The Best RAID Recovery Services and Hardware of 2020

Rarely do speed and safety go hand in hand, but that's the case when storing your company's data on a redundant array of independent disks (RAID). The technique offers the best of both worlds by speeding the flow of important data while potentially being a company-saver that can recover lost files.

Instead of sending the data to a single disk, RAID gets several drives to work together, combining their capacity. To the user, it appears as a single storage unit with one drive letter. A sophisticated controller board simultaneously accesses several of drives at once by striping the data across all the drives to speed its ability to read and write data.

Beyond death and taxes, one thing in life is certain: All drives will eventually fail. And with it goes your data. However, when a RAID drive goes south, the data can usually be recovered from the remaining data. The way RAID recovery works varies widely, based on which protocol is used, but the technique embeds recovery information, called parity data, along with the actual data. The original files can often be rebuilt using this parity data a byte at a time.

The speed and safety come at a cost: Using RAID reduces the array's usable capacity, sometimes by as much as half. Still, it's a win-win that can allow you to sleep soundly at night knowing your company's data is safe. Used by large corporations, server farms and data centers, RAID hardware and software has dropped in price to the point where it's a must-have for small businesses.

The traditional RAID route is to install a RAID controller and several drives in your company's server rack. Everything is in the same place and near electrical and data connections. The larger the data stash (think hundreds of terabytes), the more economical it can be.

When you've made all the major design considerations and you're ready to buy the RAID gear, have your checkbook handy, because pricing ranges from $500 to $500,000 for RAID systems appropriate for small businesses. Many providers have leasing options that turn a capital cost into a monthly expense.

Pay attention to the warranty. Two or three years is standard, but it is often a good idea to extend it if you have the option.

Finally, do a quick calculation as to whether the device you want is better to buy prepopulated with its drives in place or get them yourself. It only takes an extra minute or two to install the drives.

Many businesses' most precious asset is not their employees, products or property, but the data they use every day to run their operations. Nothing can protect against a data calamity better than RAID gear. It may not be perfect, but it's money well spent that could end up being the equivalent of an insurance policy against disaster.

With a RAID array holding your company's data, chances are you'll never have to apologize about lost records or files, letting you concentrate on running and growing your business.

Sometimes the data gods smile on us and even the worst data problems, like a faulty controller or a dead drive, can be solved by using the other drives to rebuild the lost data. Other times, nothing works.

Your first step here should be to try to recover the data yourself using specialized software like ReclaiMe that can work with typical software-related problems. It's free and potentially able to revive the data on the disk.

Should that fail, it's time to call data recovery experts to examine and extract all the data that is possible. These companies typically take the drive and inspect its data structure and attempt to fix simple software problems, like corrupted files, broken links, incorrect File Allocation Table entries or even a pesky virus.

The best services can also physically repair broken drives. Look for those that have a cleanroom to do the repairs and are authorized to perform the service by the drive's manufacturer.

Data In, Data Out

The prognosis is often surprisingly good, with slews of recoverable data, but it takes time and money – two things that are in short supply with small businesses. When done, the data technician puts the recovered data onto a fresh drive or can make it available via secure FTP site.

What separates good data recovery companies from the crowd? In addition to the ability to recover that last byte of data, a good data recovery firm will work quickly, accurately and securely and only charge you for the files it recovers. More to the point, insist on a certified chain of control of the disks during transit and that the company wipes your data from its system when done.

These are complex, potentially expensive networking devices that need to mesh well with your company and its future. Every minute you spend planning a RAID infrastructure is time well spent that might eliminate hours of anxiety and frustration down the road.

Be proactive, and talk to a nearby recovery firm about your storage needs and plans before setting up a RAID system. This company can give you advice and make sure it can work with your system. If it offers a reasonably priced service plan, take it.

Beneath the surface of RAID is a world of complexity that brings together up to a dozen drives to produce a unified data structure. The major ways of implementing RAID correspond to how safe or fast you want your company's data to be. On the other hand, the more foolproof the system, the more capacity is lost to storing the recovery data.

The simplest method is RAID 0, where the data is spread across two or more drives for faster access. There's also the slower JBOD technique, which stands for "just a bunch of drives." Both lack redundancy or ability to rebuild lost data and are a poor choice for business-critical data.

At the other extreme is RAID 1, where the data on one drive is continually copied to a second drive. This mirroring offers ultimate safety because, if one drive goes bad, you can always use the second drive's identical data. It may be the ultimate in data protection, but RAID 1 cuts the storage system's usable capacity in half.

The RAID Numbers Game

For RAID levels 3 through 6, all the data is striped across the drives along with parity information that can be used to recover a lost drive. RAID 3 and 4 use dedicated drives to store the recovery parity data, creating an inherent vulnerability.

By contrast, a RAID 5 setup requires at least three drives and spreads the recovery data across all the drives for greater fault tolerance. It can rebuild the data of any single drive, even while the array continues to read and write data. By contrast, RAID 6 requires at least four drives but adds faster data-writing performance and the ability to rebuild two lost drives at once. Either of these configurations can help small businesses in search of ultra-reliable data storage.

RAID 10 goes a step further by combining the redundancy of RAID 1 with the speeds of RAID 0. This technique can survive the loss of a single drive but reduces capacity considerably.

Be careful when shopping for RAID gear, because there are also nonstandard RAID techniques that some manufacturers use. For instance, the BeyondRAID protocol is like a cross between RAID 1 and 5 and can accommodate drives of various capacities.

Despite its ability to recover from a dead drive, RAID should not be your company's only line of defense against losing data. Every company – no matter how small – needs to integrate an appropriate backup policy into its operations.

Most microbusinesses don't have a dedicated server rack, not to mention the space, money or need for hundreds of terabytes of capacity to justify it. About the size of four or five books, a stand-alone RAID device can do everything a rack-mounted one can but is self-contained. A big benefit is that a stand-alone array can be put on a table or shelf or under a desk. This solution offers 8TB to 100TB of capacity at prices as low as $700.

Whether it's due to a bad RAID controller, an overheated drive, or physical damage from a fire or flood, the result is the same: inaccessible data. The key with RAID is that the remaining good drives can rebuild and recover the lost one. It works with all sorts of data, including customer records, product designs, all manner of internal documents, audio, backups, surveillance video, and even the most mundane emails and calendar entries.

Plus, the rebuilding process can occur while the storage array remains online and active. It might slow down access while the data is reconstructed on a fresh drive, but it's a small price to pay for getting lost data back.

For security-minded companies, the data can be encrypted to prevent surreptitious use. Some of the newer models feature built-in hardware encryption for scrambling data.

If you pick the details carefully, RAID can save your company's bacon. Rather than dumb storage units, RAID is the equivalent of a sophisticated single-purpose computer connected to several disk drives. Don't stint on how much RAM you get or the speed of the RAID controller's processor, because the result could be slower performance, particularly during rebuilding data.

Most products offer the choice of different RAID levels, and the gear generally includes the necessary software. Focus on devices and apps that can grow with your company as it expands in the future.

Once you know the controller and RAID level you'll be using, figure out how much data space you need and calculate how many drives you'll need. Here, an online calculator can help with estimates for the actual storage you'll get from an array.

You're not done yet. Because your company is likely going to grow quickly, add in an element for the future. The extra cost will seem minor in a year when you actually need the space.

Be aware that most RAID systems work best with drives that are matched on latency, disk speed and especially capacity. In fact, if you use drives of different capacities, most RAID systems will treat them all as if they were the smallest drive of the group, potentially leaving storage capacity on the table. This is not the case with BeyondRAID systems, though.

Digital Ports

At some point, you'll need to consider the system's ports. Sure, it will have an Ethernet connector for setting the RAID device up on your company's network. Some offer a second LAN port to allow a redundant connection or port aggregation that can yield speeds of up to 2 Gbps. Some also add USB and Thunderbolt for use as local storage devices, as might be the case at a company that does a lot of video editing.

It may seem trivial compared to the other considerations, but the device's interface is one of the most important choices you must make. Think of it as your window on a complicated machine that should tell you not only the array's capacity and how much is being used, but also the temperature and health of the drives.

Below the surface, look for the ability to adjust networking parameters, run diagnostic tests and use third-party software – for things like turning the array into a web server or backing up data – that might streamline your company's operations.

A RAID device that logs its activities is a given. Go a step further and look for one that records who accesses, moves and deletes files. This can be a big help with untangling the facts when a file is missing or corrupted.

Hot Spares

While most RAID devices require that you physically replace the bad drive before rebuilding the data on a fresh one, some have space for a standby drive. Called a hot spare, this drive is physically connected to the controller but sits idle most of the time.

In a data emergency, the hot spare is used to rebuild the lost data. In most cases, the changeover can be done from the interface without even touching the array; some manufacturers even make this rollover an automatic event when a drive shows signs of failing. At some point, though, you'll need to remove and replace the dead drive with a good one.

For companies with very sensitive data, encryption is a good way to protect the contents without impeding legitimate uses. It's a good idea to scramble Social Security numbers, customer credit card lists or health insurance information, but it might make sense to encrypt everything that's stored, just in case. Many RAID controllers offer built-in AES 256 encryption, but that can slow down operations.

Each RAID device has its own mix of creature comforts that can make using it easier. These can include LEDs that show activity on each drive or a front-mounted USB port to drive hardware that allows quick swapping of disks. Some even have an LCD screen that shows the system's status, health and temperature.

Finally, while having the physical storage onsite is reassuring, the RAID data can be made accessible to online users, like mobile workers and vendors. Look for a RAID that offers online access so that others can use the data as needed. It can be a slippery slope to a security invasion, so proceed with caution, but today's business demands this type of digital collaboration. Taken a step further, the physical storage and cloud storage can be melded together.

Rather than using traditional rotating hard disks, you can build a RAID array around solid-state flash storage that's faster and generates less heat. All-Flash Arrays, or AFAs, use semiconductor solid-state drive (SSD) chips.

On the downside, AFA systems can easily double the price tag, and each NAND SSD chip they use can be rewritten up to roughly 100,000 times, limiting its active life. There's wear-leveling software that moves the data around to keep one area from being overused, but flash storage might be the wrong choice for a business that continually rewrites existing data.