Samsung 850 EVO issues - can't read from drive in any OS

E

Eulogy

2[H]4U
Joined
Nov 9, 2005
Messages
3,124
I have an apparently dead-ish 850 Evo in my hands. This previously, in another system, was in use as the boot drive for a Win10 install. As far as I know it was GPT, and only partitioned however the windows installer would partition it (no manual partitioning).

A few weeks ago, the computer froze. Press and held power button, and got the good ol' screen saying no boot device found. Rebooted, jumped into BIOS, and, the drive doesn't even show there. I figured it was dead and moved on.

Curiosity got to me lately, and I've been wondering what exactly happened. So I plugged the drive into another PC. So it's now plugged into a working PC, which is running Win11. Even in this PC, the drive does not show up in BIOS.

When in Win11, I open disk manager. It sees the disk sitting there just fine. This strikes me as odd, since, BIOS doesn't see it at all. I right click and try to initialize the disk, and get this:
1629182919334.png


Okaayyyy. Let's look at Event Viewer to get a bit more specifics:
1629183018172.png


Well that looks pretty terrible? Let's go see what diskpart might say:
1629183168321.png


Ok, Disk 0 it is. Disk 2 is a different disk (though, also an 850 Evo). Let's take a bit of a dive...
1629183263477.png

Uhhh ok diskpart. GG. Disk, are you like, readonly or some shit?
1629183810572.png


Sooo is my partition table simply F'd? What if I try to create just a new partition and dgaf? What about clean?
1629183408366.png


Nah, it's super unhappy. Ok, well, that's about as far as I can go in Windows land... let's try things in a Linux world. First, what does lsblk show us?
Code: 
NAME   MAJ:MIN RM   SIZE RO TYPE MOUNTPOINT
loop0    7:0    0  55.5M  1 loop /snap/core18/1988
loop1    7:1    0   219M  1 loop /snap/gnome-3-34-1804/66
loop2    7:2    0  64.8M  1 loop /snap/gtk-common-themes/1514
loop3    7:3    0  31.1M  1 loop /snap/snapd/11036
loop4    7:4    0    51M  1 loop /snap/snap-store/518
sda      8:0    0 931.5G  0 disk
├─sda1   8:1    0   512M  0 part /boot/efi
├─sda2   8:2    0     1K  0 part
└─sda5   8:5    0   931G  0 part /
sdb      8:16   0 465.8G  0 disk

Alright, so, my fren here is sdb. Cool. Let's start safe and dd it to an image.
Code: 
sudo dd if=/dev/sdb of=~/rescue/original.img
dd: error reading '/dev/sdb': Input/output error
0+0 records in
0+0 records out
0 bytes copied, 0.170689 s, 0.0 kB/

Nah fren, that's not going anywhere. But why? Here's what dmesg shows:

Code: 
[ 5839.980541] ata6.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
[ 5839.980542] ata6.00: irq_stat 0x40000001
[ 5839.980544] ata6.00: failed command: READ DMA
[ 5839.980547] ata6.00: cmd c8/00:08:00:40:42/00:00:00:00:00/e0 tag 14 dma 4096 in
                        res 51/04:08:00:40:42/00:00:00:00:00/e0 Emask 0x1 (device error)
[ 5839.980548] ata6.00: status: { DRDY ERR }
[ 5839.980549] ata6.00: error: { ABRT }

Other things of note:
- Samsung Magician does not ever see the drive.
- Samsung's bootable thing for flashing firmware does not find the drive (oddly, it doesn't find the fully functional 850 Evo's either, so I suspect this is an issue with the tool itself)
- SeaTools, which I accidentally started in a sleep-deprived state, DOES see the drive and can sucessfully execute S.M.A.R.T. test against it.
- AOMEI tools do see the drive, but also cannot read from it.
- Tried booting a Win10 installer - it too sees the drive, but even if I try a clean install, it fails when trying to create partitions on the disk.

So. I'm out of ideas. I don't really care about whatever is on this drive - at best, I'd pull off some bookmarks that didn't get backed up, but, otherwise, I'm already running on a new SSD in the old PC. Curious if anyone here has any ideas on what I could do to pull data and make this drive usable again though? If it's dead-dead, that's fine. I haven't personally seen an SSD fail in this manner before, usually when they die, they are just totally dead and cannot be seen by any tool at all.
 
heh, I just had the same issue with two of these (250GB variant), a few days apart. I am about half way through the excruciatingly painful Samsung RMA process.
SSD #1: MZ7LN250HMJP
SSD# 2: MZ7LN250HAJQ

Tried registering them on their support website. That worked properly. Clicked the Service Request for RMA button. Filled out all my shipping info, then hit submit and their website errors out and says to call 1-800-SAMSUNG. (I reported to them that their website is broke -- saw other people on forums complaining about it too)

Tried calling 1-800-SAMSUNG. Talked to their robot answering system for about 20 minutes trying to get to the right department. Got hung up on at least 3x because it kept sending me to Cell Phones, HDTV's, and Samsung+ support. Finally I got someone who I made sit on the phone with me through the call transfer and had to wait about 30 minutes on hold with them. I was able to describe the issue to the Solid State Storage support department, and they sent me an email to reply to with all the info on my drives.

Manually typed in all the info, attached invoices, attached photos of the drive label on the back of the SSDs. And that's where I'm at in the process. Still haven't heard back..

There's really no excuse for how terrible their support is since they are one of the largest technology companies in the world.

Yay 5-year warranty? At least I have several more months left in the warranty to try to get them replaced, haha.
 
Anything interesting in smartctl -A /dev/sdb ?

You might try ddrescue to read the drive to try to recover things, it will continue past unreadable blocks and try them again later. If that gets a significant amount of data, there's options to try to make sense of it. Although, I wonder if at this point all reads will fail. It probably won't help, but try forcing a slower sata version too?
 
Ah, I gave up on this in the end. Every I/O op to the drive failed. Another colleague told me about ddrescue - I tried that and it wasn't able to recover a single block. My guess was something in the controller itself just totally shit itself, but likely the storage blocks were fine.
 
Well, got my SSD's sent back. One got returned back to me because they said it works. That's not the case, because it only shows up 1 out of maybe 10 boots, then subsequently causes the computer to freeze. They claim the other one is outside the TBW limit and thus the warranty is void.. again there's no way that's possible because I just used it as a VM backup drive, only writing a 30GB image to it once a month for 3 years (roughly 1TB of writes). It is completely dead, so I have no way to prove to them they're lying.

So, I'm done with buying Samsung SSDs. They didn't honor their warranty for me on two occasions.

PXL_20210928_174859619.jpg
 
shouldnt they be able to provide some CMD line report/screenshot showing the TB and the drive SN...something like when you flash it, it shows the SN or drive info...
 
Apparently they just print off a sheet of paper and check a box, saying their decision is final.
 
w1retap said:
Apparently they just print off a sheet of paper and check a box, saying their decision is final.
Click to expand...
Man, that really blows. I had to deal with Western Digital once on a raptor drive that died. The RMA process was actually not that painful and I got a new drive. I currently have a 1TB 840 evo that Used to be a windows boot drive just for random storage. Based on your situation, it may be my last Samsung drive.
 
kamikazi said:
Man, that really blows. I had to deal with Western Digital once on a raptor drive that died. The RMA process was actually not that painful and I got a new drive. I currently have a 1TB 840 evo that Used to be a windows boot drive just for random storage. Based on your situation, it may be my last Samsung drive.
Click to expand...
WDC and Seagate were both pretty painless at work for my 75+ RMAs that I have done in the 24 years I have been on the job. However I am a university customer. I have not had to RMA a single SSD.
 
drescherjm said:
WDC and Seagate were both pretty painless at work for my 75+ RMAs that I have done in the 24 years I have been on the job. However I am a university customer. I have not had to RMA a single SSD.
Click to expand...
Yeah, I would think SSDs will either work or they won't. I have had two SATA SSDs go bad on me over the years. One was a Crucial and the other was an OCZ. The OCZ was one of the ones that got wide publicity for dying young, but I think mine was quite old when it did. The Crucial was out of warranty by about a year I think. You could install and OS on it, but then you would get all kinds of weird errors. And you could replicate that behavior with fresh installs. I don't think it ever showed a SMART error.
 
w1retap said:
Well, got my SSD's sent back. One got returned back to me because they said it works. That's not the case, because it only shows up 1 out of maybe 10 boots, then subsequently causes the computer to freeze. They claim the other one is outside the TBW limit and thus the warranty is void.. again there's no way that's possible because I just used it as a VM backup drive, only writing a 30GB image to it once a month for 3 years (roughly 1TB of writes). It is completely dead, so I have no way to prove to them they're lying.

So, I'm done with buying Samsung SSDs. They didn't honor their warranty for me on two occasions.
Click to expand...
It could be they have the wrong origination ship date in their system, which means they could have it listed as out of warranty. This happened to me on a few drives, they have incorrect information on a lot of their drives ship/sale dates in the system, or the drive could have sat on the shelf at the reseller for a bit. Give them a call at the number provided and talk to a rep, email them a copy of the receipt showing the actual date you purchased the drive(s) and see if a supervisor can help. I hope this helps.
 
I already gave them all the invoices and dates. They confirmed they were still under warranty and approved the free RMA. The letter said their decision not to honor the warranty is final, so I have no recourse. Garbage company. I have rarely encountered something like this. Not really going to waste more time on this since I have 6 or 7 spares.
 
w1retap said:
I already gave them all the invoices and dates. They confirmed they were still under warranty and approved the free RMA. The letter said their decision not to honor the warranty is final, so I have no recourse. Garbage company. I have rarely encountered something like this. Not really going to waste more time on this since I have 6 or 7 spares.
Click to expand...
Your best recourse is spreading the word about their crappy service, which you are doing.
 
You must log in or register to reply here.
Back
Top