Errors are Ambiguous

[kjeffrisps]

I am using your program to validate solid state drives, but the error outputs do not make any sense. There is no description of what the errors actually are (1 bit wrong, 1 byte wrong, etc.). I am also seeing an abnormally high number of errors. Here is an example of the results of running the "cycle" run on an enterprise-grade intel SSD drive (240 GB) for 24 hours.

Read: 20,775,224 MByte (488.23 MByte/s)
Write: 10,506,396 MByte (240.3 MByte/s)
Errors: 75,202,464

I have also attached a log file for your reference.

stressdisk_enterprise_1.log

[ncw]

The error lines look like this

2019/03/07 10:55:09 0AA00000: 2D, A1 diff 8C

They are address, file1 value, file 2 value and the diff which is file1_value XOR file2_value. This will be a binary number for a single bit error so 01, 02, 04, 08, 10, 20, 40, 80.

The diffs in your log look pretty random. It looks like TST_0000 got corrupted when it was written to the disk as the errors are consistent.

The actual errors aren't that important though, you shouldn't get any - if you do then

• run memtest86 on the machine for 48 hours to check for RAM problems, if this passes then
• try the stressdisk test on another machine if you can, if this fails then
• discard or return the media

[kjeffrisps]

Thanks for the help, I will run this test and update you on whether or not that was the issue. Thanks for creating such a great tool!

[ncw]

Thanks for the help, I will run this test and update you on whether or not that was the issue.

Great

Thanks for creating such a great tool!

You are welcome :-)

I put a version of the above text in the README

[kjeffrisps]

I ran MemTest86 (couldn't run plus because the computer is UEFI boot), and strangely enough it finished all four passes with 0 errors. I'll try it on another machine.