When I tested block storage with vdbench for stability, data errors occurred.
Error type counts:
===> Data validation Key miscompare: 8
===> Checksum error on timestamp: 8
===> Logical byte address miscompare: 8
===> SD or FSD name miscompare: 8
===> Bad reserved field contents (mbz): 6
0x000* 0000020b 3a58f000 ........ ........ 20a4a58f 97108634 27eed33e 75b65a40
0x010* 01..0000 39316473 20202020 00000000 51a1f3c1 007fb741 2a30cebd 301d64b1
there are no mismatches in bytes 32-511
This question is magical.
Of the 32 bytes read out, the first 4 bytes do not know the source, and the last 28 bytes are the data in the data set generated by vdbench.According to vdbench's logic, the first 32 bytes are modified after 512 bytes are copied from the data set.It seems that the first 32 bytes of this data have not been modified.
This seems to be related to 'dedupset'. I am confused about the scenario in which this problem will arise.
Have you ever had a similar problem?
*** Moderator action (Timo): removed call out to other members who have not participated arleady in this thread.***