Re: [PATCH v3 20/21] t/perf: add tests for pack bitmaps

From: Jeff King <peff@peff.net>
Date: Dec 21, 2013, 13:40 UTC
Message-ID: <20131221134017.GD10123@sigill.intra.peff.net>
In-Reply-To: <8761r0ioyo.fsf@linux-1gf2.Speedport_W723_V_Typ_A_1_00_098>

On Sat, Dec 07, 2013 at 05:51:43PM +0100, Thomas Rast wrote:

Show 9 quoted lines

> Jeff King <peff@peff.net> writes:
> 
> > +test_perf 'simulated fetch' '
> > +	have=$(git rev-list HEAD --until=1.week.ago -1) &&
> 
> This will give you HEAD if your GIT_PERF_LARGE_REPO hasn't seen any
> activity lately.  I'd prefer something that always takes a fixed commit,
> e.g. HEAD~1000, keeping the perf test reproducible over time (not over
> changing GIT_PERF_LARGE_REPO, of course).

Good point. I just did a "git pull" on the kernel before my tests, but of course the same test with my 3-week-stale repo would be a lot less interesting.

I tried to figure out what is the right N for HEAD~N to represent a week's worth of commits. It turns out the kernel is really bursty depending on where they are in the cycle. Sometimes N is 50, and sometimes 200. I'm picking 100. It doesn't have to be a week; the main goal was to approximate "here's how it performs if you don't repack for a week", so anything in the ballpark is fine.

-Peff