The Definitive Guide to Backup

The definitive guide to backup, I am sorry to say, does not exist.  There is no one size fits all set of instructions that anyone can give you.  There are however some very key items that can help you to prepare your backup plan, and make sure that when a disaster hits, you are ready and can recover from it.  I am going to outline a number of these key items that you need to do in order to determine whether or not you have a good backup plan.

First Step in Your Guide to Backup: Evaluate

First things first, take a look at what data you have and what data you need to protect.  You will need to talk to the people that create, own, and use the data that you are looking to protect.  How important is the data to them?  How important is that data to the company or even to you?  Can the data be recreated, and if so how long would that take?  Are there regulatory requirements for how long you have to hold on to the specific data, or does it have to be encrypted in order to fulfill those requirements?  Based on that valuation, you can define how drastic of measures you need to take in order to ensure the availability of that data, or whether you need to backup that data at all.

Once you have determined what data you need to backup, you now need to figure out how long you need to keep each set of data and how quickly you need to be able to restore from it.  For instance, financial information may need to be kept for 7 years in certain industries, but in most likelihood you will only need to keep maybe 3 months of data backed up in a place that you can restore from quickly.  Whereas the older data can be stored in a somewhere that can take a little bit longer to restore from, say in the cloud, on tape, or removable media that is held offsite.  You very well might have different restore time needs depending on the data that you are backing up.

Determine How to Backup Your Data

Now that you know how long you need to keep each set of data, you will need to determine how you will do the backup of that data and where it will be stored.  In deciding on how and where to backup the data you have deemed it necessary to backup, you need to take into account the size of the data and the speed in which you need it backed up.  Say for instance, you have determined that you need to backup 6TB of data.  There are a lot of ways to go about backing that up, but say you decide to purchase a 24TB NAS to hold that backup data.  Over a 1Gbps link to that NAS it would take close to 14 hours to back that data up, assuming you get amazing speeds to that NAS.  With that amount of data you will not want to do full backups every day, instead you will want to do a full backup once a week or month with incremental or differential backups everyday, as this would make the most sense based on the amount of data that you need to backup.

If you are not deploying a 3-2-1 Backup Rule, is your data really safe?

Now that you have your primary backup location in place, you need to think about what happens when that NAS dies or there is a fire or some other disaster in your office that takes out that NAS?  This is where you need to think about what is widely referred to as a Backup 3-2-1 Rule, or the Backup Rule of Three.  It basically says that you should have 3 copies of anything you care about, in 2 different formats, and 1 of those copies of data should be offsite.  So make sure you think about how you are going to get that data offsite, be it the cloud, tape, external hard drive, replicated to another NAS device, it does not matter as long as it is offsite and safe.

The types of backup you can do differs as much as the data that you need to protect.  So make sure you take everything into account when choosing how to do your backups.  Whether you are just doing simple file-based backups, image-level backups of volumes or entire machines, application-level backups such as a SQL database, or virtual machine backups.  They all serve a purpose and you need to look at your valuation of the data and the requirements that you put in place when creating your backup plan.

Monitor Your Backup Reports

You have your backups going, your retention rules for how long to keep backups are in place, you are putting data offsite, everything is going like clockwork, or is it?  You have to make sure you are getting accurate reports of your backups.  Do you have automatic email reporting setup so you know what is going on?  Do you actually look at the report you get, and more importantly take action on that report if there is an issue?  We have seen so many horror stories in NovaStor’s Technical Support where people setup their backups correctly, saw that is was working for months without any issues, and then stopped looking at the reports. The problem is, if the software all of a sudden throws an error because the media is no longer good or something else happens, you need to be aware of it so you can act on it.  We have seen cases when that error continues to happen for months because the person that setup the backup just never payed attention to the emails that came in alerting them of the issue.  The end result in this scenario is that when a restore was needed, none could be performed because the last month worth of backups were bad. If the reporting had been looked at, the issue could have been resolved and the restores done as intended.

Test Your Backups

That brings me to my final point in the guide to backup.  Even if all the reports and verification of whatever backup software you are using tells you everything is perfectly fine, you still do not necessarily have a good backup.  You do not have a good backup until you can prove to yourself that you can restore from it and that it does contain the data that you need it to in order to fulfill your restoration needs.  About one year ago, I wrote another article about the need to make sure you can restore, it was entitled Schrödinger’s Backup: The Good and Bad of Backup.  I would highly suggest you read that article as well, as it goes into a more in-depth strategy for confirming your backup and your disaster recovery strategy.