7.1 vs 7.3 bug?

[electric_memory_ram]

If you are seeing this a lot, maybe try another model flash drive. Some might be better or worse than others in caching data on the stick (just speculation, it isn't something we have tested).

I don't think they cache anything, all cache should be in the OS, still I'll keep an eye out for varying behavior on other flash drives.

Noticed something new, the size of the log file seems to have a major impact on log write speed. ~24 hours worth of log seems to add major delays on log updates. Not at all sure why that would be, but it seems software rather than hardware.

[David (PassMark)]

Noticed something new, the size of the log file seems to have a major impact on log write speed

It shouldn't. This isn't the way file systems work. Appending data to the end of a large file should be roughly the same speed as appending to a small file.
So I still think something else is going on. (e.g. hardware issue with the flash drive)

[electric_memory_ram]

It shouldn't. This isn't the way file systems work. Appending data to the end of a large file should be roughly the same speed as appending to a small file.
So I still think something else is going on. (e.g. hardware issue with the flash drive)

That's what I would have thought, but a simple elimination of the log file seems to fix it ever time. I ran some performance tests on the flash drive (not super fast, but not terrible), and also all data written to the flash drive came back without change.

Not as an official recommendation, but what flash drive brand and model do you use most often when testing/developing memtest?

[David (PassMark)]

We have a bunch of Kingston USB 2.0 drives which seem to work fine. But my preferred unit is the Corsair Voyager GT 3.0.