Bruce Momjian wrote:
>
>
>>Btw., I was wondering, are we kidding ourselves when we display
>>microsecond precision to the user? What accuracy do these measurements
>>have in respect to what they are actually supposed to measure?
>>
>>
>
>Not sure. I think the microsecond stuff is accurate to some degree, but
>not sure how much. I think the tick is 10ms, but I thought the timing
>was much more accurate than that.
>
>
>
On my (slow) linux box the apparent time delay between 2 immediately
successive calls to gettimeofday() is 3 or 4 microsecs.
Presumably the clock tick is something much larger than that - I forget
where to look to find that out.
I think the real question is what granularity of reporting will be most
useful. For the session duration patch I restricted it to 1/100th of a
second, and didn't bother rounding the last place there because it
wasn't worth it (I thought). For query duration presumably something
smaller seems appropriate, but I'm not sure how small. I suspect
anything much below a millisec (or at most 100 microsecs) is of
comparatively little value. OTOH truncating it is also probably of
little value - if you can read 3 or 4 decimal places you can read 6 just
as easily - just be aware that the low values have little information value.
cheers
andrew
