Xen devel - Sep 2008 - Accessing the VNC using HVM stub domain - xen 3.3.0

Hello all,

I''m writing a documentation about the news features of xen 3.3.0. The
VNC is available configured to access by domain 0 but  I can''t access
configuring it to work into -dm virtual machine.
Below my conffiles:

# /etc/xen/hvmachine
kernel = "/usr/lib/xen/boot/hvmloader"
builder=''hvm''
memory = 1024
name = "hvmachine"
vcpus=2
vif = [ ''type=ioemu, bridge=eth0'' ]
device_model = ''/usr/lib/xen/bin/stubdom-dm''
boot="cd"
sdl=0
opengl=0
vnc=1
vnclisten="172.30.206.1"
stdvga=0

# /etc/xen/hvmachine-dm
kernel = "/usr/lib/xen/boot/ioemu-stubdom.gz"
vif = [ ''ip=172.30.206.1'',
''bridge=eth0,mac=00:16:3e:21:a2:94'' ]
disk = [
''file:/home/iso/debian-LennyBeta2-amd64-CD-1.iso,hdc:cdrom,r'',''phy:/dev/empirevg/pvmachine,hda,w'']

Any idea?

Regards,

-- 
Marco Sinhoreli

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Hello,

Marco Sinhoreli, le Thu 11 Sep 2008 09:47:44 -0300, a écrit
:
> I''m writing a documentation about the news features of xen 3.3.0.
The
> VNC is available configured to access by domain 0 but  I can''t
access
> configuring it to work into -dm virtual machine.
Does your network configuration permit to expose IP 172.30.206.1 on a PV
vif interface?
(apart from that your configuration files seem good)

Samuel

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Hi Samuel,

I believed that this IP was only to internal communication between the
two virtual machines. I was mistaked.  Thanks again!

On Thu, Sep 11, 2008 at 1:00 PM, Samuel Thibault
<samuel.thibault@ens-lyon.org> wrote:
> Hello,
>
> Marco Sinhoreli, le Thu 11 Sep 2008 09:47:44 -0300, a écrit :
>> I'm writing a documentation about the news features of xen 3.3.0.
The
>> VNC is available configured to access by domain 0 but  I can't
access
>> configuring it to work into -dm virtual machine.
>
> Does your network configuration permit to expose IP 172.30.206.1 on a PV
> vif interface?
> (apart from that your configuration files seem good)
>
> Samuel
>


-- 
Marco Sinhoreli

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Marco Sinhoreli, le Thu 11 Sep 2008 14:45:49 -0300, a écrit
:
> I believed that this IP was only to internal communication between the
> two virtual machines.
It is, but you still need to make sure to configure the network for that
to happen :)

Samuel

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Hi Samuel,

Thans again,. Changing the subject, I''m impressed with the stub domain
 performance! I did some benchmarks using bonnie++ and netperf and
it''s better than pure fullvirt machine.
* netperf:
stub-hvm: 25.99 Kbits/sec
hvm: 15.23 Kbits/sec

In bonnie++ benchmark, the item "write block" is impressive:
....................Per char........Block........Rewrite
stub-hvm .....20898..........41857.......21045
hvm .............19046..........17462.......19331

These values are in K/sec.

Regards,

-- 
Marco Sinhoreli

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Marco Sinhoreli, le Thu 11 Sep 2008 16:37:41 -0300, a écrit
:
> Thans again,. Changing the subject, I''m impressed with the stub
domain
>  performance! I did some benchmarks using bonnie++ and netperf and
> it''s better than pure fullvirt machine.
> * netperf:
> stub-hvm: 25.99 Kbits/sec
> hvm: 15.23 Kbits/sec
That''s what I measured too indeed.
> In bonnie++ benchmark, the item "write block" is impressive:
> ....................Per char........Block........Rewrite
> stub-hvm .....20898..........41857.......21045
> hvm .............19046..........17462.......19331
Is that disk performance?

Samuel

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

> Marco Sinhoreli, le Thu 11 Sep 2008 16:37:41 -0300, a écrit :
>> Thans again,. Changing the subject, I'm impressed with the stub
domain
>>  performance! I did some benchmarks using bonnie++ and netperf and
>> it's better than pure fullvirt machine.
>> * netperf:
>> stub-hvm: 25.99 Kbits/sec
>> hvm: 15.23 Kbits/sec
>
> That's what I measured too indeed.
>
>> In bonnie++ benchmark, the item "write block" is impressive:
>> ....................Per char........Block........Rewrite
>> stub-hvm .....20898..........41857.......21045
>> hvm .............19046..........17462.......19331
>
> Is that disk performance?
Yes, I´m using sata disks in this bench. I have servers with SAS but
this is in production.


-- 
Marco Sinhoreli

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Marco Sinhoreli, le Thu 11 Sep 2008 20:32:31 -0300, a écrit
:
> >> In bonnie++ benchmark, the item "write block" is
impressive:
> >> ....................Per char........Block........Rewrite
> >> stub-hvm .....20898..........41857.......21045
> >> hvm .............19046..........17462.......19331
> >
> > Is that disk performance?
> Yes, I´m using sata disks in this bench. I have servers with SAS but
> this is in production.
Ok. In my tests I was getting ~60MB/s in both case, the improvement was
rather of the order of just 10% (just getting a bit closer to native).

Samuel

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel