You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@cloudstack.apache.org by Pentium100 <pe...@gmail.com> on 2013/12/11 07:30:43 UTC
All Windows VMs stop unexpectedly - every half hour or so.
Hello,
I have a new problem with Cloudstack. Every 20-30 minutes or so, all
Windows 2008 R2 VMs stop.
The libvirt log says:
2013-12-10 14:40:04.881+0000: 1752: error : qemuMonitorIO:615 : internal
error: End of file from monitor
Which would mean that qemu crashed. However, qemu log says this:
qemu: terminating on signal 15 from pid 1752
2013-12-10 14:40:04.881+0000: shutting down
Which would mean that libvirt killed it (pid 1752 is libvirtd)
The hosts are Centos 6.4 and Ubuntu 13.04, the primary storage is nfs. This
happens no matter which host the VM runs in. Also, there are three Windows
VMs total and they all crash within minutes of each other. For example, VM1
crashed at 14:40:04, VM2 (same host) crashed at 14:40:34 and VM3 (different
host) crashed at 14:35:19
What can be done to prevent this?
Re: All Windows VMs stop unexpectedly - every half hour or so.
Posted by Marty Sweet <ms...@gmail.com>.
Agreed with Chiradeep, it may also be worth checking dmesg and checking the
virtual volumes with the tools provided by qemu.
Marty
On Wed, Dec 11, 2013 at 6:27 PM, Chiradeep Vittal <
Chiradeep.Vittal@citrix.com> wrote:
> Best to ask on the libvirt / KVM lists. I googled 'qemu: terminating on
> signal 15' and came up with several hits.
>
> On 12/11/13 12:39 AM, "Andrija Panic" <an...@gmail.com> wrote:
>
> >I'm having similar behaviour with CentOS 6.5 as VM (Centos 6.5 also as
> >host) - one VM keeps crashing during backup 4AM, whenever the IO is high
> >on
> >the disk/host...but this is only 1 VM, and keeps driving me crazy. I have
> >rebuilt it from scratch, and still after few months it started crashing...
> >
> >I use cloudmonkey to check if VM is running, and start it if not...
> >
> >Problem is I don't know/have any details on WHY this crash happens..
> >
> >What is the way to debug this crashing, from qemu/libvirt anyone ?
> >
> >Thanks
> >
> >
> >On 11 December 2013 07:30, Pentium100 <pe...@gmail.com> wrote:
> >
> >> Hello,
> >>
> >> I have a new problem with Cloudstack. Every 20-30 minutes or so, all
> >> Windows 2008 R2 VMs stop.
> >>
> >> The libvirt log says:
> >> 2013-12-10 14:40:04.881+0000: 1752: error : qemuMonitorIO:615 : internal
> >> error: End of file from monitor
> >>
> >> Which would mean that qemu crashed. However, qemu log says this:
> >> qemu: terminating on signal 15 from pid 1752
> >> 2013-12-10 14:40:04.881+0000: shutting down
> >>
> >> Which would mean that libvirt killed it (pid 1752 is libvirtd)
> >>
> >> The hosts are Centos 6.4 and Ubuntu 13.04, the primary storage is nfs.
> >>This
> >> happens no matter which host the VM runs in. Also, there are three
> >>Windows
> >> VMs total and they all crash within minutes of each other. For example,
> >>VM1
> >> crashed at 14:40:04, VM2 (same host) crashed at 14:40:34 and VM3
> >>(different
> >> host) crashed at 14:35:19
> >>
> >> What can be done to prevent this?
> >>
> >
> >
> >
> >--
> >
> >Andrija Panić
> >--------------------------------------
> > http://admintweets.com
> >--------------------------------------
>
>
Re: All Windows VMs stop unexpectedly - every half hour or so.
Posted by Chiradeep Vittal <Ch...@citrix.com>.
Best to ask on the libvirt / KVM lists. I googled 'qemu: terminating on
signal 15' and came up with several hits.
On 12/11/13 12:39 AM, "Andrija Panic" <an...@gmail.com> wrote:
>I'm having similar behaviour with CentOS 6.5 as VM (Centos 6.5 also as
>host) - one VM keeps crashing during backup 4AM, whenever the IO is high
>on
>the disk/host...but this is only 1 VM, and keeps driving me crazy. I have
>rebuilt it from scratch, and still after few months it started crashing...
>
>I use cloudmonkey to check if VM is running, and start it if not...
>
>Problem is I don't know/have any details on WHY this crash happens..
>
>What is the way to debug this crashing, from qemu/libvirt anyone ?
>
>Thanks
>
>
>On 11 December 2013 07:30, Pentium100 <pe...@gmail.com> wrote:
>
>> Hello,
>>
>> I have a new problem with Cloudstack. Every 20-30 minutes or so, all
>> Windows 2008 R2 VMs stop.
>>
>> The libvirt log says:
>> 2013-12-10 14:40:04.881+0000: 1752: error : qemuMonitorIO:615 : internal
>> error: End of file from monitor
>>
>> Which would mean that qemu crashed. However, qemu log says this:
>> qemu: terminating on signal 15 from pid 1752
>> 2013-12-10 14:40:04.881+0000: shutting down
>>
>> Which would mean that libvirt killed it (pid 1752 is libvirtd)
>>
>> The hosts are Centos 6.4 and Ubuntu 13.04, the primary storage is nfs.
>>This
>> happens no matter which host the VM runs in. Also, there are three
>>Windows
>> VMs total and they all crash within minutes of each other. For example,
>>VM1
>> crashed at 14:40:04, VM2 (same host) crashed at 14:40:34 and VM3
>>(different
>> host) crashed at 14:35:19
>>
>> What can be done to prevent this?
>>
>
>
>
>--
>
>Andrija Panić
>--------------------------------------
> http://admintweets.com
>--------------------------------------
Re: All Windows VMs stop unexpectedly - every half hour or so.
Posted by Andrija Panic <an...@gmail.com>.
I'm having similar behaviour with CentOS 6.5 as VM (Centos 6.5 also as
host) - one VM keeps crashing during backup 4AM, whenever the IO is high on
the disk/host...but this is only 1 VM, and keeps driving me crazy. I have
rebuilt it from scratch, and still after few months it started crashing...
I use cloudmonkey to check if VM is running, and start it if not...
Problem is I don't know/have any details on WHY this crash happens..
What is the way to debug this crashing, from qemu/libvirt anyone ?
Thanks
On 11 December 2013 07:30, Pentium100 <pe...@gmail.com> wrote:
> Hello,
>
> I have a new problem with Cloudstack. Every 20-30 minutes or so, all
> Windows 2008 R2 VMs stop.
>
> The libvirt log says:
> 2013-12-10 14:40:04.881+0000: 1752: error : qemuMonitorIO:615 : internal
> error: End of file from monitor
>
> Which would mean that qemu crashed. However, qemu log says this:
> qemu: terminating on signal 15 from pid 1752
> 2013-12-10 14:40:04.881+0000: shutting down
>
> Which would mean that libvirt killed it (pid 1752 is libvirtd)
>
> The hosts are Centos 6.4 and Ubuntu 13.04, the primary storage is nfs. This
> happens no matter which host the VM runs in. Also, there are three Windows
> VMs total and they all crash within minutes of each other. For example, VM1
> crashed at 14:40:04, VM2 (same host) crashed at 14:40:34 and VM3 (different
> host) crashed at 14:35:19
>
> What can be done to prevent this?
>
--
Andrija Panić
--------------------------------------
http://admintweets.com
--------------------------------------