user: time jump in slave machine leads to freeze

user: time jump in slave machine leads to freeze

[gnafou]

Hello

We have had several cases where  a slave machine freezes, eating all available 
cpu. ( this happends randomly, say, after 3 months of correct functionning ) 

After reboot, looking at the syslog when the freeze occured,  the few ( ~5 ) 
last lines written show a date which has jumped in the future ( roughly, 15 days 
ahead  ), but nothing related to the crash is logged.

our salves machines are ntp-synchronized 


we run under debian/lenny 
we launch the machines with the kvm command

kernel of host :  2.6.26-2-amd64  
kernel of host :  2.6.26-2-686

command launched :  

kvm -name svn -drive file=xxx -net tap -m 256 -net nic,macaddr=xxx  -pidfile 
/var/run/kvm/xx.pid -daemonize -k fr -vnc :63006 -monitor 
unix:xx_monitor,server,nowait -vnc unix:xx_vnc 


If you ever have an idea  how to solve or to debug the problem ...

Thanks


Fred


      

Re: user: time jump in slave machine leads to freeze

[Brian Jackson]

On 1/25/2011 9:21 AM, gnafou wrote:
> Hello
>
> We have had several cases where  a slave machine freezes, eating all available
> cpu. ( this happends randomly, say, after 3 months of correct functionning )
>
> After reboot, looking at the syslog when the freeze occured,  the few ( ~5 )
> last lines written show a date which has jumped in the future ( roughly, 15 days
> ahead  ), but nothing related to the crash is logged.
>
> our salves machines are ntp-synchronized
>
>
> we run under debian/lenny
> we launch the machines with the kvm command
>
> kernel of host :  2.6.26-2-amd64
> kernel of host :  2.6.26-2-686


I'm pretty sure this issue has been fixed in newer versions of qemu-kvm 
and the kernel. You might want to try using the versions from backports. 
You are unlikely to get much support on a kernel that's 10 releases old 
and iirc, the kvm that came with lenny was only a development snapshot.


> command launched :
>
> kvm -name svn -drive file=xxx -net tap -m 256 -net nic,macaddr=xxx  -pidfile
> /var/run/kvm/xx.pid -daemonize -k fr -vnc :63006 -monitor
> unix:xx_monitor,server,nowait -vnc unix:xx_vnc
>
>
> If you ever have an idea  how to solve or to debug the problem ...
>
> Thanks
>
>
> Fred
>
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe kvm" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html