From mboxrd@z Thu Jan  1 00:00:00 1970
From: Samuel Thibault <samuel.thibault@eu.citrix.com>
Subject: Re: [SPAM] Re: [SPAM] About the differences between
	Version 3.3 and Version 3.0
Date: Thu, 28 Aug 2008 14:56:32 +0100
Message-ID: <20080828135632.GT4523@implementation.uk.xensource.com>
References: <419869332.17872@bjut.edu.cn>
	<7820E802F3B24CF3B3D1BCBB1F907793@ff76d11953be449>
	<419931538.27194@bjut.edu.cn>
	<6138CCD83C6244738B8327DCAC94CECF@ff76d11953be449>
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable
Return-path: <xen-devel-bounces@lists.xensource.com>
Content-Disposition: inline
In-Reply-To: <6138CCD83C6244738B8327DCAC94CECF@ff76d11953be449>
List-Unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
List-Post: <mailto:xen-devel@lists.xensource.com>
List-Help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-Subscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
Sender: xen-devel-bounces@lists.xensource.com
Errors-To: xen-devel-bounces@lists.xensource.com
To: Fang Yuming <fangyuming@emails.bjut.edu.cn>
Cc: xen-devel@lists.xensource.com
List-Id: xen-devel@lists.xenproject.org

Fang Yuming, le Thu 28 Aug 2008 21:49:29 +0800, a =E9crit :
> >> Do these codes include the algorithm processing disk requests from
> >> different domains?
> >=20
> > No, Xen just lets linux do that, in block/*iosched.c
>=20
> Does this mean that I should modify the dom0(linux) kernel( in block/*i=
osched.c) if I want to do some experiments=20
> about the disk scheduling algorithm of virtual disk scheduling algorith=
m?

Since that's the component that does it, yes.

> >> 3. I found the newest Version is Version 3.3, can anyone tell me the=
 majors differences between=20
> >> Version 3.3 and Version 3.0?
> >=20
> > See the announces of 3.1, 3.2 and 3.3.
> >=20
> >> And is there any change in the part of virtual disk, especially the
> >> scheduling alogithm of different disk requests from domains?
> >=20
> > Again, that's just dom0's Linux matter.
>=20
> From your answers, I know that if I want to do some research on virtual=
 disk schedling algorithm, I should=20
> modify the disk scheduling part of linux kernel and it is no need to ch=
ange other parts of Xen?

Indeed. You may want to make the block backend provide the domid to the
io scheduler for e.q. better fairness, however.

Samuel