Today’s business has a lot of storage and data options. And, requirements around data control are going to continue to grow and evolve. With that in mind – let’s touch on one aspect of the IT and data center administrative process that some organizations hate to discuss: data migrations.
What if you need to move a massive amount of data? What if it’s not as simple as just re-mapping a storage repository? In some cases, you might be migrating entire storage vendors to align with specific business strategies. Either way – when dealing with critical corporate data – you need to have a plan. So, here are 8 steps to creating an enterprise data migration plan:
Business Impact Analysis. To best identify the business and operational requirements for a data migration project, a BIA should be conducted. The BIA process will involve different business stakeholders who will work to ensure that their requirements are factored into the migration plan.
During this process, four elements will be identified:
- IT team will examine and define available network bandwidth, storage, CPU needs, allowable downtime, and the migration schedule.
- As more teams become involved in the process, database and system administrators define application and database requirements.
- Key business partners will work to define the importance of, and requirements for, specific applications and types of data.
- Internal security and compliance groups define compliance requirements for a given infrastructure.
This document isn’t completed only for a data migration plan. The BIA is a critical plan which helps with disaster recovery planning and infrastructure management.
Data Discovery and Requirement Planning. The discovery step helps define the migration hardware and software environment; as well as overall requirements for the migration. Leave no feature unchecked in this process and ensure that you conduct a deep discovery of your current systems. This means understanding dependencies, permissions, user groups, how data is organized, network configurations, and more.
Data Mapping and Storage Design. Once you wrap up the data discovery phase, you’ll use this data to create your mapping and design architecture for the new ecosystem. In some cases, this’ll be easy when working with the same vendor. However, for heterogeneous migrations or working with multiple vendors – this process is especially critical. Although there are migration tools which can help, there needs to be a manual verification process to ensure proper data, storage, and configuration mapping and design.
Creating the Data Migration Plan. In many situations, you’ll be able to stand up your secondary storage or data control ecosystem in parallel to your existing environment. This allows for a seamless migration. Nonetheless, you still need to have a complete migration plan. You need to understand how this can impact users, your business, applications, and other critical workloads. In creating your plan – take these four points into consideration:
- Constraints must be identified through business and operational requirements. Make sure to take your business operations into consideration when doing a data migration project.
- Identify all of data to be migrated and all of the necessary associated attributes. When creating the migration plan, use your mapping and design metrics to ensure that all data attributes are accounted for and configured.
- Use vendor-specific migration tools if needed. You don’t have to do this alone. Aside from being able to work with a data migration or storage partner – you can leverage vendor-specific data migration tools. These can help you plan out the design, requirements, and even migration process.
- Always incorporate environment storage and application best practices. Does the VM need to be shut down? Can the VM be migrated via hypervisor migration? Can you live-migrate storage repositories? When working with virtualization, next-gen storage appliances, and very critical workloads – ensuring best practices around apps and storage is absolutely necessary. Since every environment is unique in their requirements, creating a migration plan is often a challenge. Different types of data may require different migration tools and strategies. Furthermore, business and operational requirements (downtime for example) may require creative ways of migrating the data.
The key thing to remember is that the planning phase is truly a living document. At any time, variables in the environment can alter the course of the project, or execution can lead to unexpected results. All of this can impact the migration plan as originally documented.
Data and Storage Preparation and Provisioning. During provisioning, the destination storage environment is prepared for the data move. This is where you design your repositories, LUNs, volumes, specific policies, security rules, virtualization integration, and more. Remember, we’re working with a lot of different types of storage environments today. Understand differences between hybrid arrays, all-flash, and traditional spinning disk. Furthermore, know how to provision new features like caching and even dynamic provisioning/de-provisioning features. Also, when designing around a new storage system – ensure that you’re familiar with all of the new features in the environment.
Validation Testing. This is a critical step. You can do things like load-testing and ensuring that policies and configurations properly transferred over. This is your chance to create a “production” test environment to ensure your new storage solution can handle user and business requirements.
Migration and Cutover. At this point – your migration plan is ready to go, you’ve done all of your testing, and you finalized configurations. Now, you can begin the migration and cutover. Now, there are vendor tools you can use from a destination storage technology. However, there are several factors which play into choosing the best migration methodology. These factors may include:
- The type of source and destination storage systems
- Infrastructure network and storage topology (NAS or SAN)
- Physical server and equipment locations
- Mapping requirements for specific applications
- Application and database data usage
- Specific project, customer business, technical, and operational requirements
With effective data migration, the truth is that there’s no one right way to do it. It will all depend on trade-offs based the infrastructure of the environment, core business and operational requirements, and migration experience. This is why the first planning phase in any data migration plan is so important. It provides the best-in-class migration plan for the specific migration project at hand.
Once the data has been moved, all clients must be redirected to the destination devices.
Final Migration Testing and Validation. The final and certainly very critical step is to validate the post-migration environment. This final step will confirm that all business and technical expectations have been met. From a testing side - at a minimum - network access, file permissions, directory structure, and database/applications need to be validated in their new environment.
In planning your own data or storage migration process, make sure to work with storage engineers and architects who can help guide you throughout the process. Remember, you’re dealing with very critical data resources here. Spend the extra time planning effective migrations, make sure you conduct effective backups, and work to ensure the safe transfer of your information. In some cases, this means working directly with the vendor, in other cases a partner can help. Either way, take data and storage migrations seriously. Storage outages are never fun; good planning can help avoid the situation and keep an environment up and running.