Three major risks and countermeasures for data center migration

Data center migration is a complex undertaking that can be difficult to explain to corporate executives who are funding the migration activity, and the associated business operational risks need to be understood and managed. Below we examine several of the challenges and risks associated with managing a data center.

[[126085]]

Service availability

The primary purpose of a data center is to host and run applications that provide services to the business. Whenever you consider migrating from one data center to another, you must first consider the availability of the underlying services. These services include infrastructure applications such as Active Directory and customer-facing applications such as SAP.

When services are moved from one data center to another, you must develop a strategy that takes into account when specific services are moved and the dependencies of applications on each other. A common approach to ensuring service availability is to create migration groups and group interdependent applications into the same group.

For services that support most enterprise applications, such as Active Directory and DNS, a common approach is to extend these core services across data centers. The services remain in both data centers until the migration is complete.

Hardware Migration

There are two common strategies for migrating physical servers: one is called "lift and shift" and the other is called data replication. In the lift and shift strategy, the hardware is put on a mobile truck and then installed in the new data center. The system is backed up before moving to the new location, but this strategy carries some risks.

One of the biggest risks is that the physical server may be damaged during transportation; damage during transportation will render the backup useless. Another challenge is that the two data centers are too far apart, making this approach unrealistic and unable to guarantee service availability within an acceptable period of time.

The second strategy is to migrate data over a leased line. Leased lines bring two sub-swing hardware options. One option is to perform a physical-to-physical (P2P) migration. P2P migration requires the purchase of similar hardware so that the applications and hardware of the original data center can be migrated over while ensuring the shortest downtime.

Another hardware migration option is physical to virtual machine (P2V) conversion. P2V requires converting a physical machine to a virtual machine over a leased line. P2V aims to achieve two goals: The first goal is to migrate workloads from one data center to another while ensuring that hardware costs are minimized. The second goal is to achieve data center transformation by moving to a virtual platform. P2V migration is a popular option because many engineers are already accustomed to performing this conversion as part of previous data center projects.

Data Migration

Migrating application data from one location to another can be one of the most complex aspects of a data center migration project. A simple approach is to perform a tape- or disk-based backup and perform a restore; however, like a lift-and-shift migration, backup and restore provides limited capabilities for restoring service in a timely manner. Additionally, backup and restore is not the ideal method for data migration—it is more suitable for disaster recovery scenarios where data recovery options are limited.

The primary method of choice for most data migrations is to configure a leased line. If a dedicated connection is available between the two data centers, the migration team can take advantage of hardware- or software-based synchronization mechanisms to perform data migrations. In addition to being able to migrate data, this method can also be used to perform P2P migrations, P2V migrations, and virtual machine to virtual machine (V2V) migrations.

Many enterprises decide to have multiple connections between two data centers. The connection needs at least two lines: one connection to support the usual end-user and data center to data center traffic, in order to support applications such as Active Directory and application to application traffic; the second, usually faster connection is used to perform data synchronization. The two-way connection prevents the two completely different traffic from interfering or affecting each other.