The thought of losing hours, days, or even weeks of work to data loss might be your worst nightmare. Backups are the best defence against this blight – so let’s dive into why backups are so crucial and how to get them right.
Backups are, in essence, a partial or complete replica of the data held by a particular system. They’re designed to protect systems against data loss, either through accidental file deletions, foul play from malware attacks, or hardware failures.
EPX’s “Professor” Elliot Bishop takes us through it……
Company servers
Commonly, the system being protected would be your company’s servers, especially file and application servers where your truly business-critical data resides. Backups are best created using an application created specifically for this purpose – Windows includes utilities such as Windows Server Backup built right into the operating system.
Both free and paid third-party tools are available as well, and paid versions might include extra features such as cloud replication and backup verification – more on both of those later.
How do I backup my company data?
If you were to save your backups in exactly the same place that your live data was stored, and this server later suffered a drive failure, then both your live data and backups would be lost at the same time, rendering your backup somewhat pointless.
This is why it’s important to store backups somewhere else, and ideally in more than one place. A good best practice to follow is known as the 3-2-1 rule:
- Three different copies – your original data, plus two different copies of it, in case one backup is accidentally deleted, targeted by malware, or becomes corrupted. You could even take this further and use two different backup tools, in case one stops working or fails to run.
- Two different locations – backups are saved onto storage drives just like any other data, and these drives still might fail one day. Having your backups stored on two separate forms of storage, such as an external hard drive and a NAS, can protect against hardware failure, and keeping one unplugged when not in use can help prevent malware from reaching it – this is called “air-gapping”.
- One location off-site – in the event of a truly catastrophic event, such as a fire or flood at your head office, having at least one copy of your backups in a completely different physical location, such as a satellite office or in the cloud, provides something to rebuild from. A backup app saving a copy of your backups to the cloud is sometimes referred to as “cloud replication” and can be a good way to check this box, especially if you only have one office.
Should I complete incremental backups?
Backups can be either incremental, where only data that has been modified since the last backup was carried out is included, or full, where the whole system is backed up at once. A good backup strategy should make use of both. Incremental backups are faster and can be taken more often without impacting performance too much, but it can take longer to stitch several incremental backups together again if ever you need to do a full restore.
Meanwhile, full backups can take hours to complete, depending on how much data you have, and often have a noticeable effect on system performance, but are much more comprehensive. Generally, incremental backups should be taken daily, and full backups weekly.
Windows servers also benefit from a feature called Shadow Copies – snapshots of your shared folders that can be taken every few hours throughout the day. Although these can’t be used to rebuild a server in full should something go majorly wrong, they’re perfect for quickly and easily restoring individual files or folders from an earlier point in the day to protect against human error.
Backup verification
It's also important to occasionally test that your backups are doing their job properly, so you know you can rely on them should they ever be needed – this is known as “backup verification” or “restore testing”.
This might be as simple as restoring and opening a single file after each weekly full backup, or as complex as restoring an entire server to a test environment and checking that everything works as expected. It’s a good idea to think about performing a full restore test once a quarter, or at least once or twice a year.
Backups stored in the Cloud
We’ve spoken so far about backing up on-site servers – but what about businesses that have already made the jump to the cloud, and use services such as SharePoint and OneDrive to store the bulk of their data?
Cloud services are usually more resilient to data loss straight out of the box – for example, if a PC with OneDrive set up were to fail, then that user’s OneDrive data would still be right where they left it if they signed in from a replacement PC. And SharePoint includes file version histories, allowing you to restore a file to an earlier point if needed.
But is this enough? Crucially, what these features within cloud-based services don’t protect against is cyber-attacks – a compromised account could still be used to deliberately look for and delete files in SharePoint, OneDrive, and even 365 emails.
Creating a backup strategy....
There’s a lot to consider when creating a backup strategy, but thankfully, there are tools available that cover all of these bases for you – cloud replication for disaster recovery, automated backups verification to prevent time-consuming manual tests, and even backing up cloud services to assure against cyber-attacks.
Should you have any questions about your existing strategy, or how you might be able to improve on it, our Service Desk or Partner Success team will be more than happy to help.
