201–210 of 243 posts
Earlier quoted context omitted.
You don’t need to turn echo back on at the end of your batch file. That line is pointless.
And don't need /e since /mir=/e /purge
I installed 2 x 980 Pro 2Tb in a laptop in Nov 2022. Running a daily Robocopy bat script to backup a folder in C: to D: would freeze a couple of times a week and lock the D: drive. After reboot, a drive check would find no errors and everything would work as normal. I've used the same script for years with no issues. Since the firmware update last week Robocopy has not frozen the drive at all this week.
what does the SMART data for your drive say? I'm morbidly curious how much it reports lifespan remaining for its internal write-wear-leveling system.
SMART data is as follows (both since 2022-11-25)
Primary drive C: 5.1 TBW Model Name, Samsung SSD 980 PRO 2TB Serial Number, S***** Drive Type, NVMe Result,Byte End,Byte Start,Description,Raw Data,Status ,0,0,Critical Warning,0,OK ,2,1,Temperature (K),320,OK ,3,3,Available Spare,100,OK ,4,4,Available Spare Threshold,10,OK ,5,5,Percentage Used,0,OK ,47,32,Data Units Read,6465577,OK ,63,48,Data Units Written,10998930,OK ,79,64,Host Read Commands,150273501,OK ,95,80,Host Write Commands,157439035,OK ,111,96,Controller Busy Time,1083,OK ,127,112,Power Cycles,199,OK ,143,128,Power On Hours,571,OK ,159,144,Unsafe Shutdowns,12,OK ,175,160,Media Errors,0,OK ,191,176,Number of Error Information Log Entries,0,OK ,195,192,Warning Composite Temperature Time,0,OK ,199,196,Critical Composite Temperature Time,0,OK ,201,200,Temperature Sensor 1,320,OK ,203,202,Temperature Sensor 2,328,OK ,205,204,Temperature Sensor 3,0,OK ,207,206,Temperature Sensor 4,0,OK ,209,208,Temperature Sensor 5,0,OK ,211,210,Temperature Sensor 6,0,OK ,213,212,Temperature Sensor 7,0,OK ,215,214,Temperature Sensor 8,0,OK
Secondary drive D: 2.9 TBW Model Name, Samsung SSD 980 PRO 2TB Serial Number, S***** Drive Type, NVMe Result,Byte End,Byte Start,Description,Raw Data,Status ,0,0,Critical Warning,0,OK ,2,1,Temperature (K),320,OK ,3,3,Available Spare,100,OK ,4,4,Available Spare Threshold,10,OK ,5,5,Percentage Used,0,OK ,47,32,Data Units Read,4919136,OK ,63,48,Data Units Written,6128916,OK ,79,64,Host Read Commands,164977799,OK ,95,80,Host Write Commands,94324034,OK ,111,96,Controller Busy Time,78,OK ,127,112,Power Cycles,199,OK ,143,128,Power On Hours,538,OK ,159,144,Unsafe Shutdowns,20,OK ,175,160,Media Errors,0,OK ,191,176,Number of Error Information Log Entries,0,OK ,195,192,Warning Composite Temperature Time,56,OK ,199,196,Critical Composite Temperature Time,0,OK ,201,200,Temperature Sensor 1,320,OK ,203,202,Temperature Sensor 2,323,OK ,205,204,Temperature Sensor 3,0,OK ,207,206,Temperature Sensor 4,0,OK ,209,208,Temperature Sensor 5,0,OK ,211,210,Temperature Sensor 6,0,OK ,213,212,Temperature Sensor 7,0,OK ,215,214,Temperature Sensor 8,0,OK
SSDs are complex beasts :-) I've had an OCZ SSD in the past that also became read-only, in the sense that all changes after shutting down the computer were gone. If you rebooted everything was fine, but as soon as the computer was shut down and the SSD also ran out of power, it converted to the state before turning on the computer. That was so bizarre. Once I had a hefty Windows upgrade installed and it was gone afte…
sounds ideal for a public terminal ;)
Samsung 870 EVO drives were also known to fail early including my 2TB model. https://www.techpowerup.com/forums/threads/samsung-870-evo-b...
> seems to primarily affect drives produced in January/February 2021
That is interesting, I wonder if it's another one of those cases where the supply chain shortages forced them into respinning the boards with some slightly out of spec parts. It's certainly been a major problem for anyone making PCBs.
I've been trying to find a decent endurance NVME in the m.2 form factor for write-heavy applications and it appears that true 2-bit MLC has all but disappeared, replaced by 3-bit TLC and higher (with commensurate loss of endurance) The high endurance SSDs appear to be only available in u.2\u.3\hhhl and god-help-me EDSFF form factors Any suggestions? Micron's 7450 isn't readily available
i recommend samsung pm9a3 versions. not as popular, but are enterprise products and also the endurance is like 3 times 980 pro i believe (please check it, not 100% sure). Been using in my threadrupper workstation with a lot of vms which are put to sleep every day with around .25tb written and read each time the vms are started. keep in mind these are 22110 form factor
Any firmware issue? (https://forums.servethehome.com/index.php?threads/pm9a3-firm...)
Earlier quoted context omitted.
> the firmware update last week Link to specific firmware version please?
The new firmware is version 5B2QGXA7, updated via magician on Windows. I didn't make a note of earlier firmware versions. It's still too soon to know if the ssd freeze will reoccur.
https://thomas.glanzmann.de/samsung/>
The pattern is always the same: I have them configured in a raid 1. Once a month debian does a raid check. During the raid check Debian reads all data from both devices. I get uncorrectable read errors. I no longer use Samsung SSDs and replaced them them with SSDSC2KB076T8, Micron SSDs and KC3000 Kingston NVMes. No failures since then. In 2021 I told a friend of mine about the issue. He also had a 870 EVO, issued a dd if=/dev/sdX of=/dev/null bs=8M and guess what, he got uncorrectable read errors. Due to running them in RAID 1 I caught the issue early and I had no data loss or downtime because the Linux software raid compensated for the bad hardware. However I replaced them in a hurry because I no longer trust Samsung SSDs. As you can see from the smart log they're barely used. Less than 4 months in service and 10 TB written. I also got uncorrectable read errors when evacuating data from the devices.
Earlier quoted context omitted.
I find it highly improbable that you couldn't wear out a 3-level-cell or 4-level-cell consumer grade SSD which is capable of 300-500MB/s writes with a 24x7 automated test script in just a few months. Maybe even just a couple of weeks. The published total DWPD (drive-writes-per-day) endurance on these is not that great. Even assuming a conservative 300MB per second, there's 86400 seconds in one day. That's 25920000 MB…
>Or close to 26TB per day. The samsung 960 Pro 2TB is rated by its manufacturer for a total 1200TB of write endurance lifespan. Right. So around a month and a half. In a world where hardware news drops simultaneous by multiple outlets literally within minutes of news embargoes being lifted. That's a lot of time investment to get result that are boring AF ("we tested them. they work"). Have very little real life consu…
Any outlet interested in journalism rather purely PR can purchase a retail sample on release and publish an endurance report at a later date, however long it takes.
From my experience people buy storage whenever they have a need for something faster or larger, unlike CPUs and GPUs which have peak interest around their release dates. Storage is evergreen in that sense.
> That's a lot of time investment to get result that are boring AF ("we tested them. they work").
What's boring about that? This is extremely valuable information for any perspective buyer. Either way you gain reputation for being a trustworthy outlet that people can rely on for accurate information.
> Have very little real life consumer relevance.
I disagree and I think most consumers would be extremely interested in durability of their storage devices, especially since most of them rely on it in the absence of backups.
> And the manufacturer that sent you the review sample definitely doesn't want to see (focus on edge case negative).
That's hardly relevant. Informing potential customers about extremely serious flaws in the product is quite literally their job - at least if they wish to have any semblance of integrity, trustworthiness, and respect.
Many choose to sell out and simply echo the approved selling points they receive directly from the company, but not every outlet does this and it shouldn't be held up as something that tech journalists should aspire (or be allowed) to do.
Earlier quoted context omitted.
How does this happen? Got any background info?
Poor voltage regulation from the motherboard or power supply could glitch the controller of the drive causing I/O errors or failures.