[one-users] Having trouble with VM live migration
Tino Vazquez
tinova at fdi.ucm.es
Thu May 27 05:43:49 PDT 2010
Hi Suffiah,
We can see two problems here, let's tackle them separately:
1) Cold Migration: The problems you are experiencing can be due to the
fact that "apparmor" (enabled by default in Ubuntu) is preventing
virsh from creating the checkpoint file. Please, try disabling it to
see whether this is the problem or not, in case it is you will need to
configure it properly (we can give you a hand with that).
2) Live Migration: The error you are getting with the remote virsh
connection is probably due to the inability to perform a ssh
connection paswordlessly _between_ the nodes with the <oneadmin> user.
Let's see if fixing this solves the problem.
Hope it helps,
-Tino
--
Constantino Vázquez, Grid & Virtualization Technology
Engineer/Researcher: http://www.dsa-research.org/tinova
DSA Research Group: http://dsa-research.org
Globus GridWay Metascheduler: http://www.GridWay.org
OpenNebula Virtual Infrastructure Engine: http://www.OpenNebula.org
On Tue, May 25, 2010 at 9:22 PM, Sufiah Marcellina (Excell Data
Corporation) <a-sumarc at microsoft.com> wrote:
> Here is what I get when I tried to virsh from one hosts to the other:
>
> oneadmin at euca-cloud-node2:~$ virsh --connect qemu+ssh://10.200.41.177/system
> The authenticity of host '10.200.41.177 (10.200.41.177)' can't be established.
> RSA key fingerprint is 6a:d4:21:bc:a7:1e:48:25:ad:93:6f:b0:d7:c6:53:44.
> Are you sure you want to continue connecting (yes/no)? yes
> Welcome to virsh, the virtualization interactive terminal.
>
> Type: 'help' for help with commands
> 'quit' to quit
>
> virsh # list
> Id Name State
> ----------------------------------
> 1 one-16 running
> 2 one-20 running
>
>
> It keeps on mentioning about "The auntenticity of host "x.x.x.x" can't be established, even after I say "yes" a few times. Do you think this is the problem?
> Is there any configuration that I missed?
>
>
> Regards,
> Sufiah
>
>
> -----Original Message-----
> From: Sufiah Marcellina (Excell Data Corporation)
> Sent: Tuesday, May 25, 2010 10:39 AM
> To: 'Tino Vazquez'
> Cc: Users at lists.opennebula.org
> Subject: RE: [one-users] Having trouble with VM live migration
>
> Hi,
>
> I've moved one to fixed a few things on OpenNebula, now I have shared filesystem, $VAR_DIR=/var/lib/one/vmdir.
> I am running KVM.
>
> Here are the details:
> oneadmin at OpenNebula:~/images$ onehost list
> HID NAME RVM TCPU FCPU ACPU TMEM FMEM STAT
> 3 10.200.41.208 4 800 797 400 8192660 7851128 on
> 4 10.200.41.177 4 800 797 400 8192660 7685076 on
>
> oneadmin at OpenNebula:~/images$ onevm list
> ID NAME STAT CPU MEM HOSTNAME TIME
> 16 vm-01 runn 0 524288 10.200.41.177 00 14:58:59
> 20 vm-01 runn 0 524288 10.200.41.177 00 00:05:39
> 21 vm-01 runn 0 524288 10.200.41.208 00 00:05:22
>
> On hypervisor HID=3
> virsh # list
> Id Name State
> ----------------------------------
> 1 one-21 running
>
> On hypervisor HID=4
> virsh # list
> Id Name State
> ----------------------------------
> 1 one-16 running
> 2 one-20 running
>
> Now, here is what I get when I try to do live migration:
> Tue May 25 07:34:40 2010 [LCM][I]: New VM state is MIGRATE Tue May 25 07:34:41 2010 [VMM][I]: error: unable to connect to libvirtd at '10.200.41.177': Connection refused Tue May 25 07:34:41 2010 [VMM][I]: ExitCode: 1 Tue May 25 07:34:41 2010 [VMM][E]: Error live-migrating VM, unable to connect to libvirtd at '10.200.41.177': Connection refused Tue May 25 07:34:41 2010 [LCM][I]: Fail to life migrate VM. Assuming that the VM is still RUNNING (will poll VM).
> Tue May 25 07:34:41 2010 [VMM][I]: ExitCode: 0
>
> I get similar error when I try to live migrate from hypervisor 4 to 3.
>
>
> Regards,
> Sufiah
>
>
> -----Original Message-----
> From: tinova79 at gmail.com [mailto:tinova79 at gmail.com] On Behalf Of Tino Vazquez
> Sent: Tuesday, May 25, 2010 2:40 AM
> To: Sufiah Marcellina (Excell Data Corporation)
> Cc: Users at lists.opennebula.org
> Subject: Re: [one-users] Having trouble with VM live migration
>
> Hi Sufiah,
>
> Which hypervisor are you using?
>
> I am assuming you use some kind of shared filesystem to perform at least the live migration.
>
> It would be nice to know where the VM is really running. For that, and assuming you are using:
>
> 1) KVM -> perform a "virsh list"
> 2) XEN -> perform a "xm list"
>
> in both the source and the destination host, let's see where the "one-9" VM is running.
>
> Regards,
>
> -Tino
>
> --
> Constantino Vázquez, Grid & Virtualization Technology
> Engineer/Researcher: http://www.dsa-research.org/tinova
> DSA Research Group: http://dsa-research.org Globus GridWay Metascheduler: http://www.GridWay.org OpenNebula Virtual Infrastructure Engine: http://www.OpenNebula.org
>
>
>
> On Tue, May 25, 2010 at 12:03 AM, Sufiah Marcellina (Excell Data
> Corporation) <a-sumarc at microsoft.com> wrote:
>> Hi,
>>
>>
>>
>> I'm new to OpenNebula and have been trying to get livemigration works.
>>
>> I've install OpenNebula on Ubuntu 10.04, I can add hosts and VMs just
>> fine, the trouble comes when I try to migrate (cold) and live migrate
>> my VMs, it stuck in state "migr", here is what's in the log:
>>
>>
>>
>> oneadmin at OpenNebula:~/images$ onevm list
>>
>> ID NAME STAT CPU MEM HOSTNAME TIME
>>
>> 8 vm-01 save 0 524288 10.200.41.208 00 21:10:05
>>
>> 9 vm-01 migr 0 524288 10.200.41.208 00 21:10:04
>>
>>
>>
>> /var/log/one/9.log:
>>
>> Mon May 24 08:16:55 2010 [VMM][I]: ExitCode: 0
>>
>> Mon May 24 08:16:55 2010 [VMM][I]: Monitor Information:
>>
>> CPU : -1
>>
>> Memory: 524288
>>
>> Net_TX: -1
>>
>> Net_RX: -1
>>
>> Mon May 24 08:29:36 2010 [LCM][I]: New VM state is MIGRATE
>>
>>
>>
>> Please help.
>>
>> _______________________________________________
>> Users mailing list
>> Users at lists.opennebula.org
>> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>>
>>
>
>
More information about the Users
mailing list