V.C.1     Data Center Recovery Alternatives

Data center recovery alternatives vary for infrastructure, configuration, operational state, and data migration. Management should document the reasons (e.g., cost and service level) for choosing an alternative and why it is appropriate based on the entity’s risk profile and complexity. The level of intervention required to activate the alternate sites affects both the cost and duration to resume operations. Recovery alternatives may take several forms, such as fully redundant systems at alternate sites, cloud-based recovery solutions (either internally developed or outsourced), another data center, or a third-party service provider. Data center and alternate site development is complex, and management should consider constraints in the analysis and design process. The primary objectives are for data to be available and remotely accessible. Management should maintain appropriate controls, regardless of solution. Alternative recovery site examples may include:

  • Cold site: A backup facility that has the necessary electrical and physical components of a computer facility, but does not have the computer equipment in place. The facility is ready to receive computer equipment when personnel move from their main computing location to the backup facility. This facility is usually not considered as the primary recovery option within the financial services industry because of the significant time necessary to install and activate the infrastructure. Comprehensive testing cannot occur until the infrastructure is established.
  • Warm site: An environmentally conditioned work space that is partially equipped with information systems and telecommunications equipment to support relocated operations in the event of a significant disruption. The systems are not loaded with the software or data required to resume operations and typically require manual intervention for failover and system reboots to resume critical processes. Therefore, end users may experience some disruption.
  • Hot site: A fully operational off-site data center equipped with hardware and software used in the event of an information system disruption. Hot site development is complex, and management should consider constraints in the analysis and design process.
  • Mirrored data recovery sites: Two or more separate, active sites that back up one another with each site independently supporting critical business functions. These sites provide almost immediate resumption capacity and are seamless for end users. Physical distance and its related latency present limitations for data centers that use real-time, data mirroring backup technologies. Similar to a hot site, these sites contain all of the equipment and connectivity capabilities; however, they also have a duplicate copy of the data. This method of high availability is commonly referred to as “Active-Active.”
  • Mobile site: A site that possesses capabilities between what a warm and a cold site offer and has portable structures equipped with computing equipment available to customers or personnel. Completely activating a mobile site depends on how quickly it can be delivered and backups restored.
  • Colocation facility: A facility that provides space, power, infrastructure, environmental controls, and telecommunications capabilities for multiple non-related tenants. If management relies on a colocation facility to deliver resources, there is a risk that the capacity at the colocation service provider may not be able to support the entity’s operations during a regional or large-scale event.
  • Reciprocal agreement: An agreement that allows two entities to back up each other. While these agreements may be cost-effective, they are viable only if there is adequate excess capacity at the reciprocal financial institution and both operate on the same version and configuration of core software. Consideration should be given to security and privacy, as sensitive customer information could be exposed to the staff at the reciprocal financial institution. While these arrangements may be acceptable as a short-term solution, management should not rely on them as a long-term recovery solution.
  • Disaster recovery as a service (DRaaS): A cloud-computing solution for replicating and hosting infrastructure, applications, and data that provides failover and recovery services.

 

Previous Section
V.C Facilities and Infrastructure
Next Section
V.C.2 Branch Relocation