If the subject of backups sounds boring to you, maybe this will grab your attention: Without properly implemented backups, your business is vulnerable to partial or even catastrophic data loss, which could screw your company and tank your income.
Imagine losing everything you have published. Suddenly your website resolves to a “Page Not Found” error. What does permanently losing your content mean to you? Or if it is not completely lost but needs to be re-edited and reuploaded, how long will it take you to recreate your site from scratch? For most, full recovery after complete data loss takes weeks or months, or it might be near-impossible.
If your data is not meaningfully and effectively backed up, no legal scenario or technical solution will bring your site back to life.
Fortunately, reading this article just might save your business someday. Let’s review issues that could result in data loss, as well as actionable steps to help you attain confidence and peace of mind.
Hidden Vulnerabilities
Let’s start with the most common server configuration and backup strategy: a VPS or dedicated server with cPanel, DirectAdmin, Plesk, or Webmin software installed for server management. If that’s you, and if you have enabled full backups with the default settings in your software interface, you may think you’re adequately protected — but you’re not, and neither are millions of other hosting accounts.
“But they said it’s backed up!”
Sorry, no. Regardless of what your host, admin or webmaster says, if your data is all in the same place, on a single server, this makes you vulnerable to a very long list of unrecoverable failures. Some examples:
- If the underlying physical hardware isn’t properly monitored and maintained for storage health and disk replacement, then failure of one or more drives, or of the controller or algorithms used to manage multiple storage devices, can result in complete data loss.
- A security failure, whereby hackers gain access to your host system, can also result in your data being corrupted or deleted. Remember, corrupted data is unusable, and if an “immature” backup strategy overwrites your good data, you have no point of recovery.
- “Friendly” human error, or an error in the logic of your content management system or software, can result in unrecoverable data loss, if you are only using the kind of default backup system noted above and the last known good backup is overwritten.
- Though statistically rare, fire, natural disaster or terrorism can wipe out a website just as they can damage anything else.
The bottom line: If your data is not meaningfully and effectively backed up, no legal scenario or technical solution will bring your site back to life.
Preventive Measures
How can you head off these potential scenarios? First, make the target destination for your backups off-server, at another location, such as object storage. Alternatively, give your web host login credentials and start paying for managed backups to pull your data securely to a different location.
Okay, but how many backups is enough?
Managed appropriately, two copies of your data is reasonable for most situations. The first copy is your active server. The second copy is a backup within the same data center but stored on a different server that is connected by internal network only. This is critical to protect against your server getting hacked and providing access to your backups.
Three copies is exceptional. The third copy is a disaster recovery plan: a full copy of your data in another physical location, be it another data center, your home or office.
A fourth copy of your data constitutes as close to perfection as you can come. This is what most larger companies deploy. It can be accomplished as a “hot” load-balanced spare, which essentially means having two active servers. Alternatively, a slightly less sophisticated “cold” spare, which is not kept actively in sync but rather has a slight delay in its data copy, can also bring a site back online to a different version with just a DNS change.
Doing Storage Right
Of course, it isn’t just about quantity. Quality of backups is also crucial. Remember when I said you need to back up your data “meaningfully and effectively”? Let’s look at what that means.
First, it is crucial that your data be stored redundantly in a scheme using multiple storage devices. This is referred to as a RAID (redundant array of inexpensive disks) setup. If a single storage device fails, this allows for continuous operation when a storage device is taken out and replaced with a new one. Think of it like driving a semi-truck that has two wheels in every spot where your automobile has just one. If a tire blows out, the semi can continue its journey, albeit in a degraded state of function and future reliability.
Second, it is imperative that the health of your physical storage devices be monitored, in order to alert server administrators if they fail or begin to fail. The responsibility for such health checks is on whoever is managing your server. If your space is virtualized, such as in a cloud computing or VPS instance, then your host must monitor the underlying health of physical hardware — one of many reasons why it is crucial to choose web and cloud service providers with effective policies and procedures for ensuring the continuous health and maintenance of all equipment.
Lastly, make sure your server provider uses “enterprise-grade” components: high-quality gear designed for 24 by 7 by 365-day operation, with parts engineered for 2 million-plus hours of continuous usage. Why? Because all drive types fail, and it happens much more often than you might imagine. Failure rates more than double when “consumer equipment” is put in servers rather than enterprise parts, which can reliably survive loss of power without losing your data, due to power loss protection (PLP) technology, which consumer drives do not have.
Customizing and Fine-Tuning
It is important to choose backup strategies that reflect your specific needs. There are various server configuration options to consider. For instance, I do not recommend compressing your backups, as most adult sites use media files that are not compressible. Turning off this feature will also increase the speed of backup and restoration, saving precious time when it counts the most. Since time is of the essence, be aware that factors like the size of your ethernet ports, the location of your backup and the speed of the storage devices in use also help determine how long it takes to complete or restore a full backup. You will also want to gauge what impact doing a backup will have on your live web server performance, so be certain to surf your own website to see how it is performing during a live backup.
In addition, content creators should place “masters” — the original, highest-resolution recordings — in cold storage, whether online or offline, so they can be retrieved when needed. AI and rapidly evolving video formats present future opportunities that can only be realized if you still have the full-resolution files. It can be tempting to remove these masters after using them to transcode your content into multiple formats, because scaling high-performance storage can be costly and complex. Fortunately, there are a variety of affordable cold storage solutions on the market.
Furthermore, capturing a complete and full backup of your data is important, especially the first time it is run. A good general practice is to capture a full backup and then do weekly incremental backups thereafter. This will capture the net file changes between operations, while allowing for faster subsequent backups. For most businesses, this is adequate, though it depends on how often you update your site.
Pay extra attention when designing backups for databases that contain critical information about your memberships and content library. Giving yourself a lot of database restore points can be a lifesaver, and even doing a database “dump” or export to your local storage several times per day is a great and inexpensive way to give yourself flexibility. Then, when you do weekly incremental backups, you’ll also capture those database copies.
Maintaining at least three months of backups provides a reliable window for recovery in case of data loss or corruption, but it also means careful planning is required when replacing outdated backups and their associated incremental files.
Is It Working?
Monitor backups for completion and test the restoration of specific files in order to ensure the integrity of your data. Always confirm that there is enough storage space to complete a new, full backup — and validate its success before removing any existing data, since deleting backups prematurely could leave you exposed if the new backup proves unusable. If you have managed hosting and backups, consider opening a support ticket at least once a year to have your web host confirm the date of the last backup and how far back your oldest backup extends.
Don’t let the complexities of backups intimidate you. Take the steps needed to safeguard your business. Protecting your data — and livelihood — is much more than a “set it and forget it” exercise, so be proactive, stay alert and keep refining your strategy. Reach out to your host or admin team to confirm the details of your backup plan, including last backup dates, oldest available restore points and secure off-server storage locations. By doing so, you’ll ensure that your hard work, creative content, and customer relationships remain safe and profitable for years to come.
Brad Mitchell is the founder and president of MojoHost, which has served the industry for nearly two decades and has been named XBIZ Web Host of the Year several times. He regularly shares insights as a panelist at trade shows. Contact brad@mojohost.com to learn more about the suite of services his company offers.
