Question Samsung EVO 850 SSD with corrupt data?

[Mr.CJ]

Dear smart ppl,

I am having some weird issues that I think are related to my Samsung SSD, which contains my win10 OS. The smart thing might be to just get a new one and reinstall windows, but I would like to keep configs and such. Hoping there is a way to fix my issue without a clean install.

It started with the entire computer just choking and ending with blue screen of death. Then on restart, the OS drive would be missing. I would then have to cut the power and restart. Then the disk would be fine and boot like normal. I tried to find out why this was happening. First was curious if it was an MB issue as I read there could be issues with older BIOS and SSDs. Tried updating BIOS, but was unable to. I have an old Asus Ranger VIII. This means it could still be an issue, but as I am unable to resolve it I cannot verify.

I tried some more troubleshooting. I bought a new larger M.2 drive to get all data over there(and MEGA SPEED). Moving data from other drives besides my SSD was working fine. Trying to move some games from SSD to my new drive fails. Steam gives me an error when trying to change the install folder and copying said game manually ends in blue screen of death and the same issues as described.

I tried the Samsung SW for verifying but it keeps saying all is good and it does not support further troubleshooting for the 850. Chkdsk says all is good too. Where should I go from here? Is there a tool for finding the bad sectors and repairing current data, or will this just keep getting worse until I just replace the disk as a whole?

Cheers

 

[logainofhades]

What are the errors for that BSOD? Need full system specs, including make/model of the power supply.

 

[Mr.CJ]

Appreciate your reply

Most recent BSOD:
Stopcode: KERNEL_DATA_INPAGE_ERROR
What failed: win32kbase.sys
I think I've seen variations, but not 100%sure. This is the one I've logged.

MB: Asus Maximus VIII Ranger
CPU: i5-6600K 3,5Ghz
GPU: GTX 1070
OS: WIN 10
OS disk: Samsung EVO 850 SSD 500GB (possible naughty disk)
Disk 2: Kingston SNV2 M.2 2TB
RAM: 16GB DIMM 2133Mhz
PSU: Cooler Master V750 (RS-750-AFBA-G1)

It is an older build, but it has been serving pretty flawlessly till now. Any other specs or further details, just say so

Cheers!

 

[fzabkar]

GSmartControl can display the drive's error logs. Any bad sector should show up there.

https://sourceforge.net/projects/gsmartcontrol/

 

[Mr.CJ]

GSmartControl can display the drive's error logs. Any bad sector should show up there.

https://sourceforge.net/projects/gsmartcontrol/

Thanks for the tip. There seems to be quite a bit of errors, but I do not know how to interpret them, 20774 errors, four showing. It has passed the overall health test though.

 

[fzabkar]

If you upload the report, one of us may be able to make sense of it.

 

[Mr.CJ]

If you upload the report, one of us may be able to make sense of it.

The complete error log of the four listed errors is:
Complete error log:

SMART Extended Comprehensive Error Log Version: 1 (1 sectors)
Device Error Count: 20774 (device log contains only the most recent 4 errors)
CR = Command Register
FEATR = Features Register
COUNT = Count (was: Sector Count) Register
LBA_48 = Upper bytes of LBA High/Mid/Low Registers ] ATA-8
LH = LBA High (was: Cylinder High) Register ] LBA
LM = LBA Mid (was: Cylinder Low) Register ] Register
LL = LBA Low (was: Sector Number) Register ]
DV = Device (was: Device/Head) Register
DC = Device Control Register
ER = Error register
ST = Status register
Powered_Up_Time is measured from power on, and printed as
DDd+hh🇲🇲SS.sss where DD=days, hh=hours, mm=minutes,
SS=sec, and sss=millisec. It "wraps" after 49.710 days.

Error 20774 [1] occurred at disk power-on lifetime: 19059 hours (794 days + 3 hours)
When the command that caused the error occurred, the device was active or idle.

After command completion occurred, registers were:
ER -- ST COUNT LBA_48 LH LM LL DV DC
-- -- -- == -- == == == -- -- -- -- --
00 -- 51 00 01 00 00 00 00 00 10 28 00 Error: at LBA = 0x08000010 = 134217744

Commands leading to the command that caused the error were:
CR FEATR COUNT LBA_48 LH LM LL DV DC Powered_Up_Time Command/Feature_Name
-- == -- == -- == == == -- -- -- -- -- --------------- --------------------
60 01 00 00 08 00 00 23 67 8f 00 28 01 01:02:32.000 READ FPDMA QUEUED
60 01 00 00 00 00 00 23 67 90 00 28 00 01:02:32.000 READ FPDMA QUEUED
60 01 00 00 f8 00 00 23 67 91 00 28 1f 01:02:32.000 READ FPDMA QUEUED
60 01 00 00 f0 00 00 23 67 92 00 28 1e 01:02:32.000 READ FPDMA QUEUED
60 01 00 00 e8 00 00 23 67 93 00 28 1d 01:02:32.000 READ FPDMA QUEUED

Error 20773 [0] occurred at disk power-on lifetime: 19059 hours (794 days + 3 hours)
When the command that caused the error occurred, the device was active or idle.

After command completion occurred, registers were:
ER -- ST COUNT LBA_48 LH LM LL DV DC
-- -- -- == -- == == == -- -- -- -- --
00 -- 51 00 d0 00 00 23 67 91 00 40 00 Error: at LBA = 0x23679100 = 593989888

Commands leading to the command that caused the error were:
CR FEATR COUNT LBA_48 LH LM LL DV DC Powered_Up_Time Command/Feature_Name
-- == -- == -- == == == -- -- -- -- -- --------------- --------------------
60 01 00 00 c8 00 00 23 67 90 00 40 19 01:02:32.000 READ FPDMA QUEUED
60 01 00 00 c0 00 00 23 67 8f 00 40 18 01:02:32.000 READ FPDMA QUEUED
60 01 00 00 b8 00 00 23 67 8e 00 40 17 01:02:32.000 READ FPDMA QUEUED
60 00 08 00 b0 00 00 23 67 8d 80 40 16 01:02:32.000 READ FPDMA QUEUED
60 00 80 00 a8 00 00 23 67 8d 80 40 15 01:02:32.000 READ FPDMA QUEUED

Error 20772 [3] occurred at disk power-on lifetime: 19059 hours (794 days + 3 hours)
When the command that caused the error occurred, the device was active or idle.

After command completion occurred, registers were:
ER -- ST COUNT LBA_48 LH LM LL DV DC
-- -- -- == -- == == == -- -- -- -- --
00 -- 51 00 01 00 00 00 00 00 10 28 00 Error: at LBA = 0x08000010 = 134217744

Commands leading to the command that caused the error were:
CR FEATR COUNT LBA_48 LH LM LL DV DC Powered_Up_Time Command/Feature_Name
-- == -- == -- == == == -- -- -- -- -- --------------- --------------------
61 00 08 00 18 00 00 01 4c 49 38 28 03 01:02:31.000 WRITE FPDMA QUEUED
61 00 08 00 10 00 00 00 6d 3c 30 28 02 01:02:31.000 WRITE FPDMA QUEUED
60 00 f8 00 08 00 00 23 67 93 08 28 01 01:02:31.000 READ FPDMA QUEUED
60 01 00 00 00 00 00 23 67 92 08 28 00 01:02:31.000 READ FPDMA QUEUED
60 01 00 00 f8 00 00 23 67 91 08 28 1f 01:02:31.000 READ FPDMA QUEUED

Error 20771 [2] occurred at disk power-on lifetime: 19059 hours (794 days + 3 hours)
When the command that caused the error occurred, the device was active or idle.

After command completion occurred, registers were:
ER -- ST COUNT LBA_48 LH LM LL DV DC
-- -- -- == -- == == == -- -- -- -- --
00 -- 51 00 01 00 00 00 00 00 10 28 00 Error: at LBA = 0x08000010 = 134217744

Commands leading to the command that caused the error were:
CR FEATR COUNT LBA_48 LH LM LL DV DC Powered_Up_Time Command/Feature_Name
-- == -- == -- == == == -- -- -- -- -- --------------- --------------------
60 01 00 00 a0 00 00 23 67 8f 00 28 14 01:02:31.000 READ FPDMA QUEUED
60 01 00 00 98 00 00 23 67 90 00 28 13 01:02:31.000 READ FPDMA QUEUED
60 01 00 00 90 00 00 23 67 91 00 28 12 01:02:31.000 READ FPDMA QUEUED
60 01 00 00 88 00 00 23 67 92 00 28 11 01:02:31.000 READ FPDMA QUEUED
60 01 00 00 80 00 00 23 67 93 00 28 10 01:02:31.000 READ FPDMA QUEUED

 

[Mr.CJ]

Short self test passed
Extended test completed with read failure after 10%. LBA of the first error"8344112".

 

[Mr.CJ]

Attributes shows this list:
SMART Attributes Data Structure revision number: 1
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME FLAGS VALUE WORST THRESH FAIL RAW_VALUE
5 Reallocated_Sector_Ct PO--CK 096 096 010 - 53
9 Power_On_Hours -O--CK 096 096 000 - 19066
12 Power_Cycle_Count -O--CK 097 097 000 - 2751
177 Wear_Leveling_Count PO--C- 060 060 000 - 841
179 Used_Rsvd_Blk_Cnt_Tot PO--C- 096 096 010 - 53
181 Program_Fail_Cnt_Total -O--CK 100 100 010 - 0
182 Erase_Fail_Count_Total -O--CK 100 100 010 - 0
183 Runtime_Bad_Block PO--C- 096 096 010 - 53
187 Uncorrectable_Error_Cnt -O--CK 097 097 000 - 20630
190 Airflow_Temperature_Cel -O--CK 067 048 000 - 33
195 ECC_Error_Rate -O-RC- 001 001 000 - 20630
199 CRC_Error_Count -OSRCK 100 100 000 - 0
235 POR_Recovery_Count -O--C- 099 099 000 - 89
241 Total_LBAs_Written -O--CK 099 099 000 - 180864352228
||||||_ K auto-keep
|||||__ C event count
||||___ R error rate
|||____ S speed/performance
||_ O updated online
|__ P prefailure warning

 

[Mr.CJ]

Guess it does not look good 😛

 

[fzabkar]

AIUI, there were several READ commands which produced errors at particular LBAs (134217744 and 593989888). However, the Error Register is always 0, so this doesn't tell us anything about the nature of the error. :-?

Attributes 5, 179, 183, 187, 195 are all showing that the drive is affected by bad blocks. However, the Wear Leveling Count of 841 is telling us that the drive has been extensively used.

Total_LBAs_Written would suggest that 92.6TB of data have been written to the drive.

180 864 352 228 x 512 bytes = 92.6 terabytes

 

[Mr.CJ]

AIUI, there were several READ commands which produced errors at particular LBAs (134217744 and 593989888). However, the Error Register is always 0, so this doesn't tell us anything about the nature of the error. :-?

Attributes 5, 179, 183, 187, 195 are all showing that the drive is affected by bad blocks. However, the Wear Leveling Count of 841 is telling us that the drive has been extensively used.

Total_LBAs_Written would suggest that 92.6TB of data have been written to the drive.

180 864 352 228 x 512 bytes = 92.6 terabytes

With a running time far below 20k, isn't that bad for an SSD? Or does old age affect it more?
Also, shouldn't the disk "heal" itself by isolating the bad blocks? I would belive that the games and SW with bad blocks should be able to fix those issues fairly easy instead of choking the system or causing BSOD?

I also see that my CPU is working overtime. Almost at 100% all the time with the main drivers being "Diagnostic policy service" and "Nvidia container". not sure if that means anything.

 

[fzabkar]

I don't know what to make of it.

• Total Host Writes = 92.6 TB
  
  Total NAND writes = 512 GiB x 841 = 462 TB
  
  Write Amplification = 462 / 92.6 = 5.0

The ideal WA is 1, so this would suggest that your usage pattern has worn out the drive 5 times faster than the ideal. I could be wrong, though.

The 850 EVO 500GB is rated for 150 TBW:

https://semiconductor.samsung.com/consumer-storage/internal-ssd/850evo/

 

[Pextaxmx]

I would fresh install the windows on the new drive and retire the 850EVO. I think you can save most of the windows settings by linking your current Windows with your microsoft account.
You can make an image of the 850 EVO and clone it to the new drive as well.

 

[Mr.CJ]

I would fresh install the windows on the new drive and retire the 850EVO. I think you can save most of the windows settings by linking your current Windows with your microsoft account.
You can make an image of the 850 EVO and clone it to the new drive as well.

Yeah, I guess it is time. If I clone the disk, what are the pros and cons? I guess errors will follow over?

 

[Pextaxmx]

Yeah, I guess it is time. If I clone the disk, what are the pros and cons? I guess errors will follow over?

I personally prefer fresh installations over imaging the whole drive. You never know you might have unknowingly installed unwanted spywares, or your OS might have some corrupted system files due to the SSD errors - fresh installation eliminates those concerns.
I would keep the 850 EVO disconnected (without deleting anything) while freshly installing Windows on the new drive. Reconnect the old drive after the installation and copy over all the personal files.
You can keep the old drive in your drawer, with all the data intact, just in case you missed anything. (or in emergency, boot from the old drive)

Windows setting sync
https://support.microsoft.com/en-us...3-279a-329926955708#WindowsVersion=Windows_10

 

[Mr.CJ]

I personally prefer fresh installations over imaging the whole drive. You never know you might have unknowingly installed unwanted spywares, or your OS might have some corrupted system files due to the SSD errors - fresh installation eliminates those concerns.
I would keep the 850 EVO disconnected (without deleting anything) while freshly installing Windows on the new drive. Reconnect the old drive after the installation and copy over all the personal files.
You can keep the old drive in your drawer, with all the data intact, just in case you missed anything. (or in emergency, boot from the old drive)

Windows setting sync
https://support.microsoft.com/en-us...3-279a-329926955708#WindowsVersion=Windows_10

Cheers! Will do