NaibofTabr

M-Disc

M-DISC’s design is intended to provide archival media longevity. M-Disc claims that properly stored M-DISC DVD recordings will last up to 1000 years.

If you want long-term reliability, you don’t want spinning rust.

What does the output from chkdsk #: /X look like? (# is the target volume letter)

If all else fails, SpinRite. If that can’t recover the data, probably nothing can.

Really depends on the system you’re trying to connect them to. Are we talking about a server motherboard? consumer motherboard? old laptop?

You can buy PCIe SAS cards (look for a Host Bus Adapter), but you’ll want to research whether the card you want to use is compatible with your motherboard/CPU and whether or not it has drivers for the OS you’re using before you buy one. These are intended for servers, so they’re very hit-and-miss with consumer hardware.

This is an SAS drive. It requires an SAS controller. This cannot be solved with software.

If you really want to use SAS drives, it can be done with a Host Bus Adapter PCIe card. You will need to be careful about researching whether the HBA you want to buy is compatible with the motherboard you want to use, and that drivers for that HBA exist for the OS that you want to use.

This may not be cheaper than just buying SATA drives. This is server hardware intended for use with server motherboards, there’s no guarantee that it will work properly with a consumer-grade system.

You can turn off the WiFi on your phone so that it uses the cell data connection and either test with the phone itself or tether a computer to it.

You have hundreds of GB of friends’ pictures on your computer? Assuming an average (conservatively high) file size of 10 MB per image, that’s 10,000 pictures per 100 GB.

You probably aren’t going to be able to store your AI training dataset on someone else’s platform for free, at least not in a way that is convenient to access.

Here is a discussion about undervolting an FX-4100 that you may find helpful.

Lowering the voltage will definitely reduce power consumption and heat, and possibly extend the life of the CPU. But at some point a low voltage will cause errors in the CPU operations which could corrupt data or make the system unstable. Lower it in small steps and do some testing. If you plan to underclock and undervolt, only change one at a time.

The one unexpected thing is the drive runtimes were wiped. I could verify by polling the last SMART runs which contained the runtime, and they were in the 4-5 year range.

That seems shady. Are you saying that when you ran a new SMART test the drive reported almost no runtime, but the backlog tests showed the actual runtime?

This is a good point. Surveillance drives are built with the expectation that they will be writing in video data from a security camera system 24/7, so they’re optimized for a constant data write and less so for read, and really not intended for general-purpose random read/write actions. It’ll still work, but not as well as a NAS drive would.

You should anticipate that your storage devices will fail. Drives are consumables. The system that controls the drives should be built for long-term reliability via redundancy. The actual drives should not be expected to have long-term reliability.

Typically these (and the ones on eBay) are used data center drives that have started to show signs of failure, but are not actually failed. Sometimes it’s not even that, they have just reached a certain time in service and been cycled out. For a home user these are a pretty good deal because they’re enterprise quality drives that still have useful life left, and there’s no way you’re going to load them to the level they were designed to handle so they will probably last you several years.

The risk on these failing unexpectedly is higher because you don’t really know how they were used previously. Run SMART tests on any that you buy as soon as you get them. As long as you’re building some kind of redundancy into your array, you should be fine.

SpinRite is probably your best bet for recovery.

This might help you recover the information, but the drive is toast. Get whatever you can off of it and then replace it.

Modern hard drives have some extra sectors that are hidden, and used to replace active sectors as they go bad (built-in redundancy). Once your drive starts reporting more than one or two bad sectors, it’s done. The survivability is gone. Rescue the data as soon as possible and then trash the drive (e-waste recycling, preferably).