Backing up is hard to do...

01/30/2009 10:27

Have you ever experienced the nauseating sensation of data loss? If so, you are most likely a firm believer in backing up your information. Whether it was a deleted proposal that could have save you hours of work, or a presentation that someone wrote over, or a hard drive that died, there is no teacher quite like experience.

As a result, we tirelessly change tapes/drives and diligently rotate them offsite in case of a disaster at the office. But what exactly does our information backup accomplish? A good information backup strategy should be one component of an overall business continuity plan. But is all of this tedium still necessary, or is there an easier way?

We know that information backup is vital to our business’s continued existence, but let’s face it: it’s a chore. Every morning, the offsite tape/drive needs to be brought into the office, then the tape/drive has to be changed out, and every night the proper tape/drive has to be taken offsite and swapped with the previous one.

Let’s take a look at what backing up accomplishes. Generally, the information backup has 2 primary goals: disaster recovery and data restoration.

Disaster Recovery

The first goal of information backups, disaster recovery, is basically having the latest information copied to an alternate location. Then if a drive fails, the server crashes, or your office is flooded, you can continue to do business after this information (and the required infrastructure) has been restored.
Ironically, the most important component of disaster recovery backups is the most often overlooked: RECOVERY. The process of recovering from a disaster is a usually more time sensitive and stressful, but is the most likely to be ignored until it is needed: usually in an emergency. That being said, your information backup should provide an acceptable recovery time estimate, and this process should be regularly tested. It’s not just the technology recovery components that need to be regularly verified (such as the tapes or drives being written to) but the recovery procedures need to be practiced as well.

Data Restoration

The second goal of information backups, data restoration, involves restoring individual files and items (such as emails or database records) instead of recovering the entire system. These files or items may not necessarily be from the latest backup; but could be days, weeks, or even months old. Data restoration requires the ability to turn back to the clock and restore older or previous versions of files that may have been changed, deleted, overwritten or corrupted.

Traditionally, data restoration requires a significant investment in tapes, drives, or other media to store old versions of the information backups. A typical recommended backup media rotation includes daily media for at least 3 weeks, plus a monthly media for each month end, and a yearly media for permanent archival. Assuming the information backups are only occurring on the weekdays, that means you need to purchase 27 tapes/drives, plus 1 for each year and spares in case a tape or a drive breaks.

New solutions

While the overarching goals of information backup remain the same, technologies are maturing that can meet these goals in a more timely, efficient, and cost effective manner. Managed backup solutions have emerged that coordinate technologies such as hard drive imaging, server virtualization, data encryption and offsite data synchronization.

This type of solution images the server(s) hard drives at a regular interval, providing both disaster recovery and data restoration. These images can be used to fully restore a server, sometimes to new or different hardware, and can also be used to restore individual files or emails.
When imaging is used in conjunction with virtualization, the recovery time is significantly reduced. Virtualization allows a computer to operate within another computer: so Server1 and Server2 can both run on Server3. Server3 provides the hardware which Server1 and Server2 share. When this virtualization technology is combined with imaging, it allows a virtual version of the original server to be restored and running within hours instead of days, or can even be configured to provide failover recovery.

Encryption comes into play as we need to secure the data when it is taken offsite. The offsite component can be accomplished through a traditional rotation of media, or by uploading the data over the internet to a data backup service, or some combination of the two. Of course, encryption of backed up data should be always be done, regardless of the backup method.

A managed backup solution can seamlessly provide the security, reliability and productivity, and above all, peace of mind, that legacy backup solutions have historically failed to provide.