Data Verify failure with different vendor SSD

[kenhe]

Hi,

I have tried to run the burn in test, but it shows data verify error message during the test.
This failure symptom still can be reproduced by different vendor M.2 SSD.
Is there any cause from detailed event log can tell why it will fail with different vendor?



Test tool:
BurnInTest V9.1 Pro 1008 (64-bit)

OS version:
Windows 10 Professional Edition build 19042 (64-bit)



**************
RESULT SUMMARY
**************
Test Start time: Fri Dec 4 11:19:34 2020
Test Stop time: Fri Dec 4 11:34:36 2020
Test Duration: 000h 15m 02s

Test Name Cycles Operations Result Errors Last Error
CPU 56 5.515 Trillion PASS 0 No errors
GPGPU 58 171 Trillion PASS 0 No errors
Memory (RAM) 44 810 Billion PASS 0 No errors
2D Graphics 65 33183 PASS 0 No errors
Video Playback 67 861 PASS 0 No errors
Disk (C)130 532 Billion FAIL 1 Data Verify failure
TEST RUN FAILED

******************
DETAILED EVENT LOG
******************
LOG NOTE: 2020-12-04 11:19:36, Status, Main Tests started
LOG NOTE: 2020-12-04 11:19:36, Video, Video test starting: Duty cycle = 75. Test on primary monitor (0). (x,y) = (0,32). WxH = 750x493.
SERIOUS: 2020-12-04 11:25:30, Disk, C: Data Verify failure
LOG NOTE: 2020-12-04 11:25:30, Disk, File: C:\BurnInTest test files\~bittestC00016, Value Read: 190, Value Expected: 80, Offset in file 1208582080, Ops: 215071326208
LOG NOTE: 2020-12-04 11:25:31, Disk, Disk Error (C): Data Verify failure in Cycle 60 with pattern Random data, random seeking
LOG NOTE: 2020-12-04 11:34:35, GPGPU, GPGPU ops for 0: Intel(R) UHD Graphics: 108298.40 billion
LOG NOTE: 2020-12-04 11:34:35, GPGPU, GPGPU ops for 1: NVIDIA Graphics Device: 63560.68 billion
LOG NOTE: 2020-12-04 11:34:51, Status, Test run stopped

[David (PassMark)]

You aren't using the latest release of BurnInTest, but there have been no changes in this test for a few years
https://www.passmark.com/products/bu...st/history.php

If you have results from a few different drives, can you post them. Maybe there is a pattern?

Does it still happen if you turn off all other tests?

[kenhe]

Hi ,

Update the different vendor result as below.
Are you saying only run test disk verify without any other test in burn-in test?
Is there any possible reason to run all test and cause disk verify error?

Vendor one

**************
RESULT SUMMARY
**************
Test Start time: Fri Dec 11 11:19:40 2020
Test Stop time: Fri Dec 11 11:34:42 2020
Test Duration: 000h 15m 02s

Test Name Cycles Operations Result Errors Last Error
CPU 55 5.627 Trillion PASS 0 No errors
Memory (RAM) 105 787 Billion PASS 0 No errors
2D Graphics 65 33317 PASS 0 No errors
Video Playback 67 862 PASS 0 No errors
Disk (C)103 368 Billion FAIL 1 Data Verify failure
TEST RUN FAILED

******************
DETAILED EVENT LOG
******************
LOG NOTE: 2020-12-11 11:19:41, Status, Main Tests started
LOG NOTE: 2020-12-11 11:19:41, Video, Video test starting: Duty cycle = 75. Test on primary monitor (0). (x,y) = (0,32). WxH = 375x230.
SERIOUS: 2020-12-11 11:30:19, Disk, C: Data Verify failure
LOG NOTE: 2020-12-11 11:30:19, Disk, File: C:\BurnInTest test files\~bittestC00013, Value Read: 54, Value Expected: 56, Offset in file 537493440, Ops: 267965693952
LOG NOTE: 2020-12-11 11:30:20, Disk, Disk Error (C): Data Verify failure in Cycle 76 with pattern Random data, random seeking
LOG NOTE: 2020-12-11 11:34:49, Status, Test run stopped


Vendor two

**************
RESULT SUMMARY
**************
Test Start time: Sat Dec 12 14:51:50 2020
Test Stop time: Sat Dec 12 15:06:52 2020
Test Duration: 000h 15m 02s

Test Name Cycles Operations Result Errors Last Error
CPU 55 4.691 Trillion PASS 0 No errors
GPGPU 29 106 Trillion PASS 0 No errors
Memory (RAM) 9 828 Billion PASS 0 No errors
2D Graphics 64 32838 PASS 0 No errors
Video Playback 67 862 PASS 0 No errors
Disk (C) 67 254 Billion FAIL 1 Data Verify failure
TEST RUN FAILED

******************
DETAILED EVENT LOG
******************
LOG NOTE: 2020-12-12 14:51:52, Status, Main Tests started
LOG NOTE: 2020-12-12 14:51:52, Video, Video test starting: Duty cycle = 75. Test on primary monitor (0). (x,y) = (0,32). WxH = 750x493.
SERIOUS: 2020-12-12 15:04:40, Disk, C: Data Verify failure
LOG NOTE: 2020-12-12 15:04:40, Disk, File: C:\BurnInTest test files\~bittestC00007, Value Read: 0, Value Expected: 2863311530, Offset in file 1083703296, Ops: 172314575328
LOG NOTE: 2020-12-12 15:04:40, Disk, Disk Error (C): Data Verify failure in Cycle 42 with pattern Binary data pattern 1 (1010)
LOG NOTE: 2020-12-12 15:06:52, GPGPU, GPGPU ops for 0: Intel(R) UHD Graphics: 106045.96 billion
LOG NOTE: 2020-12-12 15:07:07, Status, Test run stopped

[David (PassMark)]

Results seem pretty random. No obvious pattern.

I would think there is either a hardware failure on the system, or there is some buggy device driver corrupting data.

So does it still happen if you turn off all other tests, except the disk test?

You might want to also try MemTest86 to do a more intensive RAM test.

[kenhe]

Hi David,

Thanks for your recommendation.
i'll try to run the disk test firstly and see whether we can get positive result.

[Nelson]

HI David (PassMark).

After we re-install Memory to system, we can get positive result and no found disk error with 72 hours.
Could you help to share more further information to point out why memory will impact disk error?
if we get it, we will confirm it.

[David (PassMark)]

Nelson,

Is this some new issue, or are you talking about Kenhe's machine?

Does "positive result" mean you found disk errors or didn't find disk errors?

RAM is used for everything. It holds the code that is executed by the CPU, the disk cache and the I/O buffers read and written to the disk. If the RAM doesn't work correctly, then any kind of random behavior is possible.

[Nelson]

HI David (PassMark).

Is this some new issue, or are you talking about Kenhe's machine?--> Yes, I talking about Kenhe's machine.

Does "positive result" mean you found disk errors or didn't find disk errors?--> we cannot find out "Disk errors" after re-plug Memory on system.

RAM is used for everything. It holds the code that is executed by the CPU, the disk cache and the I/O buffers read and written to the disk. If the RAM doesn't work correctly, then any kind of random behavior is possible.--> I am confusing for the result is pass after re-plug Memory on system. so, we can check RAM side firstly, Right?. Thank you very much.

[David (PassMark)]

If you suspect bad RAM, then as suggested above, you might want to also try MemTest86 to do a more intensive RAM test.
Re-seating RAM (if that is all you did) can fix problems as it cleans gunk of the connector, and can sometimes give better electrical contact.