only migrating a few resources?
- Will resources migrated include physical or logical or both?
- What are the technical and business objectives?
The above questions and many others are good starting points for organizations migrating between datacenters. However, many other things must be considered that may be unique to the current datacenter topology utilized by an organization and the destination datacenter targeted for migration.
Considerations for Migrating Datacenters to the Public Cloud
Perhaps one of the most common scenarios entertained by organizations today is migrating resources from on-premise datacenters over to the public cloud. Public cloud datacenters are powerful options for organizations today, however, they also present unique challenges, especially in the way organizations plan and design network topologies. All of the major public cloud providers have network configurations that are layer 3 only. This means organizations must think about legacy applications that may require layer 2 broadcast connectivity or have VLANs hard coded. The size of network deployments in public cloud environments also must be thought through carefully. Many cloud storage providers such as Amazon require picking the size of your network once you provision the VPC which can result in issues down the road for organizations who have not sized this correctly and didn’t account properly for growth.
Additionally, security may need to be reconsidered for public cloud resources. Organizations using more legacy or traditional approaches to security can be presented with real challenges in public cloud environments. Traditional firewalls, VLANs, and application security may not be sufficient or even apply in the public cloud environment. Organizations certainly want to take this into consideration.
Datacenter Migration – Planning Testing Executing and Post Analysis
In its simplest form, we can divide a datacenter migration into three phases – planning, testing, and executing. Organizations must give proper attention to each phase of the process to avoid running into issues along the way. Let’s take a look briefly at each phase and what they may entail.
The Planning phase of a datacenter migration is crucially important. In this phase, organizations define the scope, identify resources to be migrated, and develop a timeline among other things. This process could literally take months to accomplish as it can be a monumental undertaking to fully understand the above action items. It can involve the following:
- Site surveys – Especially important if migrating from a physical datacenter to another physical datacenter
- Documentation – Everything must be documented all the way up the stack – from layer1 cabling, through the networking details and up the application layer to fully understand everything that is required for migration
- Identifying dependencies – Often this slips by when supporting an application. What are the true dependencies? Until every nuance is understood, planning and documentation must continue
- Understanding Network needs – As mentioned earlier, networking challenges can be real issue especially in public cloud environments. Are these fully understood?
- Supporting applications – How do applications currently work? Can they work this same way during and after the migration? What is the expected behavior as resources are migrated?
Organizations must invest the needed time and resources into the planning phase to expect a successful datacenter migration.
During the Testing phase of the datacenter migration, organizations perform mock testing of migrating resources. Whether it be test virtual machines or other resources that can be staged to test the migration process, the process of the migration is tested during the testing phase. This ensures that any areas that may have been missed in gathering information regarding underlying requirements or system dependencies can be caught during this phase and not during the actual production migration.
During the testing phase also, the key members of teams who will be involved need to coordinate the events exactly as they will happen during the production cutover. Making use of the same personnel and resources helps organizations to work through communication challenges and possible lack thereof during the staged test migration. Testing the migration may include:
- Running through the actual migration playbook
- Coordinating with the various team members who will be involved during the production migration
- Documenting any issues along the way so that critical flaws in the process can be identified and remediated before the next test phase
- Multiple iterations of testing may need to be performed until the process is executed with the degree of success that is acceptable to the migration plan
Once the testing phase yields the results that are desirable for the actual migration, organization will execute the production datacenter migration. This is the moment that all the preparation, planning, and testing has lead up to – performing the actual migration. During this production cutover or migration and just prior to, organizations will want to do the following:
- Not deviate from the migration plan and the methods that were prepared during the testing phase
- Have good communication between all team members involved as well as contact information of all internal and even external resources who may be needed during the migration
- Notify all customers of any potential system outages or performance degradation during the migration. An informed customer is generally happier than one who is surprised by unexpected errors, performance issues, or system downtime
- Be ready with customer service as well as an escalation team ready to address any unexpected issues that were not discovered during the testing phase
Finally, during the post analysis phase, organizations perform a post mortem of all the events prior to the migration, during, and even after. During this phase, organizations will want to make sure of the following:
- Perform thorough testing of all applications and systems after the datacenter migration to uncover any unexpected system errors or even performance issues
- Have a team prepared for any residual customer facing issues
- Be sure and notify customers and others once the maintenance period has passed and when the system is performing as expected
- Perform a thorough post mortem with all team members involved to glean any information that can serve as learning points for any future system migrations
Thoughts
In today’s highly volatile world for businesses as well as technologies in use, organizations have to be prepared to move or migrate resources as needed. The new modern world of public cloud vendors allows organizations to be more agile and powerful than ever before. No longer do organizations have to even rely on physical datacenters of their own that have to be built, maintained, or upgraded. Those looking to perform a datacenter migration need to be sure to plan well. In fact, following the action plan to plan, test, execute, and perform post analysis can ensure the best chance for a successful datacenter migration.
Experience modern data protection with this latest Vembu BDR Suite v.3.8.0 FREE edition. Try the 30 days free trial here: https://www.bdrsuite.com/vembu-bdr-suite-download/
Follow our Twitter and Facebook feeds for new releases, updates, insightful posts and more.
