IBM Systems Lab Services

Storage data migration 101: The migration process

Share this post:

Every day, organizations of all kinds undergo storage data migration projects. Migrations can be challenging, and in IBM Systems Lab Services we get a lot of questions about the 101 of storage data migration, so in this blog series, I wanted to explain some basic concepts of storage data migration for you.

In my previous blog post, I described the difference between the two general types of storage data migration: online and offline storage data migration. Now let’s look at how different applications access storage and what that means for data migration, and then we’ll talk about the storage data migration process.

How is your application accessing storage?

It’s critical to understand how your application is accessing storage. Experienced consultants customize and optimize the data migration method for applications based on the way that they are accessing storage. The typical storage input/output (I/O) access types are as follows:

Block I/O: Block I/O means the application is accessing block volume(s) provided by the storage subsystem. The application can either access the raw device or file system created on the block volume. Block I/O is the most used access type in the data center, and online data migration is possible for applications using block I/O access.

File I/O: File I/O means the application is accessing the file system or share provided by a storage subsystem, using mostly the Network File System (NFS) or Common Internet File System (CIFS) protocol. Online data migration is usually not possible for applications using file I/O access, but the downtime could be narrowed to a few minutes if it is well planned.

Object I/O: Object I/O means the application is accessing object storage using the Swift or Amazon S3 protocol. Object storage is a hierarchy-free method of storing data, typically used in a cloud environment. An application that has massive amounts of unstructured data can benefit from object storage. Object storage migration usually requires application downtime.

Other private API: Very few applications use a private API to access storage data, such as an application that uses the EMC Content Addressable Storage (CAS) API to access data in an archive platform. A data migration involving an application using a private API would require very close collaboration with the application team, and it usually requires application downtime too.

Storage data migration process

A successful storage data migration includes not only the data migration phase itself but also the planning phase prior to data migration and the validation phase after data migration. Those three phases comprise the whole storage data migration process.

Planning phase: The planning phase is very important to a successful data migration. The following tasks should be done in the planning phase:

  • Understand your requirements and targets
  • Collect current storage environment information
  • Create a data migration plan
  • Finalize the data migration method for each application
  • Develop data migration procedures
  • Develop a risk assessment and rollback plan

Migration phase: During the migration phase, you will perform the data migration according to the migration plan and procedures. Usually, the higher the complexity of the client environment and the greater the quantity of data, the more time the migration phase will take.

Validation phase: Once the storage data migration completes, it’s critical to validate that the data on the new storage subsystem is an accurate copy of the data from the original storage subsystem. Usually this is a joint effort with application administrators.

Contact Lab Services for support

Hopefully this storage migration 101 has been a helpful introduction. Migrations can be tricky depending on the complexity of your environment, and there are experienced storage leaders available to help. IBM Systems Lab Services has a team of consultants with a wealth of experience doing data migrations, and we’ve helped many organizations worldwide to migrate data to new storage platforms. Contact us today if you’ve got questions about a storage data migration.

More IBM Systems Lab Services stories

Open treasures: Why we contribute to open source projects

IBM Systems Lab Services, Open source, Technical

Open source projects are having a transformational effect on the industry and driving innovation in the IT ecosystem. More organizations are leaning in: more

Four ways to get ready for Power Systems Virtual Servers in IBM Cloud

IBM Systems Lab Services, Power Systems, Technical

Are you investigating running your AIX or IBM i workloads in the cloud, but don’t know where to start? IBM Power Systems Virtual Servers in IBM Cloud provide significant value as an addition to your on-premises IBM Power Systems environment. more

Fast-track your adoption of the latest IBM Z technologies

IBM Systems Lab Services, Mainframes, Technical

Extensive encryption for data in flight and at rest. Managing data privacy across your hybrid cloud network. Centralized key management for data sets. Accessing security and privacy solutions like these is a top priority for many businesses: more