5 Tips for Avoiding a DR Disaster
Disaster recovery technologies and procedures are only as good as the DR plan they’re based on. Here are five quick tips for pulling together a solid plan.
By Martin Sinderman
Well, maybe not. Those technologies and procedures are only as effective as the DR plan they are based on, according to Jeff Byrne, senior analyst for Taneja Group Inc., a storage and storage-centric server technologies consultant. A solid plan is the foundation of any DR effort, according to Byrne, who offers the following tips for creating such a plan:
1. APPOINT a senior member of your team to lead the effort. “And if you don’t have the internal bandwidth or expertise,” Byrne says, “you may want to [partner with] an IT consultant who has experience in DR planning and execution.”
2. DECIDE what data, applications, systems, and other resources need to be protected and immediately recoverable in the case of a disaster.
3. CONDUCT a risk assessment, including a simple classification of threats (such as a fire or earthquake) and vulnerabilities (such as an unpatched operating system or unprotected network). Determine what type of response will be required to recover from each threat.
4. ESTABLISH recovery time and recovery point objectives (RTOs and RPOs) based on the risk assessment, as well as your client’s overall service-level requirements. Be prepared to modify them based on what the client is able and willing to spend: “DR strategies always involve a set of trade-offs,” notes Byrne, “between the value and recoverability of critical data and applications, and the resources required to plan, set up, test, and execute your DR processes.”
5. QUALIFY, procure, and/or license the necessary technology and product components. “Choose disaster recovery best practices that fit well with your client’s IT competencies and level of infrastructure.”
Byrne also advises keeping up on the latest developments in DR by engaging with other IT professionals, as well as participating in DR conferences and tutorials. “DR is an ongoing, highly visible, living and breathing process you’ll need to continue to invest in over the long haul, and there will always be room for improvement.”
One of the newer DR developments includes server virtualization technology, which significantly reduces the dependence of applications and data on hardware platforms, while greatly increasing their mobility. When combined with wide-area network (WAN) acceleration, these technologies open up new disaster recovery opportunities for small businesses.
A popular DR approach in VMware environments is to take a snapshot of a running virtual server—including its operating system, configuration settings, data, and applications. Then replicate that snapshot image to a remote site. Since incremental snapshots are now both space-efficient and nondisruptive, they can be taken as frequently as required to satisfy RTO and RPO metrics. For SMBs that cannot afford a single, designated disaster recovery site, you can set up data replication paths between different branch or remote sites, to at least cover the risk of a single-site failure.
Going one step beyond snapshot replication, virtualization, and data protection, vendors are now offering automated failover capabilities for virtual machines that will be supplemented in the near future with automated failback functionality. Though not too many SMBs have invested in these capabilities yet, they will likely become more mainstream (including easier to use and support) over the next couple of years.