IT professionals utilize Disaster Recovery to prevent or minimize the damage of infrastructure downtime or failure. A “disaster” in the IT field can be anything that knocks an infrastructure offline. It can be as dramatic as a fire or flood destroying a data center or as unspectacular as a power outage or hacking attempt disrupting a network. Disaster Recovery is a comprehensive response to such an event. An approximately equal amount of tasks of a Disaster Recovery effort must be performed before and after a disaster. In anticipation of infrastructure downtime or failure, a company must back up its data and assign Disaster Recovery tasks to certain internal or external IT personnel. In addition to data, some companies will rent entire infrastructures (processing, networks, storage) as backups. Backup infrastructures can either be hot sites (an entire, functional, ready-to-go infrastructure) or warm sites (a scaled-down ready-to-go infrastructure). Companies should also perform tests of their Disaster Recovery preparations. They need to confirm in advance the functionality of their backup storage and infrastructure. Once the disaster occurs, IT personnel need to assess the damage to the existing infrastructure and, to the extent that the situation requires, route users and data to the functional backup systems. The first step should be to resuscitate the still-functional original infrastructure or move all critical applications and data to the temporary infrastructure. Only after a stable IT arrangement has been established should the company attempt to recover less critical applications and data or rebuild the entirety of the original infrastructure.
The resources involved in and the intensity of the Disaster Recovery effort will be planned for (but not guaranteed) by the recovery point objective (RPO) and the recovery time objective (RTO). A company sets its RPO as the maximum amount of data (expressed in time) it can afford to lose in an infrastructure failure. With an 8-hour RPO, a company will back up all of its data every 8 hours. Implicitly the company then accepts the risk of losing a maximum of 8-hours’ worth of data. Due to bandwidth and storage costs, extremely low RPOs can be expensive. Companies have to balance with their budget their desire to retain the largest possible amount of data. With an RTO, on the other hand, a company decides the maximum tolerable amount of downtime for an infrastructure or other service. A 24-hour RTO means that a company has determined that if a knocked-out infrastructure has not been restored within 24 hours the company will experience severe (usually financial) consequences. A company lowers its RTO the more resources and preparation it invests in Disaster Recovery.
According to this article from Continuity Central, companies that want to handle their own Disaster Recovery efforts should avoid a few common pitfalls. First, they should not accept cookie-cutter solutions or copy the plans of other companies. To be effective, they need to customize their Disaster Recovery plan according to their specific infrastructure and IT requirements. Second, companies need to double-check that their most important files have been copied by their backup systems. The article also advises the companies be sure not to store the master copy of the Disaster Recovery plan on the infrastructure that it pertains to.
Companies can avoid the uncertainties and complexities of Disaster Recovery by selecting virtual desktop infrastructures and hosted applications from IronOrbit. IronOrbit promises an RPO of 4 hours and an RTO of 12 hours. As this article from InfoWorld points out, virtualized infrastructures can be backed-up, repaired, and replaced much faster and easier than non-virtualized infrastructures. Additional Disaster Recovery features offered by IronOrbit include weekly server snapshots, SAN storage arrays, and disaster-proof, geographically-separated data centers. IronOrbit managed and resolves IT disasters for our clients, allowing them to turn their focus from potential emergencies to revenue-driving, day-to-day work.