Whether it's Mother's Day, Veteran's Day, or Administrative Professional's Day, we like to assign days to remind us to pay attention to important things. World Backup Day is no different. Started a few years ago by a group of Redditors, World Backup Day is primarily aimed at reminding people to protect their digital lives with backups.
Many of us in IT would benefit from reminders to test backups, too. But there's a litany of additional items IT has to stay on top of to keep operations running smoothly and to aid recovery when disaster strikes. Along with ensuring that backups are made and frequently tested, here are five details you might not think about every day. Consider using this awareness day to double-check that you have all your ducks in a row, not just your backups.
1. Software media and licenses
If an application needs to be reinstalled following a restore, do you have the media readily available? What about the licensing?
This may seem not to be a big deal in a day and age where you can download virtually any piece of software from the Internet. However, I've seen several instances where a client needed to reinstall software released a few years ago, only to find it could no longer download the required version. Worse, the client had made customizations that didn't work with a newer version. Worse still, the license keys on record didn't cooperate with the new version without paying for an upgrade.
It's wise to keep a software repository that contains installation media (installers, ISO images, and so on), as well as copies of the license keys for every piece of software you run. As with network device configurations, don't store these on a device that could itself be affected by the disaster you might find yourself recovering from. A small NAS or even an external hard drive can do the trick.
2. Labeling
If I pull two cords out of out of a switch in your data center, how quickly can you figure out where they should go? How about if I pull them from the back of a server rack? If you're on vacation and need to describe to a nontechnical coworker which server he or she should turn back on after a power failure, could you do it over the phone?
If you answered no to any of those questions, it could be because you haven't labeled your inventory very well. Labeling is one of those tasks that some in IT (typically the perfectionists) can't get enough of, but many of us can't stand to spend our time on. After all, the data center changes so frequently -- why bother labeling anything? The answer is that it can save you a tremendous amount of time in a disaster (think swapping out a failed high-density switch) or even in a simple accident.
3. Network device configurations
If a switch or firewall on your network fails, do you have the configuration backed up so that you can easily restore it to a replacement? If you don't, imagine how long it might take you to reconfigure your firewall from scratch. If you don't have good documentation, it could go on forever. That's especially true if you're running site-to-site VPNs: Getting those running again could require coordinating with the owners of all the VPN peers.