Destroying hard drives - waste or necessity?

Not everything is eternal

One of our team members often says, "glass is glass, and glass breaks, and electronics are electronics, and electronics fail". It may sound unprofessional and perhaps a bit silly, but if you think about it - it's a valid thought. Computer hardware is not eternal. Some components will work correctly even after forty years, while others will end up in the trash after two. How does this relate to hard drives?

Let's assume that we use our data storage devices properly. We provide them with adequate cooling, do not abruptly cut off the power supply, and in the case of classic mechanical hard drives, we avoid subjecting them to shocks. Since we have perfect conditions, when can we expect a failure? Manufacturers of storage devices provide an answer to this question, as they have tested it for us.

Each series of hard drives has a defined MTBF parameter. It determines the durability of the devices and is calculated statistically. For example, the Seagate ST1000LM035 hard drive has an MTBF parameter of 1,400,000 hours. This means that if we provide it with perfect operating conditions, it should work without any failures for precisely that time. Will it break afterward? No, but the chance of failure will be very high. Can the drive fail before this time elapses? Of course, and in the case of consumer drives, it is almost certain. Hard drives never operate under ideal conditions. This is especially noticeable in laptops, where temperatures are not ideal, power outages are quite common, and shocks are omnipresent.

And what about the world of SSDs? Here it is very similar: semiconductor storage devices also have an MTBF parameter, but in this case, completely different factors are taken into account. Each SSD consists of memory cells. Each cell can store data, which then can be read from it. Each cell has a limited number of read and write operations - when this limit is exceeded, it "dies" and is replaced with a cell from the backup pool. When the pool is exhausted, the disk is damaged. The user has very limited control over this phenomenon. Will the SSD work after exceeding the aforementioned limit? It may work, but it doesn't have to. Can it fail prematurely? Naturally, such cases occur - usually with cheap, low-quality drives.

That's enough theory, but what does it look like in practice? What do data centers have to do with drive destruction?

Does security policy justify drive destruction?

Data centers should operate continuously and without failures. This means that breakdowns are unacceptable, and administrators of such facilities do everything to minimize the risk of their occurrence. Armed with knowledge from the previous section of the article, we might think: "They probably constantly monitor drive operations, comparing them with the MTBF parameter. They also consider the read and write cycles of SSDs." However, the truth is quite different, and it all comes down to one word - warranty.

Mass storage devices are usually sold with a five-year warranty. Data centers typically withdraw the drives from service after the warranty expires. What happens to the drives that have been "retired"? Drives storing less sensitive data are saved, but it is estimated that 90% of drives are destroyed after the end of their service life. Usually, these drives are fully functional and far from failure. Can't they just be cleaned and reused for other purposes?

Large data centers justify drive destruction by their security policy. However, the main reason is risk management. Such facilities have a zero-risk policy. What does that mean? Simply put, not a shred of information can leak from any drive. For this reason, data centers are obliged to destroy the drives. The irony is that shredders do not guarantee the permanent deletion of data. Recovery specialists can retrieve information even from small pieces of disks. There is an extreme method of data destruction, which is burning and melting the drives, but in addition to data loss, it also ensures the loss of valuable resources.

Given technological progress, there must be a more sophisticated method of data erasure, right?

We can afford more - "humane" data erasure

Alright, so what can be done to safely get rid of data without resorting to barbaric methods of destroying drives? It all depends on whether these drives are intended for reuse within the same company or will be resold to someone else.

The least secure method is to "wipe" the drive. It is often called "formatting." All data appears to be deleted, but it can be recovered using specialized tools. In simplified terms, how does this method work? Imagine a classroom. All students are present and sitting in their places. At this point, the teacher crosses out one of the students from the attendance list. The student is not asked to leave the room - they are merely removed from the documentation. The student remains physically in the class. The same happens with data during formatting. Only entries indicating their existence are removed, and the data itself remains on the drive until overwritten. This method is sufficient if the drive is to be reused within the same company.

However, there is an alternative method that guarantees the permanent deletion of data. It involves overwriting the entire drive with new data, usually random and meaningless. For certainty, the operation can be repeated several times. The advantage of this method is complete data erasure. However, it is time-consuming because erasing a high-capacity drive can take up to two days. Utilizing modern technology can shorten this time.

By comparison, cryptographic erasure takes only a few seconds. Many modern drives have built-in encryption, which means that the data on them can only be read if we have the encryption key. If this key is deleted, all data will be encrypted. It will still be there, but it cannot be read. The drive can be safely resold.

Circular Drive Initiative and Chia Network come to the rescue!

Some time ago, we mentioned the Circular Drive Initiative (CDI) in our website. This project is the result of collaboration between global leaders in the digital data storage, data center, and blockchain technology industries. The project aims to reduce the amount of electronic waste, greenhouse gas emissions, and other factors. To achieve this, the Circular Drives Initiative intends to promote and facilitate the safe reuse of data storage devices. One of the co-founders of this initiative is the Chia Network.

Seagate, a well-known hard drive manufacturer, has also joined the project. According to Amy Zuckerman, the company's Director of Sustainable Development and Transformation, all drives can be restored for use if customers are confident that they can be safely wiped. In the 2022 fiscal year, Seagate renewed and resold 1.16 million hard drives and solid-state drives, avoiding over 540 tons of electronic waste. This applies to drives returned under warranty and drives repurchased from customers. The pilot program in Taiwan allowed the recovery of three tons of e-waste. Zuckerman claims that the challenge now is to scale up the program.

Seagate's renewed drives are tested, recertified, and sold with a five or seven-year warranty. Currently, they are used by small data centers and cryptocurrency farmers. If the Circular Drive Initiative actions achieve a global scale, and the practices promoted within the project are accepted by server communities, all users will benefit, especially Chia farmers.