> value of 1-9 - only 0, 10 or more than 10. (and I think it's
> actually an OS problem on Windows, not a CF issue - can any
> Linux/Unix user confirm the bug on their platform?)
According to the MM guy who explained this problem to me, the underlying
reason for this is that CF simply isn't capable of accurately dealing with
time increments smaller than 10 milliseconds. Of course, this was a while
ago - back in the CF 3 and 4 days - so things may have changed since then.
I've observed the same behavior on Solaris with CF 4 and 5, if I recall
correctly.
> This is exactly the reason that you do may operations in a
> loop when testing this stuff - you need to get things out of
> that initial 10ms "gray area". Basically any timings higher
> than 10 ms are "trustworthy" since it's not the individual
> operations being timed but the interval between the reports.
The problem with this approach is that it only tells you what will happen if
a single request repeats the operation within a loop. There have been
several cases I've observed in which this isn't reflected at all when the
test involves multiple requests repeating the operation in parallel.
Dave Watts, CTO, Fig Leaf Software
http://www.figleaf.com/
phone: 202-797-5496
fax: 202-797-5444
[Todays Threads] [This Message] [Subscription] [Fast Unsubscribe] [User Settings]
