From mboxrd@z Thu Jan  1 00:00:00 1970
From: bugzilla-daemon@bugzilla.kernel.org
Subject: [Bug 42600] New: Live migration of very large vm get's stuck
Date: Wed, 18 Jan 2012 14:53:43 GMT
Message-ID: <bug-42600-28872@https.bugzilla.kernel.org/>
Mime-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
To: kvm@vger.kernel.org
Return-path: <kvm-owner@vger.kernel.org>
Received: from bugzilla.kernel.org ([198.145.19.204]:52155 "EHLO
	bugzilla.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1757720Ab2AROxn (ORCPT <rfc822;kvm@vger.kernel.org>);
	Wed, 18 Jan 2012 09:53:43 -0500
Received: from bugzilla.kernel.org (localhost [127.0.0.1])
	by bugzilla.kernel.org (8.14.5/8.14.5) with ESMTP id q0IErhCF006897
	for <kvm@vger.kernel.org>; Wed, 18 Jan 2012 14:53:43 GMT
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

https://bugzilla.kernel.org/show_bug.cgi?id=42600

           Summary: Live migration of very large vm get's stuck
           Product: Virtualization
           Version: unspecified
    Kernel Version: 2.6.32-131.6.1.el6.x86_64
          Platform: All
        OS/Version: Linux
              Tree: Mainline
            Status: NEW
          Severity: high
          Priority: P1
         Component: kvm
        AssignedTo: virtualization_kvm@kernel-bugs.osdl.org
        ReportedBy: florian.rustedt@unitedexperts.de
        Regression: No


We have several high load Tomcat servers virtualized on Scientific Linux via
KVM.

Configuration is between 10 and 30 GB of RAM and between 8 and 12 cores per VM.

What i can see so far is, that after round about 20% of migration the progress
get's slower and slower and finally the migration runs into a timeout,
sometimes also with a broken source vm, means stuck and needs to be rebootet :(

Both hypervisor servers are connected via 1 GB interface, so bandwidth should
be fine.
The Hypervisors are Intel modular server blades with 2x6cores and 96GB of RAM,
connected to a shared storage.

>>From theory i would asume two possible explanations:
1. dirty RAM is changing too frequently and migration is transfering slower
than the changes are differing in size
2. Too much CPU's and therefore running threads break down migration
performance

If one of theese explanations is true, this is not a bug, but a problem ;)

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching the assignee of the bug.