[Tarantool-patches] [PATCH 1/2] box: make box.cfg idempotent function

[Konstantin Osipov]

* Alexander Turenko <alexander.turenko at tarantool.org> [20/05/06 15:58]:
> On Wed, May 06, 2020 at 02:49:21PM +0300, Konstantin Osipov wrote:
> > * Alexander Turenko <alexander.turenko at tarantool.org> [20/05/06 14:17]:
> > > https://gist.github.com/Totktonada/407855389ed4da93bf0175cf8a11c7b9
> > > 
> > > I made those measurements on fixed CPU frequency and 10 times run each
> > > implementation 10 times: 300 runs at whole. I see that results becomes
> > > worse for all implementations from run to run. Looks strange.
> > 
> > Try with jit off?
> 
> Are there any production setup with jit.off()? This measurement would be
> artificial I think.

Why do you think you can assume that this trace will be jited in production?

> Anyway, I collected the results and they are quite noisy again (~10%
> variation). I would not say anything based on them. But in average ffi
> call gives -0.5% (worse) and ffi call with caching -1.0% from the
> baseline (master). Maybe it is just noise.
> 
> WBR, Alexander Turenko.

-- 
Konstantin Osipov, Moscow, Russia
https://scylladb.com