Managing data migration: how to choose the right solution for your organization

The migration of business IT infrastructure to the cloud is accelerating day by day. A survey carried out by 451 research projects which, by 2020, 60 percent of all business workloads will be in the cloud. But migrating applications and data from an enterprise data center to a cloud platform presents a daunting challenge.

Migration may be mandatory due to a merger or acquisition, in which data from another organization needs to be migrated to a new or existing environment, or because your business segment has been sold, which requires you to migrate storage elsewhere. In many cases, migration is voluntary in the quest for better service delivery. Cloud services can reduce costs, increase flexibility, and in some cases ensure better service performance

With the increase in the amounts of data being generated at exponential rates, the question of how to migrate data from an organization’s data center to its new cloud home has become particularly pressing. Some published reports indicate that it would take around 120 days to migrate 100TB of data using a dedicated 100Mbps connection. Unless a data migration is well planned and executed well, it could significantly disrupt an organization’s workflows. As Rebecca Hennessy, Marketing Manager at Experian Data Quality, says: “Without a holistic approach to data migration, any planned improvements for innovation, performance and growth can be seriously delayed, or worse, completely derailed.

This is especially the case when trying to migrate a production environment to the cloud. In such cases, extreme care should be taken to ensure that operations are not disrupted by the migration process. And that means choosing the right migration strategy for the unique circumstances of your business.

There are three main approaches to data migration: big bang, phased, and parallel. When planning a migration project, the first step is to determine which of these approaches will provide the best opportunity for success. Let’s take a look at each.

The big bang approach

The inevitable frustrations with any migration are intense, but you eliminate them within a contained period of time (hypothetically speaking). Costs tend to be lower, and you avoid dealing with middle-of-the-road solutions or using two operating systems at the same time.

This approach often means that the migration is done over a single weekend. When users log in at the start of the next week, they log into the new system, and the old one is completely offline. This avoids having to run the old and the new system simultaneously. Since no production operations take place in the interval between shutdown of the existing system and commissioning of the target system, the need to synchronize the two systems is eliminated.

However, due to this interval during which old and new systems are necessarily offline, the big bang approach is only suitable for businesses that do not need their systems to be online 24/7. 7. Additionally, since there is a specific and limited window of time for the change to be made, any issues that arise during the migration process could have a serious impact on the business operations if this window is passed.

The bad news about big bang migration is that it is high risk. Because of these exposures, the big bang approach is considered relatively high risk and works best when the amount and complexity of data to be migrated is low. Failure can lead to long downtime and intense frustration as users are locked into a new system all at once. Big bang migrations can work well if businesses are moving only small amounts of data, or if they are migrating data from just a few offices. A big transition done in the big bang style can lead to major disruptions, which most organizations cannot afford.

The parallel (or parallel) approach

The new system is installed next to the old one and the two work in tandem during the transition. Updates are released on both systems until the migration is complete. Once the correct operation of the new system has been validated, the old one is switched off. Parallel migration mitigates the risk somewhat since the old environment remains functional while the parallel environment is established.

The advantage of this approach is that current production is not interrupted and migration issues can be fully addressed before the target system takes over. It’s the the least risky of the three strategies because, if there is a problem with the new system, you can revert to the old system.

Establishing and managing two complete environments at the same time becomes expensive, both in terms of infrastructure usage and staff costs, so migrating everything in this way very quickly becomes prohibitive.

The progressive approach

Also called iterative migration, this approach means that data is migrated in small increments over time, by module, by volume, or by subsystem. As each increment is transferred to the target system, bugs can be fixed and any required user recycling done in small chunks, rather than having to be done system-wide all at once. The result is less risky than with a big bang migration, but with a much longer changeover time. Due to the longer time required to complete the migration, the costs may be higher.

You can migrate one or more desktops at a time, or you can start with applications that have little or no interdependencies. Gradual migration gives users a chance to get used to new ways of doing things. It can also be more complex to manage when you link the old and new storage together to keep apps functional during the transition.

One of the best ways to ensure a smooth migration with the phased approach is to have remote mirroring by volume, which is also a great disaster recovery and backup solution. It creates a natural and transparent gradual migration. Because replication is snapshot-based, only changed data is replicated and only the most recent change is synchronized, which saves bandwidth and ensures minimal impact on service performance. When all the volumes are migrated to the cloud, you can start running your new workloads there.

Good planning is paramount with a gradual migration, as dependencies between modules must be carefully mapped in advance so that modules do not become “orphan” in legacy or target systems. In fact, gradual migration is the option most frequently used today. Dylan Jones, editor-in-chief of Data Migration Pro, notes that their recent data migration research study indicates that 62% of migration projects use the phased approach.

The important thing to understand are all the pros and cons of each approach, and at the same time, realize that you need to choose the one that best suits the specific needs of your organization. The truth is, most migrations involve a mix of big bang, parallel, and gradual, but the right balance depends on your budget, your schedule, and your risk threshold.

Kerry Telling, Sales Manager, Northern Europe at Zadara storage

Image Credit: Alexskopje / Shutterstock

Sean N. Ayres