From: Yanfei Zhang <zhangyanfei@cn.fujitsu.com>
Subject: Re: [PATCH v2 0/5] Export offsets of VMCS fields as note information
 for kdump
Date: Tue, 29 May 2012 15:06:33 +0800
Message-ID: <4FC47579.2040504@cn.fujitsu.com>
References: <4FB35C48.30708@cn.fujitsu.com> <4FB92D5A.3060507@redhat.com> <4FB9A92D.7050108@cn.fujitsu.com> <4FB9FE08.4050905@redhat.com> <4FBA05F6.8070804@cn.fujitsu.com> <4FBA0C8A.2050003@redhat.com> <4FBB0ACA.2040907@cn.fujitsu.com> <4FC30C40.80500@cn.fujitsu.com> <4FC37D94.3080404@redhat.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=GB2312
Content-Transfer-Encoding: QUOTED-PRINTABLE
Cc: mtosatti@redhat.com, ebiederm@xmission.com, luto@mit.edu,
	Joerg Roedel <joerg.roedel@amd.com>, dzickus@redhat.com,
	paul.gortmaker@windriver.com, ludwig.nussel@suse.de,
	linux-kernel@vger.kernel.org, kvm@vger.kernel.org,
	kexec@lists.infradead.org, Greg KH <gregkh@linuxfoundation.org>
To: Avi Kivity <avi@redhat.com>
Return-path: <kvm-owner@vger.kernel.org>
Received: from cn.fujitsu.com ([222.73.24.84]:57549 "EHLO song.cn.fujitsu.com"
	rhost-flags-OK-FAIL-OK-OK) by vger.kernel.org with ESMTP
	id S1752706Ab2E2HJY convert rfc822-to-8bit (ORCPT
	<rfc822;kvm@vger.kernel.org>); Tue, 29 May 2012 03:09:24 -0400
In-Reply-To: <4FC37D94.3080404@redhat.com>
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

=D3=DA 2012=C4=EA05=D4=C228=C8=D5 21:28, Avi Kivity =D0=B4=B5=C0:
> On 05/28/2012 08:25 AM, Yanfei Zhang wrote:
>>
>> Dou you have any comments about this patch set?
>=20
> I still have a hard time understanding why it is needed.  If the host
> crashes, there is no reason to look at guest state; the host should
> survive no matter what the guest does.
>=20
>=20

OK. Let me summarize it.

1. Why is this patch needed? (Our requirement)
  =20
We once came to a buggy situation: a host scheduler bug caused guest ma=
chine's
vcpu stopped for a long time and then led to heartbeat stop (host is st=
ill running).
  =20
we want to have an efficient way to make the bug analysis when we come =
to the similar
situation where guest machine doesn't work well due to something of hos=
t machine's,=20

Because we should debug both host machine's and guest machine's sides t=
o look for
the reasons, so we want to get both host machine's crash dump and guest=
 machine's
crash dump at the same time when the buggy situation remains.=20

2. What will we do?
  =20
If this bug was found on customer's environment, we have two ways to av=
oid
affecting other guest machines running on the same host. First, we coul=
d do bug
analysis on another environment to reproduce the buggy situation; Secon=
d, we
could migrate other guest machines to other hosts.=20

After the buggy situation is reproduced, we panic the host *manually*.
Then we could use userland tools to get guest machine's crash dump from=
 host machine's
with the feature provided by this patch set. Finally we could analyse t=
hem separately
to find which side causes the problem.

Thanks
Zhang Yanfei

  =20