The Complete Guide to Backup
An introduction to everything about how to backup home and business computers.
How to Backup
This guide is designed to help everyone who needs to back up data, from the casual home user to the power administrator. While we do talk about KeepVault, this guide is designed to be useful no matter what backup service you use.
We hope you’ll find this to be a valuable reference, both now and in the future.
How to Use This Guide
This guide is a reference, and not necessarily meant to be read straight through (though if you do, you get a cookie). Make use of the index to find sections relevant to you. In some sections there is breakout information for specific user types. Find the user type that is relevant to you for information specific to your needs.
If you are looking for something on this page, use your in-browser search (CTRL+F) or in Google or Bing use the site-specific search filter “site:keepvault.com/how-to-backup” followed by your search terms.
We’re always working on improving this document. If you have suggestions or corrections, feel free to contact us through our super simple feedback system.
In the past, people didn’t think much about backups. “Put a copy on a floppy” was the way it was done. Now as people increasingly integrate technology into their lives, as well as accumulate files over many years, the need for persistent backups is more critical than ever.
Despite that fact, 9 out of 10 people and 6 out of 10 businesses do not back up all their important data. Those numbers are even more shocking when you consider how easy online backup has become, the prevalence of high speed internet, and how storage keeps getting cheaper.
This is a great time for consumers and professionals because we’re at a point where the need has never been greater, but the solutions has never been easier or more affordable. Now is the best time to be in the business of backups, whether it is for yourself or as a service you provide.
The first steps you need to take is to determine what needs to be backed up, estimate how much data that is, and determine how much down-time you can sustain.
Causes of data loss:
- Hardware fatigue
- Power grid surge
- Static electricity
- Fire or fire sprinklers
- Spilled beverages
- Rapid temperature change
- Malicious deletion
- Accidental (full) deletion
- Government seizure
- Magnetic damage
- Physically dropping
- Vibration or g-forces
- Three year olds
Ask yourself: if all of your computers failed at this moment, what could not be replaced? You can always reinstall your operating system, or software, but it’s the files that are unique to you that cannot be replaced if lost. This could include:
- Medical records
- Finances and tax records
- Invoices and receipts
- Electronic books
- Manuals and documentation
- Personal documents
- Audio books
- Application files
Application files are the files your various software programs create based on your use of them. This is how they remember what recent files you opened, what settings you configured, and most importantly, some of the information you actually saved through the program.
For example, you could reinstall your tax software if you lost it, but you would not get back the information you put into it before.
There are different levels of precision when it comes to choosing what to backup. We could summarize them as:
- Select files and folders
- Select folder groups
- Everything except the operating system
- Absolutely everything
You may ask, why not back up absolutely everything? The answer is: you might. Backing up everything is the surest way to know you have anything you could want backed up, and sometimes the most convenient when you lose everything, because you don’t have to reinstall the OS or software. (We say sometimes because in some cases reinstalling the software is faster than restoring from the cloud, if time is more important than effort.)
But that level of protection and convenience comes at a price. Depending on what you put on your computer, it could cost ten times more to have everything backed up. Beyond price consideration, there is an issue of processor resources and network bandwidth (discussed more in the Security section below).
To mitigate those two issues, best-practice is to do a full-system backup scheduled nightly. This saves your processing power and network resources for when you aren’t using them. The disadvantage to this is you lose your instant backup protection through the day, though some people consider that a small sacrifice.
Depending on your habits, most of your irreplaceable files are in the /users folder in Windows. The /users folder contains many of the folders you are familiar with such as /My Documents, /Pictures, /Downloads, etc. The /users folder also contains your hidden applications folders, which has that program-specific information we talked about earlier.
Note: the Applications Folder is a hidden folder. Make sure your backup software has an option to include hidden folders if you want to back it up (KeepVault has a checkbox).
For this reason many people back up their /users folder as a minimum first step. We would call that a select folder group, because the /users folder encompasses all your computers users, along with each users’ folders.
But it is often the case, especially if you have a second hard drive on your computer, that you save files outside of the /users directory. It is important to have an awareness of where you are saving your files.
If, like many of us, you aren’t precisely sure where all your important files are, we recommend a popular freeware utility to help you find them. WinDirStat is a small program that indexes your folders by size. It’s a great tool for finding data that you no longer need, as well as data you don’t want to lose. It also presents them by type, so you can see all the different locations of .pdf’s or .jpegs. This is great for ferreting out files in places you may have forgotten.
Note: Be aware that WinDirStat shows you all of your files, including critical Windows system files. Do not delete anything without being 100% sure it is safe to delete.
Not only can WinDirStat help you find the files you want to save, it can also help you estimate how much storage you need. It displays the size of any directory, so you can see the total size of your /users directory, as well as any other directories you deem important. You can also see the total size of your hard drive or drives.
You now have everything you need to identify folders to back up, as well as estimating the total storage size of your critical files. Armed with this information, you can now set about deciding how much storage you need to purchase.
You need to do this for every computer you with to protect, including laptops. This does not apply to tablets that run on a tablet operating system (like iPad, Android, or Kindle), but there are some laptops that come in a tablet form factor, such as the Microsoft Surface Pro, HP ElitePad, Dell XPS series, or Asus VivoPad which are effectively laptops and can be backed up with the KeepVault software.
Note: The distinguishing factor for a tablet that is a laptop is that it runs a full version of Windows 8. KeepVault is compatible with Windows 8 and Windows 8.1.
Your data can be transferred and stored encrypted or unencrypted. Unencrypted files can be stored locally more quickly because the encryption process takes a lot of processing power. However unencrypted files are more vulnerable to data theft.
Often on a secure local network, such as a secure home network, local backups do not necessarily need to be encrypted. On a USB drive there is no potential issue of network sniffing so it is very common to not encrypt backups on a USB drive unless there is concern of the drive being physically stolen.
Over the internet, or an unfamiliar network, you absolutely want your data encrypted to protect it against theft.
How secure is the encryption?
There are two parts to the answer to this question. The first deals with when the encryption happens, and the other deals with what the encryption is.
The when is very important because KeepVault uses end-to-end encryption. That means your data is encrypted by the KeepVault software on your computer before going out on the network. The beauty of this is your data is encrypted at all times--even on unsecured networks. If someone sniffs your backup data at an internet cafe, all they’ll get is useless gibberish.
The encryption itself is 128 or 256 bit encryption with KeepVault Home and Pro respectively. That is the same level of encryption commonly used to protect your credit card or password information online. Some services offer higher levels of encryption, but be aware of when it happens (some services advertize very high encryption when it doesn’t happen until it reaches their servers), and whether the extra work required by your computer is necessary.
The main threat to your data is your identity. Ensure that the e-mail you use for KeepVault has a strong password, and that you never access your e-mail unencrypted in a public place. Also physical access to your computer is a vulnerability, so ensure your operating system has a good password. If you use web access to access your files, be sure to use a strong password. Your data is only as secure as the weakest link in your chain.
In the Requirements section you went through the process of determining what folders you needed to back up, and roughly how much data they total. Now we need to look at that information to develop a backup strategy.
In a typical case, a home user will have a smattering of documents and pictures ranging from 20-40GB on each computer. With the prevalence of DSLR cameras, you might have a pictures folder somewhere with several hundred gigabytes. If you are really clicky or are into video, you can easily find you are using one or more terabytes.
Note: A terabyte (TB) is 1000 gigabytes (GB).
Starting from a typical case again, if you have a desktop and a laptop computer each with about 40GB in the /users folder, then a 100GB online storage account would be adequate, split between the two computers.
It is best-practice to also have a local backup to a USB hard drive, USB thumb drive, or to a network drive.
Businesses can fall into two categories broadly: those with a server, and those without. Those without a server are configured like a home setup, with each workstation backing up on it’s own.
Those with a server often have a policy of keeping work files on a network drive. They also typically utilize Microsoft Exchange for email and user profiles. By centralizing the data on a server (or servers), backup can be handled at the server level.
If the policy is to save files on a workstation hard drive, you could either back up those folders directly to the cloud, or back them up onto the server. You then have the option of backing the copies on the server up to the cloud.
A strong plan includes mitigating the loss of the server itself. This almost always involves backing it up to the cloud, or to another company location via remote network backup. Because of the load and bandwidth these backups entail, they are almost always scheduled to run during the night or slow hours.
Your solutions will be a determined by whether the customer is a home or business. In either case you can have all the devices under your KeepVault account.
Aside from software that works for their clients needs, a major consideration for MSP’s is administrative controls. Those will be discussed more in the Execution section.
If you do have much larger storage needs, you may need to consider a combination of local and online backup.
Local backup, as the name implies, is backing up to a USB hard drive, or another hard drive in your network, whether that is on another computer, a home server, or a NAS device. The backup is local because the backup hard drive is physically in the same building as it’s source.
Local backup has the advantage of being generally cheaper and faster, particularly for very large amounts of storage. The drawback is that if fire, flood or lightning destroyed your building, you’d lose both your computers and your backups.
Online or cloud backup (synonymous) has the advantage of being off site. Theft is not an issue if the data is encrypted. If someone were to somehow manage to steal a hard drive or rack of hard drives from the cloud storage facility, all they’d get is useless gibberish. Online backup is a little slower, and generally costs more.
NAS stands for network-attached storage. In essence it is a hard drive or bank of hard drives connected to your home network via a machine designed for that purpose. NAS devices are ideal when you want to backup multiple computers to a drive, have large amounts of storage, use RAID, and do not wish to dedicate an entire PC for the task.
Not all NAS devices are the same. It is important to research which manufacture and model are most likely to work well for you.
There is one more technical consideration when comparing local and online backup: Online storage uses a system called RAID to distribute duplicates of data across multiple hard drives. If one hard drive were to fail, a RAID system can automatically re-assemble it from the other drives.
This is how data storage providers can be certain that they won’t lose your data.
It is possible to have a RAID system in your own home/office for local backups, but this is recommended for users with a good familiarity of the technical aspects of networking and advanced disk management, and is generally the domain of IT professionals.
Note: There are different types of RAID configurations that affect how much storage is actually available. The more redundant the RAID, the more protected your data is, but redundancy consumes space. So, for example, if you had a NAS with five 1TB hard drives (5TB), you would only have 4TB available with a RAID5 and 3TB available with RAID6. When you calculate your needs, remember to compute what will actually be available against your needs.
But lets say you do have a lot of family pictures and video, totalling 1.4TB. How should you proceed?
There are three common options for this:
- Backup your critical files except your pictures to the cloud encrypted, then backup the pictures to a USB hard drive unencrypted
- Backup everything to the cloud
- Backup everything to the cloud (encrypted) and a USB hard drive (unencrypted)
This is by no means the only way to go, but it is common.
For the first option you would simply purchase a 2TB or larger external USB or Firewire hard drive.
Note: USB is backward compatible with itself. So for example you can use USB 2 ports for a USB 3 drive. It will simply perform at a USB 2 speed. If you use Firewire, be certain you have a Firewire port and note whether it is the large 6-pin or the small 4-pin type.
For the second and third option you need to be willing to purchase at least 1.5TB (or 1,500GB) of online storage.
You can get creative with your solutions. For example, if you store all your raw photos locally, there is nothing preventing you from storing just your edited photos (theoretically a much smaller set) in the cloud.
To actually do the backups you need software. Windows has built-in tools, but they are notoriously unhelpful, and not nearly as robust as software dedicated to the task.
The software needs to meet your requirements. If you need local backup, online backup, or a combination of both, the software you choose should support it.
All the major cloud services offer their own software, but they aren’t all the same. Software that runs as an application must be restarted every time you reboot. Software that runs as a service will automatically run in the background, much like antivirus software does.
Keep in mind that with a cloud service their software is usually included, but not always. Check to see if they charge separately for their software. This is often the case for companies who sell software for only network backup only.
Not all cloud services are equal either. Though most major services are similar, there are differences that should be considered:
Hot Storage vs. Cold Storage
Some backup services put your data into cold storage. That usually means it is actually transferred to digital tape or DVD and locked up somewhere. The key difference is that it’s much cheaper, but it can take days or weeks to get your data back.
Data kept in hard drives (like the RAID arrays mentioned earlier) are instantly accessible. This is important if any of your data is time sensitive, which is often the case if you have a home business. Be aware of how your data is stored and how long it takes to restore when shopping for a cloud service.
Also, be aware of pricing structures for restoration. Some companies charge one fee for everything, including storage and bandwidth to and from their cloud. Others break these out separately, and may charge for storage, bandwidth to backup, and bandwidth to restore.
There is nothing necessarily wrong with these different storage and pricing models, but it is important that you are aware of what you are buying and how it fits your needs.
Imaging is a way of capturing a snapshot of your entire computer. Imaging can be resource intensive, but it is the best way to restore an exact copy of your entire hard drive.
Because of the time and bandwidth often involved in imaging, they are nearly always scheduled on a weekly or nightly basis. Users who desire a very high level of protection use realtime backup throughout the day, while making an image at night. This gives them the best of both worlds, but is very resource intensive!
There are as many different ways to do imaging as there is software that does it. Some support differential imaging while others do not.
Differential backup, as well as differential imaging, is a process where only changes in files are backed up, rather than whole files. This process happens in the software, and only matters to the end user because it can drastically reduce the time and bandwidth needed to update their backups.
Deduplication is a feature where identical data is not backed up twice. For example, if you are doing whole system backup, software that exists on two separate computers will not be backed up twice if they are identical. This is typically most useful in image-based backups. (KeepVault does not yet support imaging or deduplication, but does provide client-side differential backups.)
At this point you need to have decided what folders need to be stored, and where they should be stored. You should have a cloud service provider which includes software that does what you need it to do. Finally you should know what you need to buy for what type of storage you intend to use.
The details of execution are largely dependent on what software/cloud provider you choose. Follow their own documentation for the best results. The important thing is that you follow through. Data is never secured by good intentions alone!
Because this is KeepVault we’ll show you the KeepVault way:
You can sign up for a subscription to our cloud service on this site. We offer a Home plan with a basic feature set, and a Pro plan with an expanded feature set and server support. You can see the difference on our Features page.
Subscribe to one of the plans. If you don’t know which, start with Home. You can always upgrade later on if you need to. You can then download the KeepVault software from our Downloads page. Find your operating system on the list and download the installer package.
Once downloaded, install the program. After the program is installed, launch it.
Note for IT Professionals:
If you’re backing up a number of clients to the server, KeepVault has a connector for that purpose.
If you have done everything right, you should see your progress indicators tell you that the backup is in progress. Depending on how much data, your network and internet speed, and encryption options, it make take a few days for all the data to be initially backed up.
If the backup reduces your computers performance, you can pause the backup process until you are ready for it to resume again. You could also schedule the times you want the backups to run.
Some backup services allow you to mail them a hard drive full of your data, which they can then quickly transfer to their servers. That can be ideal when the backup would otherwise take several weeks. This process is called seeding.
Administration for MSP's
An important aspect for Managed Service Providers and IT Professionals is their ability to manage their cloud storage account, users, and devices. If you are a reseller it is important that your service is structured to support sub-accounts. Those that do vary in specific features.
This is the part that is unfortunately often neglected. All backups should be tested, but the more complex the setup the more important testing becomes. Even Pixar suffered a data loss followed by a restore failure.
There are two reasons why testing is important:
- It ensures your data was not corrupted in the backup process
- It familiarizes you with the restoration process
If you have time-sensitive needs, it is also useful to test restoration in order to build an accurate timeline. One service we (and others) offer for time-sensitive recovery is disk by mail. We can copy your data onto a USB hard drive and mail it to you (overnight in some cases). This avoids a bandwidth bottleneck when you are dealing with large amounts of data.
For specifics, refer to the documentation that came with your software and service. As always, we will show you the KeepVault way right here:
A spot check is when you download one file or a reasonably small folder tree, and check to make sure it opens properly. This is a quick if not through way of getting a sense if things are working. The critical thing isn’t just that you can download it, but also that it is not corrupted, which is why you should make certain the files actually work.
A fire drill is when you take an assessment of where you are at a given moment. It’s 3PM on a Friday, what would happen if you lost all the computers in your facility? This exercise can rage from a theoretical exercise to a complete practice of a recovery situation, similar to our next test, RTO.
Recovery Time Objective (RTO)
RTO is a type of recovery plan where you evaluate the potential business impact of an outage, and craft a recovery plan that prioritizes which systems should be restored first in order to mitigate harmful effects.
To add to the jargon, there is also Recovery Time Actual (RTA) and Mean Time To Recovery (MTTR). RTO is a series of goals--what you want to happen under certain crisis levels. RTA is what actually happens. MTTR is a more general number that should be somewhere between RTO and RTA until you have multiple RTA’s to sample from.
Put another way, RTO answers “when should it work again?”, MTTR answers “when will it work again?” and RTA answers “when did it work again?”.
It is best if you can actually dedicate one or more workstations or servers to a full restore. Remember when restoring to multiple machines simultaneously your bandwidth will be divided. For that reason restoring a single computer won’t necessarily accurately demonstrate a time-critical restore.
Some backup services offer something called “cold failover”, where your computers will directly map to the backup providers drives until a restore (failback) can be made. This can sometimes reduce downtime to minutes instead of hours or days.
Setting up and maintaining readiness for failover can be a challenging, and it is unfortunately common for a company to try to use it in an hour of need, only to learn they had not tested it adequately. It is also significantly expensive, but still much better than losing business.
For “hot failover” (uninterrupted service) talk to your IT professional. That type of system is usually configured on-site and is outside the scope of backup services.
One of the best ways to know how your backups are doing is to receive notifications. This way you can be alerted of errors or capacity issues.
KeepVault has account level and user level notification settings. This enables IT professionals to limit the alerts to themselves, rather than having them go out to customers.
The process of restoration is highly dependant on the software you are using. Restoration should be part of your backup plan.
If you are doing image backups, that image needs to be flashed onto a working hard drive.
If you are doing file/folder backups, you usually need to install your operating system on your new hard drive, reinstall any important software that you have (including the KeepVault desktop application), then perform a restore to download your files back into your system. If you make good backups, this should be all it takes to have you back in good working order.
Many backup providers offer a restore by mail service. This is basically the opposite of seeding, because your backup provider mails you a hard drive with a copy of your backups. In many cases 2 days of shipping is much faster than possibly weeks of restoration over the net.
Take every restoration opportunity to gauge time. This will serve you in the future as you improve your MTTR (Median Time To Restore) estimates. Remember the fewer machines using bandwidth, the faster the restore will be.
This completes the Complete Guide to Backup. If you have suggestions, corrections, or even just questions about what is discussed here, don’t hesitate to contact us. We’d love to hear from you.View Plans and Pricing