Factory Recertified Drives & The Case For Sustainable Storage

When it comes to lowering your total cost of storage ownership, factory recertified drives make for an attractive, sustainable option. By understanding more about recertified drives, you can help your company lower costs while procuring the storage inventory you need. 

Although tested refurbished drives typically offer the lowest cost-per-terabyte, there are several factors when determining which device is right for your organization. In this article, we’ll explore use cases for factory recertified HDD and how they support sustainable storage solutions. Before diving in, let’s get a better understanding what a factory recertified HDD is and the benefits FR drives provide.

What Are Factory Recertified Drives?

Factory recertified (FR) drives are often confused with used or refurbished drives. While a common misconception, there are important differences.

“Used” or “refurbished” is a broader category which includes drives with many power-on hours (POH). They vary in quality, from seller-tested drives sold by reputable dealers to untested drives sold with misleading information. By contrast, FR drives are returned devices that have limited or zero power-on hours (POH) and which are certified as 100% error free.

As Good As New

With little or no use history, factory recertified drives are often as good as new. For example, a manufacturer like HPE might receive a shipment of drives and find a 4% failure rate within that batch. If the failure rate exceeds its acceptability threshold, the entire shipment might be rejected. 

Since these drives have already been invoiced and assigned specific serial numbers, they cannot be sold as new. The same is true with lightly used devices used in sales displays. Instead, the manufacturer puts them through their test process for new drives, often undergoing more tests than drives fresh off of the assembly line. Drives that pass these rigorous tests are then labeled and resold as “factory recertified”. By contrast, a refurbished drive is a broader category for used drives that have been tested to determine whether they are eligible for reuse.

Cost Advantages of Factory Recertified

Factory recertified hard drives allow IT managers to lean into the cost advantage of HDD over SSD. An FR drive can usually be purchased 20-30% cheaper than a new hard drive. For tested refurbished drives, the savings compared to new HDDs may be 50% or greater.

“In addition to the reduced price, factory recertified drives often come with a 6-month to 1-year warranty from the seller.”

This cost difference is material for large-scale deployments. However, since hard drives are typically in use for five years before being decommissioned, refurbished drives will often be only available in lower capacities. This limitation isn’t normally the case for FR drives, as even batches of leading-edge large capacity drives can be recalled due to defects in a small percentage of units.

	12TB New HDD	12TB FR HDD	12TB Refurbished HDD
JBOD (60 Bay)	$4,600.00	$4,600.00	$4,600.00
Storage	$12,000.00	$8,400.00	$6,000.00
Hardware cost	$16,600.00	$13,000.00	$10,600.00
Total Storage in TB	720	720	720
Cost per TB	$23.05	$18.05	$14.72

Are FR Drives Reliable?

Factory recertified drives are affordable, but are they reliable? Let’s put numbers behind the likelihood of a hard drive failure.

Generally speaking, drive defects are exposed within the first thirty days of operation. This means that if you have a defective drive, you’ll know fairly quickly, and can make use of any warranty agreement that came with the purchase. 

Several drive measurements predict failure at the drive level:

• Mean Time Before Failure (MTBF): Statistical measure of the average number of hours a drive is expected to last before failure, based on the time before failure of similar drives.
• Annualized Failure Rate (AFR): Percentage of hard drives of a given type that are expected to fail in a year. 
• Terabytes Written: This tells us how hard the drive has been worked. Odds of drive failure increase significantly if TBW exceeds certain thresholds set by the manufacturer. 

Perhaps surprisingly, factory recertified drives may even have a lower failure rate than new drives, due to the extensive tests they undergo.

Putting HDD Failure into Context

While a 10% AFR for used drives and even a 2% failure rate for factory recertified drives may seem like a lot, consider it in the context of a risk-tolerant data center setup. Many architectures use parity setups, where data is split into m pieces, and an algorithm is used to generate c “clue” pieces that enable data reconstruction in case of failures. 

For illustration, let’s notate these setups as m/n, where m represents the original data pieces, and n is the total number of pieces (original + clue pieces).

We’ll focus on two different parity setups, 4/9 and 9/11, and calculate how many drive failures the system can tolerate. Since FR drives are as reliable as new drives, we’ll use an AFR of 1.57%, Backblaze’s annualized failure rate for its 301,120 drives across 2024. 4/9 means your data is split into 4 pieces, and 5 extra “clue” pieces are created. You only need any 4 of these 9 total pieces (original data plus clues) to get your data back, meaning you can lose up to 5 pieces without a problem.

By using binomial distributions, we find that the probability of the 4/9 setup failing is only about 0.00000012% in a given year. In practical terms, this means the chances of losing data across your HDD with a 4/9 setup is close to zero. You would need to host hundreds of thousands, if not millions, of drives for the risk of losing data to become a realistic concern.

Crunching the Math  

Our calculation for 4/9 begins with the “probability mass function”, which tells us the odds that the number of drives that fail in a year is exactly k. We add values of this function from k=9 (all drives fail) to k=5 (if one less drive failed, redundancy would preserve data).

For 9/11, we use a similar calculation. In this setup, data is split into 9 pieces, and 2 extra, the “clue” pieces are created. You need any 9 of these 11 total pieces (original data plus clues) to get your data back, meaning you can only lose up to 2 pieces without losing data. Sounds risky, right?

But plugging this into the same formula, the percentage chance of data loss is still only 0.058%, which corresponds to a one in 1,724 likelihood of happening in a given year. In layman’s terms, that means a couple of hundred drives will yield around a 1% annual risk of data loss—not such a failsafe as 4/9, but still a low probability by most measures. 

(Note this that this is simplified calculation for demonstration purposes)

What If My Drive Fails?

Irrespective of drive condition, data integrity relies on the architecture in which the drives are placed and on software options designed to protect data. Thankfully, a bit of forethought means you can sharply reduce the risk of data loss, as we started to outline above.

RAID & Erasure Coding Setups

There are many ways to build redundancy into your data center architecture. The most common way is “Redundant Array of Independent Disks”, or RAID. RAID is actually a collection of architectures, which comes in different “levels.” 

RAID 1 “mirrors” data in one drive onto another drive. This requires twice the amount of disks, so higher levels of RAID use “parity” information, which allows you to reconstruct the contents of failed drives without having all of their contents duplicated. 

This parity information can be consolidated in a dedicated drive, or, more commonly, distributed amongst other data-bearing drives, enabling data rebuilding in case of drive failure (RAID 5 and 6).

RAID (Redundant Array of Independent Disks) is a widely accepted approach. Network Attached Storage (NAS) devices often incorporate built-in RAID capabilities, providing similar data protection. Another option is “erasure coding”, which is a scalable and space-efficient, albeit CPU intensive, option which makes use of some advanced mathematics (specifically, the Reed-Solomon algorithm).

Considering Software-Defined Storage

Software also plays a key role in maintaining data integrity, with many examples of software defined storage (SDS) that utilize parity information or erasure code to secure your data.  

Exploring Bulk HDD Use Cases

Now that we have demonstrated that HDDs in good working order—including factory recertified—are safe to deploy in bulk, let’s explore some use cases, particularly with cost-effectiveness in mind.  

Evaluating storage options almost always boils down to a balance between cost and performance, both of which should be evaluated in terms of what you want to do. 

If performance is the business driver, then SSD is going to be preferable, due to its speed, increased storage density, lower power consumption, and reduced latency. If cost is your primary driver, HDDs maintain a strong advantage in cost per terabyte giving an average 6.6x price advantage over SSD. This makes hard drives ideal for bulk storage applications where performance is less critical than capacity and affordability.

Putting Hard Drives to Work

Here’s a broader look at HDD use cases, including factory recertified and tested used drives

• Data Warehousing: HDDs offer cost-effective storage for large data repositories used for analytics and reporting. While query speed matters, the volume of data often makes HDDs a good choice. 
• Media Storage: HDDs provide cost-effective storage for large media files (videos, images, audio), especially when streaming or real-time access isn’t crucial. 
• Cold Storage: HDDs, especially larger capacity drives, are ideal for rarely accessed data (cold storage). 
• Backup/Disaster Recovery: HDDs are often the core of backup and disaster recovery due to their lower cost per terabyte, though SSDs may be used for rapid recovery of critical systems.
• File Servers: HDDs are cost-effective for general file sharing. SSDs can improve performance, but this isn’t always necessary. 
• Surveillance: HDDs are commonly used in surveillance for their capacity and cost effectiveness. 
• Software-Defined Storage (SDS): As indicated above, HDDs are attractive for building large SDS clusters due to their cost advantage. Recertified or used drives can be incorporated into SDS deployments with planning.

However, contrary to breathless tech coverage about SSDs and HDDs competing for shares of exabytes, the choice between SSDs and HDDs is not always an either-or proposition.

Instead, a hybrid approach is often preferable. One can effectively balance dual pressures of performance and cost by using both technologies in a tiered storage strategy. In such setups, SSDs are typically used for frequently accessed data and applications, while HDDs are used for less frequently accessed data and bulk storage.

HDDs can also leverage parallelism to execute multiple read/write operations simultaneously, helping to offset the speed limitations of mechanical drives. Techniques like RAID 0 distribute data across multiple disks, enabling simultaneous access and boosting throughput. Additionally, operating systems optimize I/O requests to minimize head movement, further improving performance when handling multiple pending requests. While parallelism is effective for random access patterns, excessive use can introduce overhead, potentially negating its benefits.

Factory Recertified and Sustainability Storage

When purchasing hard drives, there is also the sustainability component to consider. Electronic waste, including waste from discarded hard drives, is a growing problem and needs to be addressed—enter the case for drive reuse. 

According to data from the Critical Materials Institute, reuse reduces up to 9x as much emissions as magnet reuse, and up to 900x as much emissions as metal recycling. By extending the lifetime of a HDD by as little as two years, you can significantly reduce a drive’s carbon footprint while delivering tangible cost savings.

Recently, the growing circularity movement has been promoting the reuse of storage. The Circular Drive Initiative provides a great deal of information behind the reuse initiative.  

When an “FR Drive” isn’t an FR drive

When procuring hardware, it’s important to know how to check whether hard drives are used or not. There are some instances when bad actors adjust internal stats to hide usage history before slapping on a new drive label.

Thankfully, there are steps one can take to safely purchase drives in the open market. Starting with knowing your supplier. If buying a factory recertified or refurbished drive, make sure the supplier has the proper certifications. These should include both ISO 9001 and 14001. In addition, the supplier should have either R2, e-Stewards or ADISA. 

Factory recertified drives have limited or zero power-on hours (POH). When it comes to checking POH, SMART stats are not always reliable as a guide, as bad actors can sometimes reset SMART data. Where available, Field Access Reliability Metrics (FARM) are more reliable. Found in many Seagate drives, the FARM logs contain statistical information such as POH, total read and write commands, and other drive metrics.

Other tips:

• A reliable supplier should also be able to provide test documentation for the drives you purchase. 
• For factory recertified hard drives, look for identification on the drive indicating official recertification. 
• Factory recertified hard drive will not have hours and absolutely should test to new drive standards.
• A visual check can also reveal signs of prior usage. Things to look for include misplaced or scratched labels, and signs of wearing around the screw holes. 

Optimizing HDD in Your Storage Mix

The argument for sustainable storage through the use of factory recertified and refurbished hard drives presents a compelling case for cost-effectiveness and environmental responsibility. 

While concerns about data safety may arise, robust hardware and software solutions, including RAID configurations, erasure coding, and software-defined storage all offer multiple layers of protection against data loss, regardless of drive condition. 

Modern drives, whether new, recertified, or refurbished, are remarkably reliable, and failure rates can be further mitigated with appropriate redundancy measures. The key differentiator lies in cost, where factory recertified drives offer significant savings compared to new drives, and refurbished drives provide even greater discounts. 

While the decision to utilize FR or refurbished drives should be made on a case-by-case, they present an attractive option for when cost—rather than performance—is what’s needed most.

Get in touch to find out how our recertified and refurbished drives can reduce the total cost of ownership of your storage hardware.