Many people wonder how to preserve digital data for many years, and some have already lost it. At the same time, most ordinary users are not experts in archival storage of information, and therefore often either do not think about how to organize such storage or draw conclusions from successful but statistically insignificant personal experience that are not always correct.
This article details what you should consider if you are concerned with ensuring the availability of important files years from now, with a focus on ordinary users. So, our goal is to preserve data for as long as possible: decades, and if we are lucky, even 100 years.
Storing data for decades — what does it take?
The first thing to consider if you want to ensure that your current and new data is preserved for a really long time is that you don’t need a specific type of storage device, the one with the longest lifespan.
The main “ingredient” is a preservation strategy that takes into account possible points of failure. A system that you will adhere to and adjust as necessary.
Storage devices are an important part, but not the only factor to consider. If you decide to write something important once to the “most reliable disk” and place it in a secure storage location, you may still find that the data becomes inaccessible after a period of time, and often it will not be 100 or even tens of years.
Possible reasons for failure
When it comes to long-term archival data storage, the list of reasons why you might fail is much longer than in the “save data until tomorrow” scenario.
What can prevent data from being saved? Here are a few examples, broken down into categories.
Physical risks
- Hidden recording failure or “phantom copy.” One of the most unpleasant but quite likely problems, especially with the proliferation of counterfeit flash drives and SSDs on the market. This can also happen with a completely normal drive: due to memory cell or SSD controller failures, or errors in the optical disc drive. The bottom line: the copying process, whether manual or using backup tools, appears to be successful, but in fact the data is either not written at all or is written with integrity violations.
- Degradation of the drive during storage, including without external influences (light, heat, moisture, impact, electric fields). This applies to all types of drives available to the average user: hard drives, SSDs, memory cards and other flash drives, optical discs.
- Loss or destruction of storage devices: theft, destruction of storage space due to external events, loss (or “disposal” by a family member).
- “Silent” data corruption (bit rot). Random changes in bits on the media, possible on any type of storage device. In this case, files may be visible on the storage device and even open, but individual fragments are inaccessible. This is especially dangerous for archives, compressed and encrypted data — for them, damage to even individual bits can lead to the inability to read even the information that has been successfully saved.
Technological risks
- Lack of access to compatible devices that could be used to read the storage device, especially if it is a specific storage device. It is usually easier with more common storage devices: for example, even today, in 2025, you can find a working drive for reading 3.5-inch floppy disks.
- Discontinuation of support for proprietary file formats. Example: if you once created documents using Adobe PageMaker or Adobe/Macromedia Flash, it is still possible to open them, but it is not as easy as it used to be. It has been 20 and 5 years, respectively, since support and development for these formats ended.
External and economic risks
- Finances. For example, today you are able to invest in setting up a storage system. But you cannot be sure that in a few decades, you will not need to invest in reading the data, for example, to purchase replacement equipment that has failed, which is beyond your means at that time.
- Inaccessibility or destruction of data by cloud providers. The latter is unlikely.
But the first is something that could well happen in today’s world and, in some ways, is already happening: from the inability to authorize to the lack of access to certain servers.
Human factor
- Forgotten or lost passwords for encrypted data.
- The absence of a person who knows or understands the contents and value of the storage. Have you organized storage in a “special” way that only you understand, without providing the most accessible instructions possible? The data will be lost if the person who inherits it after you lacks sufficient curiosity and skills.
The list of issues is broader than just longer-term storage: for truly long periods, factors such as language changes, inaccessibility (or very high cost) of electricity, stagnation in semiconductor production, and legislative restrictions that are difficult to predict now may come into play.
These are just a few examples. In reality, there may be more ways in which your strategy could fail, and these should be considered in advance when creating a storage system that takes into account as many possible events as possible that could lead to failure.
Reliability of storage devices for offline data storage
When it comes to data storage, most users focus on storage devices: which ones are more reliable, what risks they are exposed to, and how many years data can be stored on each of them.
A simple answer to these questions can be presented in the form of a table showing data when stored in a state disconnected from the computer:
Hard disk drive (HDD)
Storage life: 3–10 years
Key factors and risks:
Mechanics and magnetism. Degradation of lubricant in the motor, “sticking” of heads to plates (stiction), gradual demagnetization of cells (bit rot).
Recommendation for archiving:
Use with caution. Requires periodic (once every 1-2 years) connection to refresh magnetic fields and check mechanics. Not for “set and forget.”
SSD drive (TLC/QLC)
Storage life: 2–10 years
Key factors and risks:
Charge leakage. The charge in the memory cells “leaks” over time. The rate of leakage depends heavily on the storage temperature (the warmer it is, the faster it leaks) and cell wear.
Recommendation for archiving:
Use with extreme caution. Absolutely not suitable for long-term cold storage. Data may degrade after just a couple of years in less than ideal conditions.
USB flash drive / Memory card
Storage life: 1–5 years
Key factors and risks:
Low-quality flash memory.
Similar to SSD, but typically uses less reliable chips and controllers. High risk of complete data loss without warning.
Recommendation for archiving:
Not recommended. Only suitable for data transfer, not storage.
Optical disc (CD-R, DVD-R)
Storage life: 5–20 years
Key factors and risks:
Degradation of the organic layer. The recording layer made of organic dye is destroyed by light (especially UV), humidity, and temperature fluctuations. The quality of the disc is crucial.
Recommendation for archiving:
Conditionally suitable. Only for less critical data. Requires discs from trusted manufacturers and ideal storage conditions (dark, cool, dry).
Optical disc (BD-R HTL)
Storage life: 15–50+ years
Key factors and risks:
Stability of the inorganic layer. An inorganic phase-change layer that is resistant to light is used. Risks: physical damage, disc delamination (adhesive degradation), oxidation of the reflective layer. Manufacturing quality is a key factor.
Recommendation for archiving:
Recommended. A good and affordable option for long-term archiving. It is worth choosing discs from top manufacturers (Verbatim/MCC, Sony, Panasonic).
Optical disc (M-DISC)
Storage life: 100+ years (claimed up to 1000)
Key factors and risks:
Super-durable inorganic layer. Uses a patented “stone” layer that is physically burned by a laser. Maximum resistance to external factors. Nuance: The greatest advantage and most studied technology is on DVD M-DISC.
BD-R M-DISC is also very reliable, but the difference between it and high-quality BD-R HTL is less significant.
Recommendation for archiving:
Highly recommended. The best option available for a “write and store” strategy for your most valuable data.
Tape storage (LTO)
Retention period: 15–30 years
Key factors and risks:
Magnetic tape stability. Professional standard for archives. Tape is very stable, but requires specific and expensive equipment for reading/writing and compliance with storage conditions.
Recommendation for archiving:
Recommended. The gold standard for corporate and professional archives.
Usually too expensive and complicated for home use.
In reality, it’s more complicated. The numbers in the table are not guarantees, but statistics and probabilities that will be influenced by many factors: from storage conditions and the quality of the drive’s manufacture to the specific production batch and the conditions of its delivery to you.
Personal experience is also irrelevant here: you may find a 15-year-old memory card with data that is still fully readable. However, this does not mean that it is a reliable way to store data. It is just luck compared to today’s manufacturing processes. You should rely on statistics.
Why are we talking about storage in a state disconnected from the computer? The reason is to reduce risks. Yes, a connected SSD suffers less from charge leakage, but the likelihood of losing data due to erroneous actions, malware, or OS failures is higher.
NAS storage devices are not completely secure either: vulnerabilities or other firmware flaws are frequently discovered in them. Automatic synchronization with cloud or network storage? The results of an encryption virus can also be synchronized.
This is great for “operational” archives. But it would be wise to also have completely “offline” copies of your data.
How to reduce the likelihood of information loss
Now let’s talk about the basic principles that will help you not guarantee data preservation, but increase the likelihood of success.
- Never limit yourself to a single copy, as this will most likely lead to failure. There is a well-known rule, a kind of standard: “3-2-1,” which means: 3 copies of data, on 2 different types of media, 1 copy stored in a different geographical location. This is not always possible for home users, but it is worth taking seriously. There are other approaches, such as 4-3-2 or GFS.
- Avoid proprietary and compressed file formats. The former may be difficult to open when support ends or the software becomes unavailable. The latter are more sensitive to data corruption: an uncompressed format can be successfully read when corrupted, albeit with minor losses, while a compressed format may become completely unreadable (however, archives may include additional data for recovery in case of corruption, if you took care of this during archiving).
- Perform regular data audits (checks) and migration as fundamentally new or more reliable/suitable storage devices become available. If the data was originally written to write-once storage devices, do not discard them during migration.
- Consider the factor of bit rot. The only way to make sure that nothing undesirable is happening to the data on the storage device is to check the checksums before and after backup, as well as in the future. Here, it may be useful to study tools such as ExactFile, QuickSFV, TeraCopy, HashCheck, and CrcCheckCopy. The list is exhaustive: just choose what works best for you. There are file systems where the necessary mechanisms are already built in, such as ZFS, but when using them, there is a risk that someone other than you will find the disk with this file system to be faulty due to the lack of widespread support for the FS in the most popular operating systems.
- Keep documentation for the data for yourself in the future or for those for whom the data is potentially intended: in the root of the drive, in the same location where the data is stored on the physical medium, with each copy. It makes sense to describe not only the data itself and its value in the documentation, but also how to access it: what seems obvious today may not be so in 10-20 years.
- Learn how to organize digital data storage. There are various approaches: simple hierarchy, Johny Decimal, PARA, chronological, tagging, and others. Find one that is suitable and understandable not only for yourself, but also for those for whom the data is intended.
- Data that can also be presented in physical form, such as photos or documents, should be stored not only in digital form. This adds reliability, reduces dependence on reading devices, and increases the likelihood that someone will pay attention to them if you are not storing them for yourself. Digital data is fragile and cannot be read with the naked eye.
- For important videos, it is a good idea to record them not only as files, but also in the format of regular Blu-Ray discs for playback on home players. You will be able to find such a player or game console with a working drive (and they can also play videos) even many years later, you will not need to search for ancient software, and built-in error correction algorithms will increase the likelihood of successful playback.
- If you have excess funds and the opportunity to “archive” the reading devices themselves in a dry place, it may be useful.
Additional information
In the first version of this article, written more than 10 years ago, and today in its update, I mentioned optical discs several times — Blu-Ray, M-Disc DVDs, and others. Some may ask: is this really relevant in 2025, when no one uses them anymore?
In my opinion, yes. This type of storage device is not relevant for fast data transfer between people and devices, but it is useful for archival data storage:
- This is one of the most reliable ways to store important data offline that is available to ordinary users, and it can be safely used as a supplement to other types of storage devices, such as HDDs.
- High-quality BD-R HTL and M-Disc discs are still available for purchase: you may have to look on a well-known Chinese marketplace, but they are not unavailable.
- Some companies, notably Sony, are producing corporate storage systems on optical disc cartridges today. With a 50-year data retention guarantee. And you can purchase the same discs from the same production lines (Sony BD-R 128 Gb).
- M-Disc DVD has proven its exceptional reliability in both organizational tests and experiments by ordinary users, which can be found in the Russian-language part of the Internet, for example, on the IXBT forum. There are also M-Disc Blu-Ray discs: despite the larger amount of data that can be recorded on them, tests show that their reliability is close to that of other high-quality BD-R discs and is not unattainable for them.
My recommendation is not to dismiss this type of storage device if you need to create an additional reliable and inexpensive copy of truly important data. But remember: the quality of the disc is of paramount importance.
Conclusion
The idea of storing data for many decades is bold, but it is achievable with the right approach. The main components of this approach are:
- Planning and consistency
- Redundancy of data copies
- Active management
- Open data formats
- Use of different types of storage devices
- Documentation and organization of information
