Data Backup Options

[darrenyorston]

I'm interested in what options people are using to automate the backup of essential data from their server. I was backing up remotely about a TB of data however I am considering moving to backing up physical disks to an offsite location. In the long run its cheaper and it negates the security issue of people having my data.

 

What ways is everyone doing their backups?

[Scott Pederson]

With regard to backups, there are LOTS of things to consider before deciding on the right solution for your environment:
 - Is this for personal use or for business use?
If you are using this for personal or business, one of the first questions to ask is - how long can I live without the data before it becomes critical? For example, if you're using this for business, and you must be up and running 24x7, then backups (part of your disaster recovery strategy) become crucial for the livelihood and survival of your business. If you cannot afford an hour's worth of downtime, then your backup / recovery strategy must be able to lose / notify / restore / restart within that hour.

If however, you're using this for personal use, and (for example) driving to the location of your offsite backup is housed might be a day or two, or this weekend, and that is acceptable - by all means it's a good strategy.

 - How much is this data worth? Now, I'm not talking on the open market - but how long did it take you to collect the data in your backups? If your backups contain pictures of your friends and family as they've grown up, then it could be priceless - you can never recreate those pictures. As such, it's very important for a solid backup strategy.

 - Does your backup / recovery strategy take into account the different types of data loss:
1) Someone deleted a spreadsheet by accident, they need it back
2) Somehow your data was compromised by Crypto-Ware, and cannot (or will-not) pay the ransom.
3) The building your system(s) exist in have burned down, swept away by flood, or your cat accidentally stepped on the DELETE key.

The more you think about "What If's", the better you will be able to recover.

So - (TL;DR) Many admin folks try to follow a 3-2-1 strategy. 
3 - copies of your data locally
2 - two concurrent backups (in-case one gets corrupted, tape gets eaten, disk crashes etc)
1 - offsite and, even better, OFFLINE backup for those super-emergencies.
How often to replicate, backup, and rotate off-site will depend on how much time you would need to restore it. If you must have it available at all times, you're looking at a pretty hefty load-balanced cluster with failover options, high-speed restoration, multiple site and even cloud-based solutions.
If you are using a storage array at home to make life easier, but want some option in-case disks fail, power supplies, etc. then keeping an offsite solution is a good idea. removable disks are low-cost insurance.
Hope this helps!

One quick follow-up:
How to automate it? Lots of software tools exist for running backups. You can even script it yourself using things like RoboCopy for Windows, or RSync for *Nix. For MACs, Time-Capsule.

Microsoft has a system utility called "SyncToy" that has been used by many to make copies of your data to another device or folder. It's smart enough to skip over the 80% of files that already exist, and only copy the 20% that have changed since last backup. (Same for Robocopy and RSync for that matter)

I also did not mention things like Versioning - keeping multiple versions of the same file, or EOM or EOY backups for accounting, as well as compliance requirements (like keeping 7 years' worth of IRS backups for auditing purposes).  All kinds of cool things to think about.

Edited February 4, 2020 by Scott Pederson