svn commit: r254986 - head/sys/ufs/ufs

[Davide Italiano]

On Wed, Aug 28, 2013 at 12:31 PM, Ivan Voras <ivoras at freebsd.org> wrote:
> On 28 August 2013 12:25, Davide Italiano <davide at freebsd.org> wrote:
>
>> do you have any evidence that this change impacts positively (or
>> negatively) performances for some workloads? If yes, can you share?
>
> Yes, observation of my own servers. Without this, dirhash is basically
> useless since in certain situations everything gets evicted after 5
> seconds and it never grows to its full potential.
>

Oh, well. Your servers doesn't necessary reflect FreeBSD general user
so you can just maintain this change local.
I've always thought FreeBSD policy was that of leaving reasonable
defaults for the general case and giving people possibility to tune
them according to their needs. Also 'in certain situations' makes
relatively little sense to me unless you provide a testcase that
mimics such situations or explain to people how to reproduce it.

>> Also, why did you choose the '60' value (rather than something else)?
>
> Personal experience.
>

Do you realize that this is a driven by commit change, right? And
there's no technical motivatiion about this or experimental data that
confirms you've chossen the right value?


>> I don't see any 'Reviewed by:' line in your commit message neither I
>> remember a public discussion on -current or -arch or -fs about this.
>> OTOH I think such changes deserve a wider discussion.
>
> See discussion in @stable.

I see from the archives that you didn't get any feedback by anyone
working in the VM layer or by some UFS maintainer.
I don't even comment about discussing -CURRENT changes in stable at .

Thanks,

-- 
Davide

"There are no solved problems; there are only problems that are more
or less solved" -- Henri Poincare