Every organization relies on data that must be protected and backed up in a reliable and secure way by authorized personnel. While the content or criticality of the data may vary, the processes behind a successful backup strategy are universal. This checklist will help you establish a standard framework for backing up files that you can count on in a crisis.

Assess the environment
Take a look at your systems and data and identify your needs. Where is critical data stored? How often does it change? How much data is there, and what type does it represent? Do you need to arrange firewall access across subnets so a centralized backup server can connect to desired hosts or can these hosts can be configured to back up data to the cloud (if applicable)?

Factor in user workstations and/or mobile devices. No system should have a single copy of data, and since they’re often used to generate or change information, a plan should be devised to product this material.

Always follow any security guidelines per your IT or security departments. If no such departments exist, research backup security best practices.

Determine backup selections
Decide which systems and/or data need to be backed up. Separate the data by three categories of importance: low, moderate, and high. Take note of the locations of the data on your systems so you can plan the backup implementation.

Sometimes you don’t actually need to back up data; you may be able to synchronize it to other systems, such as by using a Windows batch file to map drives and xcopy or robocopy data or a using a Linux cron job to rsync information. This can be useful for low-level data that would be nice to have available if a system fails but which won’t break the company if lost. Keep in mind that if you use this method, corrupt data will synchronize just as easily as clean data and the latter may get overwritten by the former, rendering it useless.

Similar principles apply to directory data. While it’s a good strategy to back up Active Directory in case you have an entire site failure, multiple domain controllers that share the environmental data ensure the loss of one server won’t affect the company.

It may be necessary to back up entire operating systems, but not always. Servers can be built fairly rapidly in today’s automated and virtualized environment, so focus more on the data these systems hold rather than whether you can restore Windows or Linux itself.