ℒӱḏɩę 💾☮∞🎶♲☀🔋It would seem that some of my ancient 10+ year old SSDs in the 60-128gb range are experiencing sudden death after sitting around for years. They're basically bricked and can't be written to.
EDIT: This should be a PSA. SSDs are not for cold storage.
🇺🇦 haxadecimal@Lydie
SSDs are rated to retain data unpowered for six months for consumer grade devices, and for three months for enterprise grade. They need to stay powered up so that they can scrub (refresh) memory cells that lose their charge over time. This is because each data bit is actually stored by a surprisingly small number of electrons in modern NAND flash devices.
Spinning drives, despite their potential mechanical failure modes, are actually better for cold storage than SSDs.
SSDs are rated to retain data unpowered for six months for consumer grade devices, and for three months for enterprise grade. They need to stay powered up so that they can scrub (refresh) memory cells that lose their charge over time. This is because each data bit is actually stored by a surprisingly small number of electrons in modern NAND flash devices.
Spinning drives, despite their potential mechanical failure modes, are actually better for cold storage than SSDs.
skategoat 🐐 🇵🇸@brouhaha I had no idea D: that kinda scares me!
@felipe
In practice, if not stored at high temperature, they'll probably hold data a lot longer than those guaranteed minimums, but you shouldn't expect five years or more.
If you do use an SSD for offline data storage, you should probably power it up periodically, and do something that forces a read of the entire device. In Linux or macOS, something like "dd bs=1M if=/dev/nvme0n1 of=/dev/null" (substitute correct device name) would be suitable. I'm not sure of an equivalent for Windows.
In practice, if not stored at high temperature, they'll probably hold data a lot longer than those guaranteed minimums, but you shouldn't expect five years or more.
If you do use an SSD for offline data storage, you should probably power it up periodically, and do something that forces a read of the entire device. In Linux or macOS, something like "dd bs=1M if=/dev/nvme0n1 of=/dev/null" (substitute correct device name) would be suitable. I'm not sure of an equivalent for Windows.
@Lydie @felipe
and I know that Windows does assign internal pathnames to raw disk devices, but I don't know how one finds them. It's not just e.g. "C:", but rather something like "\\.\PhysicalDrive0". There are published C code examples to enumerate the physical drives from a program. There's probably a way to do it in Powershell, but I haven't found it.
and I know that Windows does assign internal pathnames to raw disk devices, but I don't know how one finds them. It's not just e.g. "C:", but rather something like "\\.\PhysicalDrive0". There are published C code examples to enumerate the physical drives from a program. There's probably a way to do it in Powershell, but I haven't found it.
sijmen
Roger Lipscombe@rogerlipscombe @sijmen @felipe @Lydie
Is there a trivial translation, given a wmic command, to the Powershell equivalent?
Is there a trivial translation, given a wmic command, to the Powershell equivalent?
@brouhaha @sijmen @felipe @Lydie dunno; it's been years since I last looked at this stuff.
Maybe start here? https://powershell.one/wmi/commands#querying-information
Jernej Simončič �
penguin42@penguin42 @felipe
As I understand it, the drives from reputable brands do some background scrub process that does over time read all of the blocks, do error correction, and if necessary, reallocation. Since the manufacturers consider such details proprietary, I agree that a periodic read of the entire device is a good idea. I do that about once a month, scheduled for the wee hours of the morning.
As I understand it, the drives from reputable brands do some background scrub process that does over time read all of the blocks, do error correction, and if necessary, reallocation. Since the manufacturers consider such details proprietary, I agree that a periodic read of the entire device is a good idea. I do that about once a month, scheduled for the wee hours of the morning.
@brouhaha @penguin42 @felipe I remember the early Samsung TLC drives (840 IIRC) had a problem where reading rarely accessed data would become really slow; this was solved in a firmware update, but you also had to run a manual refresh when updating. Shouldn't be a problem with newer drives.
Infoseepage@brouhaha @felipe There is an easy to use windows tool called DiskFresh which is free and will do a full disk read or read+write refresh of a disk.
https://www.puransoftware.com/DiskFresh.html
I've use it primarily on older hard disks to refresh the surfaces and take a look at SMART logs afterwards to look for signs the drive is going south. The software will also report bad sectors and such that it encounters during a refresh operation.
tom jennings@tomjennings @Lydie
As I just wrote in another reply, that's the rated minimum over the full temperature range, so in reality it generally isn't quite that bad. It's sort of the storage industry's deep dark secret. It's not actually a secret, but the manufacturers definitely do not want to call attention to it.
NOR flash has historically been rated for decades of retention, so many of us incorrectly assumed that ultra-high-density NAND would be similar, but unfortunately it isn't.
As I just wrote in another reply, that's the rated minimum over the full temperature range, so in reality it generally isn't quite that bad. It's sort of the storage industry's deep dark secret. It's not actually a secret, but the manufacturers definitely do not want to call attention to it.
NOR flash has historically been rated for decades of retention, so many of us incorrectly assumed that ultra-high-density NAND would be similar, but unfortunately it isn't.
F4GRX Sébastien@tomjennings @brouhaha @Lydie yes it's sad, awful and infuriating. Fortunately all my data and backup drives are spinning rust.
The Penguin of Evil@tomjennings @brouhaha @Lydie a lot of random other flash just degrades too. USB sticks, SD cards, even on device firmware slowly dribbles it's brains out.
Dave Fischer@etchedpixels @tomjennings @brouhaha @Lydie We've been discovering at the museum that it's often harder to get old computers from the 90s running than stuff from the 60s or 70s, because of all the batteries and nvram everywhere.
@davefischer @etchedpixels @tomjennings @brouhaha varta cancer of the 90s 😭. 80s stuff is usually OK too
Too, densities were low (components and recorded data) and lots of standardized parts that appear in catalogs were used. Ditto old automobiles; my early 60's Rambler have *complete parts catalogs*!
By the 80's, quantities were up enough such that custom ASICS and masked ROMs and such became common. Now everything is a brick.
We did demand this...
sen@davefischer this reminds me of 1920s steam locomotives (big levers, pins, valves) vs. 1980s locomotives (largely undocumented PLCs) in preservation, too.
@davefischer @etchedpixels @tomjennings @Lydie
I think it will be possible for CHM to have their restored IBM 1401 and DEC PDP-1 computers, introduced in 1959 and 1960, respectively, still running 50 years from now, with some components replaced, but not fully a "Computer of Theseus."
For computers from the 1990s and newer, that's not at all likely.
I think it will be possible for CHM to have their restored IBM 1401 and DEC PDP-1 computers, introduced in 1959 and 1960, respectively, still running 50 years from now, with some components replaced, but not fully a "Computer of Theseus."
For computers from the 1990s and newer, that's not at all likely.
Right! My daily drivers were all over 50 years until this year.
There will not be 50 year old priuses. Or 50 year old 21st C cars, at all. Too much shit plastic and bespoke parts.
@tomjennings @davefischer @etchedpixels @Lydie
The oldest car I've owned was a 1979, which originally belonged to my grandparents. Unfortunately I totalled it. Currently I have 1992, 2004, and 2021. That's one more that I can actually justify. The 2021 is my daily driver and the 2004 is a 4WD SUV for hauling stuff and for bad road conditions.
The oldest car I've owned was a 1979, which originally belonged to my grandparents. Unfortunately I totalled it. Currently I have 1992, 2004, and 2021. That's one more that I can actually justify. The 2021 is my daily driver and the 2004 is a 4WD SUV for hauling stuff and for bad road conditions.
Bob Collins@davefischer @etchedpixels @tomjennings @brouhaha @Lydie
I expect the primary problem of getting middle aged computers working is the lack of availability of the NVROM code.
@BobCollins @davefischer @etchedpixels @tomjennings @Lydie
In the short term. In the longer term, everything with ICs beyond SSI/MSI will become impossible to repair short of fabricating new emulation devices. I expect basic silicon transistors (but not germanium small-signal transistors) and passive components to still be available for a long time.
In the short term. In the longer term, everything with ICs beyond SSI/MSI will become impossible to repair short of fabricating new emulation devices. I expect basic silicon transistors (but not germanium small-signal transistors) and passive components to still be available for a long time.
@BobCollins @davefischer @etchedpixels @tomjennings @Lydie
In logic circuits, often germanium transistors can be replaced with silicon. Not as authentic as one might prefer, but still much more authentic than replacing an LSI chip with a new FPGA.
#ComputerOfTheseus
In logic circuits, often germanium transistors can be replaced with silicon. Not as authentic as one might prefer, but still much more authentic than replacing an LSI chip with a new FPGA.
#ComputerOfTheseus
@BobCollins @davefischer @etchedpixels @tomjennings @Lydie
More VLSI chips have internal firmware in floating-gate EPROM or flash than you might expect. That includes chips that seem on the surface to have simple functionality. Those chips will fail much sooner than general chip failure due to e.g. metal or ion migration.
More VLSI chips have internal firmware in floating-gate EPROM or flash than you might expect. That includes chips that seem on the surface to have simple functionality. Those chips will fail much sooner than general chip failure due to e.g. metal or ion migration.
Andres@etchedpixels @tomjennings @brouhaha @Lydie petition to make "dribbles its brains out" an industry standard term for bitrot of unused devices
I hadn't considered usb sticks; I've always thought of those as short term transport but I bet, post your post, lots of people don't think that.
Norman Wilson@brouhaha Thank you, Doctor Memory.
Alex Kahn
Savely
Jérôme Petazzoni@brouhaha whoa! What are the keywords to look up to learn more about this? (I tried "nand scrub" and things like that, but it didn't yield much, except a stack overflow answer saying that leaving ssds unpowered for extended periods of time can actually help fix issues caused by trapped electrons, so now I'm even more confused 😕)
@jpetazzo
Manufacturers are generally not willing to publish any information on their error correction and scrub algorithms, which is why just applying power periodically isn't even enough, you have to read all the data as well, to guarantee that the firmware does error correction.
Some sources claim 12 month retention for consumer drives, but still only 3 months for enterprise.
I've never heard of any SSD "trapped electron" problem. The PURPOSE of a flash memory cell is to trap electrons.
Manufacturers are generally not willing to publish any information on their error correction and scrub algorithms, which is why just applying power periodically isn't even enough, you have to read all the data as well, to guarantee that the firmware does error correction.
Some sources claim 12 month retention for consumer drives, but still only 3 months for enterprise.
I've never heard of any SSD "trapped electron" problem. The PURPOSE of a flash memory cell is to trap electrons.
Eric Brombaugh@brouhaha True! When law enforcement confiscates computers as evidence, they image the SSDs onto mechanical drives to ensure that the data is not lost in the interval between when the computers are seized and when they are examined for trial (which can be years!).
malte
Lyskar@brouhaha @Lydie Very true. My hard drives from 20+ years ago still fire up fine as long as one turns them on once a year to avoid stiction (the motor getting jammed due to the lubricant solidifying). Meanwhile, SSDs... well I haven't experienced it yet, but I definitely don't trust them for cold storage.
Jonathan Hendry@jonhendry @Lydie
It would be best if it actually did a full drive read every so often, since we don't actually know any details about scrub operations. That way it could also provide status, e.g., drive 3 all blocks read successfully, or even just green and red LEDs.
It would be best if it actually did a full drive read every so often, since we don't actually know any details about scrub operations. That way it could also provide status, e.g., drive 3 all blocks read successfully, or even just green and red LEDs.
Tim Ward ⭐🇪🇺🔶 #FBPE@TimWardCam @brouhaha It certainly wouldn't hurt, and having things updated more often is a good side effect
highvoltage@brouhaha @Lydie I'm kind of amazed how long some older SSDs retain their memory. I have some early SSDs that haven't been turned on in 10-20 years and they still have all their data. And from my understanding after 10 years of not having them powered up, you shouldn't be surprised if they are dead.
@highvoltage @brouhaha SLC NAND is pretty darn resilient!!