Five Disaster Recovery Tips for Businesses
New tips that can help businesses protect their data and recover quickly from disasters.
I was reading a report about how a mid-sized firm successfully slashed the backup window for restoring its SQL database from 17 hours to just two. Titled "How One SMB Slashed Disaster Recovery Time," it was an interesting piece about how wholesale distributor Hit Promotional Products dealt with the massive and increasing load of data that the 10-person, developer-centric IT team had to work with.
I thought that some of the insights shared by CIO Eric Shonebarger are relevant for small-sized businesses as well, especially where it pertains to disaster recovery.
Virtualization for Easy Disaster Recovery
Though it wasn't mentioned, the virtualization of physical servers into virtual machine packages undoubtedly contributed to an easier time in terms of system backups and disaster recovery. For unlike physical servers, digital copies of virtual machines can be easily made and restored - given adequate storage resources.
Time Taken for Disaster Recovery Matters
The strategy of retrofitting existing technology and processes into a virtualized environment is not a perfect one though, and can lead to its own set of problems. By treating virtual machines like physical machines, the company soon realized that managing a large virtualized infrastructure may require new tools given the likelihood of entire virtual machines being erased by mistake.
Specifically, Shonebarger came to the realization that a complete restore of the company's infrastructure from backups would take days - hardly tenable in an era of cloud computing and always-on connectivity. After some exploration, the company eventually settled on Veeam as its backup and recovery tool of choice. While the road taken by Hit Promotional Products may or may not be suited for a particular SMB, the takeaway here is simple enough: Disaster recovery takes time, and businesses should take into consideration the amount of downtime that their business can tolerate. This can be used to establish realistic recovery time objectives (RTO), which I wrote about in "SMBs Should Approach Disaster Recovery Differently."
"Don't confuse high availability with backup and recovery," Shonebarger said, underscoring the importance of data backups as opposed to focusing only on high-availability systems. On this front, Shonebarger advocates recovery testing, which is when a company simulates a data recovery by performing a recovery according to documented procedures.
Admittedly, the amount of resources and time required for a small business to perform recovery testing can be significant. It is important, however, that SMBs understand how this can help catch loopholes or flawed procedures in the advent of a total IT failure. After all, what good is it to establish a backup regime when not all data is recoverable in a disaster?