A mirrored site instantaneously recovers a company’s mission-critical applications from disaster. These sites reside in a geographically distinct data center, and maintain identical pairs of servers, networks, storage devices and the like. Selecting a geographically diverse N+1 data center along with multi-carrier route-optimized Performance IPTM ensures complete facility-level diversity to ensure the mirrored site is accessible if something goes wrong at the production site.
A hot site is similar to a mirrored site in many ways – all infrastructure components are prepped, running, and ready. System configurations and application data, however, are sometimes replicated asynchronously and at failover, applications will typically be restarted at the hot site. Hot sites can be deployed using Active-Active failover or Active Passive failover.
A warm site differs in terms of architecture because the failover isn’t automated via a site recovery manager or stretched cluster. This means recovery of individual applications and servers must be accomplished manually.
Automated failover programs perform some of the following tasks:
- Synchronizes final data changes between sites
- Gracefully shuts down protected VMs or physical devices at failed site
- Completes power-up of devices in accordance with the DR “Run Book”
- Suspends data replication scripts
- Automates re-protection and failback
A cold site is appropriate for systems and applications that can be recovered in greater than 6 days without significant impact to the business. Backups of system configurations, data, and applications are periodically shipped to an offsite data vault. When a disaster occurs, equipment is typically procured and deployed to the cold site and then provisioned with backups from the offsite data vault.
Learn more about Disaster Recovery planning for IT organizations in our video series: