On Tue, 2009-06-16 at 09:08 +1000, NeilBrown wrote:
> On Tue, June 16, 2009 8:53 am, Amit Gud wrote:
> > Trond Myklebust wrote:
> >> It boils down to this: I'm extremely reluctant to get into an arms race
> >> of adding ever more mount options for increasingly finer-tuned attribute
> >> cache timeouts. While such an option may meet _your_ machine room
> >> requirements here and now, it becomes tiresome over the years as
> >> machines gain power, and people first start lobbying for millisecond
> >> timeouts, then microseconds, nanoseconds, pico... in order to catch up.
> >>
> >
> > It only makes sense to make machines more powerful and put that power to
> > optimum use. If keeping up with increased resolution is an issue, you
> > think something like "-o unit=ms acregmin=..." would be better? 'unit'
> > can apply to all the time related options defaulting to seconds.
>
> Wasn't this problem solved hundreds of years ago by the invention
> of the decimal point :-)
>
> -o acregmin=0.0012
>
> Write the code to use as much precision as makes sense given the value
> of HZ - and round up. No new mount options.
That works for me...
Trond