> 1- But why is it 2000 not 1000 or 3000 ?!!
Well, you can derive those figures by multiplying the result by 0.5 resp 1.5.
Seriously. I am not the inventor of this particular benchmark, but aren't all benchmarks arbitrary? In the sense the number can be anything, as long as you take the same number when comparing things. Yes, the key is, they are for _comparison_!
> 2- Any why is it this text not another ?!!
> "Lorem ipsum dolor sit amet, consectetur adipiscing elit. .."
You can take "aaaaaaaaaaaaaaaaaaaaaaaaaaa..." as long as the character count is the same.
;)
> 3- Why if Writing file performance is 0.5 second is a good performance not a bad performance ? What is the criteria for the good or bad performance ?
Probably the developer has run his benchmark on different servers, good and bad, and defined these ranges empirically.
Note that the numbers are in seconds. So lower the better.
Also note that the developer wants to add those individual time measurements together for an overall figure. So they need to have the same level of magnitude.
Other than that I don't know what was in the mind of the inventor. Did you go through the original thread I've linked earlier? If not sufficient, run the advanced search on this forum:
https://moodle.org/mod/forum/search.php?id=5&forumid=94&subject=benchmark.
If you ask me, I won't look for absolute truth in a benchmark.
;)
BTW, you've completely overlooked my final question:
>> P.S. I see that you had a very similar issue some time back:
write performance issue with OCI FSS. Was it not resolved then?