From mboxrd@z Thu Jan  1 00:00:00 1970
References: <87C61423-7C13-4516-88B5-C2ABA7D32AA9@me.com>
	<A1D8C0A6-BB79-42EB-9874-C3F4A8FCC1F6@9srv.net>
	<919294DB-2559-46FB-9800-DB0FBDB63AF6@me.com>
From: "steve@quintile.net" <steve@quintile.net>
Content-Type: text/plain;
	charset=utf-8
In-Reply-To: <919294DB-2559-46FB-9800-DB0FBDB63AF6@me.com>
Message-Id: <59A40CDB-21A4-495B-B4D7-843DEACE0176@quintile.net>
Date: Tue,  5 May 2015 16:46:57 +0100
To: Fans of the OS Plan 9 from Bell Labs <9fans@9fans.net>
Content-Transfer-Encoding: quoted-printable
Mime-Version: 1.0 (1.0)
Subject: Re: [9fans] fossil+venti performance question
Topicbox-Message-UUID: 4f04c546-ead9-11e9-9d60-3106f5b1d025

I too see this, and feel, no proof, that things used to be better. I.e. the f=
irst time I read a file from venti it it very, very slow. subsequent reads f=
rom the ram cache are quick.

I think venti used to be faster a few years ago. maybe another effect of thi=
s is the boot time seems slower than it used to be.

sorry to be vague.

-Steve





> On 5 May 2015, at 15:47, KADOTA Kyohei <lufia@me.com> wrote:
>=20
> Thanks Anthony.
>=20
>> I bet if you re-run the same test twice in a
>> row, you=E2=80=99re going to see dramatically improved
>> performance.
>=20
> I try to re-run =E2=80=98iostats md5sum /386/9pcf=E2=80=99.
> Read result is very fast.
>=20
> first read result is 152KB/s.
> second read result is 232MB/s.
>=20
>> Your write performance in that test isn=E2=80=99t really
>> relevant: they=E2=80=99re not hitting the file system at all.
>=20
> I think to write 1GB data to filesystem:
>=20
>    iostats dd -if /dev/zero -of output -ibs 1024k -obs 1024k -count 1024
>=20
> Write result of dd is 31MB/s.
> But this test may just write to fossil. It may not write to venti.
>=20
>> I=E2=80=99m not sure why you=E2=80=99d see a difference in a
>> fossil+venti setup of a different size, but the
>> partition size relationships, and the in-memory
>> cache size relationships, are what=E2=80=99s mostly important.
>=20
> My hardware has 2GB memory.
> Plan 9 configurations are almost default. (except /dev/sdC0/bloom)
> To increase memory size is difficult,
> because memory size is determined by public QEMU/KVM service plan.
>=20
> =E2=80=94
> kadota