Let’s face it, stuff happens. Computers crash, hard drives fail, access to the Internet goes down, and electricity goes dark. We rely on our infrastructure to keep our data safe, yet sometimes the unexpected happens. Thankfully, we can be proactive and prepare for most disasters. Backing up our data is an essential part of mitigating those risks, and we’re going to discuss some viable business options.
We are big fans of the 3-2-1 backup strategy.
Always keep 3 copies of your data,
2 of those copies on-site and on 2 different devices.
1 of those copies off-site
Let’s address the two local copies. One copy is the original data itself: what you keep on servers, hard drives, or network storage. The second copy is going to be your regular backup. We recommend continuous or daily backups, locally. This gives you the first two on-site copies.
The generally accepted approach to fulfill this is to keep a local copy of all your data and in some cases your software and configurations. Usually on affordable, locally connected hardware. Depending on how your network is structured, there are numerous options for backing up to some form of a physical storage device. Directly connected external hard drives and network-attached storage arrays are good options. Some are portable, and others offer increased security, or some provide a single copy, where others have built-in redundant failovers. Regardless of which is best for you, it can’t be the same hardware medium that the original data lives on.
A physical device connected to your system offers peace of mind that data is there when we need it. On the other hand, data backups to physical devices are only as reliable as the backup software or tools used to make the copy and whatever accountability steps need to be taken to regularly test recovery. Testing data restores, locally and from remote backups is critical. Nothing is more painful than going to restore from a backup and failing because the backup was bad.
Network-attached storage devices can be inexpensive and have a small footprint. Scalability is simply about adding as much space as needed. Regardless of the size, these units back up data on multiple hard drives simultaneously. This characteristic is vital because hard drives tend to fail for many reasons, and redundancy has saved countless companies from backup disasters.
The final piece is the off-site copy of your data. This is to protect you in the event the local copies of your data are damaged, stolen, or compromised somehow. There are many natural and unnatural ways for this to happen and they happen more often than you would think, hence the need for an off-site backup.
In the early days of computing, you would use tape backups that a designated person from the office would be tasked with rotating tapes that they brought home or to another branch office or something like that. These days, we have cloud storage with encryption, high-speed connections to said services, and FedEx when you need all your data restored pronto.
Backing up to dedicated cloud storage services like BackBlaze and Amazon Web Services is both affordable and reliable. You could conceivably back up your entire server for about $10/month, in some cases, less! Generally, if you are going to restore a file or group of files, you are better doing it from local storage backups (see above), but sometimes that’s not possible and that’s when you rely on your off-site backup to restore. Now, for a larger-sized restore (like, the whole server), this could take weeks to complete over the Internet, so it’s not practical for what is called a bare-metal restore. In this case, some backup services will copy your data to an external drive and ship it to you so you can treat it like a local drive and restore it more quickly. There is a premium cost for this “emergency” service, but it could be the difference between being up and running in a few days versus weeks or months. Think ransomware.
Data backed up to a reputable online platform should be protected with 256-bit AES or similar encryption and a historical record of file versions if you need to recover from a malicious computer hack or accidental file deletion. Most restores are a single file or folder here and there and generally can be restored within minutes, even from an online service.
There is a misconception that using a service like DropBox or OneDrive is the same thing as a backup. The thinking is that because you can turn on versioning, that you get a copy of every version of the file you saved, so you get history, much like a typical backup would give you. For a single user, this may work because there is a single source for the data, and restoring would be back to a single or secondary computer like a laptop. This completely falls apart in typical business settings where data is shared by many users, usually from multiple sources. Services like DropBox don’t centralize your data. They maintain “shares” per user, so doing a mass restore requires a business-grade account and you would have to access any one of the user’s shared data. Again, most business environments use centralized file-sharing models and this “sync” option doesn’t scale or work well, even if you remember to turn on file versioning. The best solution is to backup locally and then back up your copies to an external, off-site source. We should also mention that “sync” solutions don’t work for database or email servers or a multitude of other applications based on databases that are always “live” and can’t be synced to effectively backup, it simply doesn’t work.
We always say that “good backups are the single most important thing you can do in IT” and we believe that with every ounce of our being. This is why the 3-2-1 backup strategy is what we constantly recommend and implement for our clients. More often than not, backup solutions for most small businesses are very affordable, easy to set up, and are worth every expense because ultimately you are protecting data that is so much more valuable than the cost of any backup will ever be.