Backup Management – Five Tips for Success

backup servers

Five Reasons Why Managing Your Backups Isn’t as Easy as You Thought

Lying-Business-ManJust select the folders you want to back up and pay the bill.

You’ll never need to restore unless the building burns down.

Wrong and more wrong! If anybody tries selling you on a ‘quick and easy’ backup service with the above hogwash, they are lying through their teeth.

When it comes to backing up your system, you cannot afford to be lazy…and I mean that in the most literal way possible. One power outage, one hiccup in the system, one act of nature can just about wipe your company’s system out of existence. Such losses have crippled more companies than I care to think about.

Fortunately, an attentive backup policy and a little vigilance can prevent this type of catastrophe. Consider these five points when deciding how much time (and money) you want to spend on your backup systems.

  1. If you’re not backing up the whole server or PC, you’re wasting your time.

You could spend hours picking through your servers to whittle away the directories that you think don’t have any valid data in them…but for what? Just to reduce your backup storage by a few Gigabytes or your bill by a few dollars? If the backup software you use is any good, it should have:

• A 99% de-duplication rate, which won’t even back up the Windows folder

• Default exceptions, which will exclude those anti-virus definition folders anyway

A more likely situation is that some software vendor or someone else at your company will save some application critical files outside of those folders you selected to back up, and then you’ve wasted all that time setting up your backup sets.

  1. Don’t let your backups fail for more than one night.

a-time_is_money-1502072Many other very specific scheduled events take place throughout the software that your company uses. If payroll puts out their weekly report only on Friday, and the estimating guys upload their pictures after lunch then delete them the next day, how many of these events could you miss should your backup fail for three nights?

Of course there are those fluke network or power blips, which can explain a singular failed back up one night. But wouldn’t you rather check that log and find that it rebooted during the back up instead of finding out that after three nights of failures the firewall changed and the network guys can’t fix it for another day?

“Time is money” is one of the most overused phrases I’ve ever heard, but I’m sad to say that it’s the best one that applies here. If your users have to spend extra days or weeks re-creating the files they couldn’t restore then you’re going to have some ‘splaining to do.

  1. Don’t ignore the exceptions.

Unless it says ‘successful’, it’s not a real backup.

If it’s yellow, it might as well just be red.

Exceptions to your backup jobs usually come from the most important applications you back up. If your job log shows that some of the databases backed up successfully, but that big one with all the complicated settings didn’t, chances are that database is the only one you can’t afford to not back up. If your mail backup is barking about logs missing, I bet that when you go to restore it, it will give you the same kind of message.

My advice is to hunt down every exception and resolve it as quickly as possible, because it will save you heartache when you try to restore.

  1. Educating your Users is not enough.

It’s almost impossible to back up every single file, database, and email in your company. Sure, you can run reports and audit those reports to make sure all the workstations are added to the backups. You can send emails to all your users urging them to back up, but there’s always going to be some rogue or clueless employee who doesn’t save their data in their mapped drive or doesn’t join their laptop to the domain.

teaching workersIn cases like these, covering your end of the bargain as a backup administrator just isn’t enough. When those files are long gone, someone with a “C” at the beginning of their title isn’t going to care whose fault it is if they don’t have a backup of their reporting data.

My advice is, don’t make it a choice for users. Before that laptop goes out the door it should have a backup agent on it and a group policy from the domain to lock it down so they only have one choice where to save their data. That way you know that there’s no reason why it doesn’t show up in your reports.

Before that server goes into production it should be added to the backups and a human should monitor it for a few days after its first back up. You’re not going to make any friends when users get their totally locked down laptop, but you’ll make it up to them when they leave it in an airport and need their files back.

  1. Your backups are only as good as the time it takes to restore.

Every nightly backup report that you see for your big ol’ file server says that it backed up hundreds of files overnight without any problems. But have you ever had to restore the whole thing from your off-site backup?

My guess is that you worked pretty hard to get all of your servers backed up, ironed out the kinks and then didn’t take the time to test how long it would take to restore. Some cloud backup providers are nice enough to ship you a hard drive with your Terabytes of lost data on it, or will let you spin up your server on their hardware. Like I said, those are the nice ones and the nice ones cost lots of money. If you’ve skimped on your off-site backup provider or your ISP you might be waiting for a week for your data to restore over the Internet.


Pat Vince DRS

Patrick Vince

Cloud Services Engineer at DRS LLC