The surprising truths about testing the recover-ability of your data backups.
When it comes to backing up data, too many people are still going at it with the wrong mindset. No matter how good of a backup technology you rely on, you’re the one who’s ultimately responsible for the availability and recovery of your data.
Even a 100% secure and reliable backup process can fail if not implemented properly. For example:
- It’s common for personal and home computer users to forget or overlook backups of their Microsoft Outlook email files. These files are often stored in hard-to-find folders, making it easy to forget.
- Larger companies must deal with” virtual server sprawl” or frequent IT changes. It’s often the case that servers are moved, modified or added without notifying the backup administrator. This can lead to systems being left out of the backup process.
Because of these and other reasons, it’s a good idea to test backups for consistency at least once per month, and perform full data recovery drills once per year. (If you have an online backup service, they probably already monitor your backups for consistency for you… and you can obtain reports from them directly)
When it comes to full recovery drills, it isn’t simply enough to restore your files. There are actually many different scenarios in which you’d need to recover backup data , and each of these can vary in different ways. When you perform you data recovery drills, keep each of these scenarios in mind.
- Recovery Point: How much data can you afford to lose? If you perform monthly backups, you may lose up to 30 days of data. If this is unacceptable, you may want to consider using a real-time backup which can minimize the data loss window to just a few seconds.
- Recovery Time: For legal, customer service, PR and other reasons, a company may need to recover quickly after a server outage. (Some studies have estimated the downtime cost for a midsized company at nearly $6000 per minute!) If this is a priority for you, make sure you have a means to retrieve data quickly.
- Bare metal recovery: Sometimes, loading backup data is insufficient. You may need to re-build entire servers from the OS up. You need to think about hardware conflicts and other potential issues that could complicate a bare metal restore of your systems.
- Off-site recovery: If your primary location is destroyed, be ready to quickly set up and re-launch at an alternate location. This is why it’s a good idea to always maintain one off-site set of backups.
- Point-in-time recovery: The most common types of data loss are minor incidents related to human clumsiness. Maybe you accidentally deleted a file or overwrote an important file… then forgot about it for a few days. It should be easy to “roll-back” your backups and restore the original data.
- Recover roaming machines: Many companies allow employees to leave the office with their computers or mobile devices. These machines often contain lots of important, proprietary information which isn’t located on the main servers. It’s good to have a centralized backup system which backs up remote machines, and allows all users to be administered from a single management portal. This way, IT can recover data from a stolen laptop, even if it hasn’t been in the office since the documents were created.
- High availability: In extreme cases – such as is common in banking, emergency services and manufacturing – you may need to ensure little or no downtime under any circumstances. If this is the case, you may require a replicated temporary staging site at a remote datacenter. If you shut off your home servers, operations should quickly switch over to the other site without noticeable service interruption.
- Legacy recovery: You may need to recover a 7 year-old-file, as is commonly done for legal reasons. Not only must this data be easy to locate, but you must actually be able to restore it. Often, older data can’t be restored because the backup media is physically damaged or the legacy formats are obsolete and inaccessible. Don’t let this happen to you.
As you can see, data backup and recovery is a dynamic process with lots of variables to consider. And every individual or company has differing requirements when it comes to the protection and restoration of their digital data.
Put together a backup testing plan for your own systems, and consider all of the possible scenarios that may come into play.
About The Author: Canadian Online Backup leaders since 2001, Storagepipe know how to keep corporate data safe and ensure it can be restored reliably, securely and in a timely manner.