On Fri, 2009-06-19 at 20:14 +0200, Daniel Veillard wrote:
> On Fri, Jun 19, 2009 at 05:45:33PM +, David Lutterkort wrote:
> > Agreed .. that format wouldn't help much with static checking.
>
> Okay, well I think the recursive definition is really the worse
> for validation and processing.
On Fri, Jun 19, 2009 at 05:47:44PM +, David Lutterkort wrote:
> On Fri, 2009-06-19 at 10:47 +0100, Daniel P. Berrange wrote:
> > IMHO that results in a bad structure, because its anot associating
> > the related info together, eg having an separate element to turn
> > on/off IPV6, and then list
On Fri, Jun 19, 2009 at 05:44:25PM +, David Lutterkort wrote:
> > There are 4 possible arrangements of physical NIC, bond and vlan,
> > each of which can use a bridge. This gives 8 total configs
> >
> > 1. Physical NIC
> >
> > 2. Physical NIC + bridge
> >
> > 3. Physical NIC + bond
> >
>
On Fri, Jun 19, 2009 at 08:19:24PM +0200, Daniel Veillard wrote:
> On Fri, Jun 19, 2009 at 05:39:57PM +0100, Daniel P. Berrange wrote:
> > On Thu, Jun 18, 2009 at 05:53:59PM +, David Lutterkort wrote:
> > 5. Physical NIC + 2 * vlan
> >
> >
> > eth0
> >
> >
> >
>
On Fri, Jun 19, 2009 at 06:16:20PM +0100, Daniel P. Berrange wrote:
> On Fri, Jun 19, 2009 at 07:05:23PM +0200, Daniel Veillard wrote:
> > > Of course, you could actually have a hybrid of 7/8, where some vlans
> > > where bridged, and others direct endpoints. That's trivally handled
> > > there of
On Fri, Jun 19, 2009 at 05:39:57PM +0100, Daniel P. Berrange wrote:
> On Thu, Jun 18, 2009 at 05:53:59PM +, David Lutterkort wrote:
> 5. Physical NIC + 2 * vlan
>
>
> eth0
>
>
>
> vlan42
>
>
> eth0
>
>
On Fri, Jun 19, 2009 at 05:45:33PM +, David Lutterkort wrote:
> On Fri, 2009-06-19 at 17:17 +0100, Daniel P. Berrange wrote:
> > On Fri, Jun 19, 2009 at 06:13:37PM +0200, Daniel Veillard wrote:
> > > On Thu, Jun 18, 2009 at 10:50:10PM +, David Lutterkort wrote:
> > > > On Thu, 2009-06-18 at
On Fri, 2009-06-19 at 10:47 +0100, Daniel P. Berrange wrote:
> IMHO that results in a bad structure, because its anot associating
> the related info together, eg having an separate element to turn
> on/off IPV6, and then listing addresses:
>
>
>
>
>
> vs having the direct association
On Fri, 2009-06-19 at 17:17 +0100, Daniel P. Berrange wrote:
> On Fri, Jun 19, 2009 at 06:13:37PM +0200, Daniel Veillard wrote:
> > On Thu, Jun 18, 2009 at 10:50:10PM +, David Lutterkort wrote:
> > > On Thu, 2009-06-18 at 20:48 +0100, Daniel P. Berrange wrote:
> > > > On Thu, Jun 18, 2009 at 07
On Fri, 2009-06-19 at 17:39 +0100, Daniel P. Berrange wrote:
> Here is a proposal that is a compromise between the single hierarchy, and
> completely flat. The break point is only introduced where VLANs appear,
> which is acceptable because when defining VLAns, you don't need to define
> the underl
On 06/19/2009 01:09 PM, Daniel P. Berrange wrote:
> On Fri, Jun 19, 2009 at 12:37:11PM -0400, Cole Robinson wrote:
>> The storage driver arranges its parsing routines in a way that make them
>> difficult to use in the test driver for non-default file parsing. This
>> refactoring moves things to be
On Fri, Jun 19, 2009 at 05:54:47PM +0200, Matthias Bolte wrote:
> 2009/6/19 Daniel P. Berrange :
> >
> > ACK, looks fine to me - though I'm wondering if you might also want to
> > explicitly turn off support for lists ? IIIRC the VMware config values
> > were only scalars, so for safety you might w
On Fri, Jun 19, 2009 at 07:05:23PM +0200, Daniel Veillard wrote:
> On Fri, Jun 19, 2009 at 05:39:57PM +0100, Daniel P. Berrange wrote:
> > On Thu, Jun 18, 2009 at 05:53:59PM +, David Lutterkort wrote:
> >
> > So the possible configs would appear as XML like
> >
> > 1. Physical NIC
> >
> >
On Fri, Jun 19, 2009 at 12:37:33PM -0400, Cole Robinson wrote:
> If specifying a custom test driver, virtual networks were not 'activated'
> on driver init. This differs from the behavior of domains and storage pools,
> so fix it. Also improve a couple error messages in that area.
>
> Signed-off-b
On Fri, Jun 19, 2009 at 12:37:21PM -0400, Cole Robinson wrote:
> If using a custom test driver, storage pool file parsing was broken, and
> storage volume parsing was never implemented. Fix these issues, and add
> some examples in docs/
>
> Signed-off-by: Cole Robinson
> ---
> docs/testnode.xml
On Fri, Jun 19, 2009 at 12:37:11PM -0400, Cole Robinson wrote:
> The storage driver arranges its parsing routines in a way that make them
> difficult to use in the test driver for non-default file parsing. This
> refactoring moves things to be consistent with the way domain_conf and
> network_conf
On Fri, Jun 19, 2009 at 12:36:51PM -0400, Cole Robinson wrote:
> virEnumFromString doesn't check for a NULL string, and will segfault if
> passed one. Lots of calling code protects against this, but at least
> /pool/@type parsing does not.
>
> Signed-off-by: Cole Robinson
> ---
> src/util.c |
On Fri, Jun 19, 2009 at 12:36:51PM -0400, Cole Robinson wrote:
> virEnumFromString doesn't check for a NULL string, and will segfault if
> passed one. Lots of calling code protects against this, but at least
> /pool/@type parsing does not.
Agreed, ACK
Daniel
--
Daniel Veillard | libxml G
On Fri, Jun 19, 2009 at 12:01:52PM -0400, David Allan wrote:
> * src/node_device.c src/node_device_hal_linux.c: check in both
> /sys/class/fc_host/hostN and /sys/class/scsi_host/hostN for the vport
> operations files
> ---
> src/node_device.c | 23 ++-
> src/node_
On Fri, Jun 19, 2009 at 05:39:57PM +0100, Daniel P. Berrange wrote:
> On Thu, Jun 18, 2009 at 05:53:59PM +, David Lutterkort wrote:
> > On Thu, 2009-06-18 at 18:06 +0100, Daniel P. Berrange wrote:
> >
> > > VLANs are tricky, because you can define VLANs on a physical
> > > inteface or a bond
On Thu, Jun 18, 2009 at 05:53:59PM +, David Lutterkort wrote:
> On Thu, 2009-06-18 at 18:06 +0100, Daniel P. Berrange wrote:
>
> > VLANs are tricky, because you can define VLANs on a physical
> > inteface or a bond interface, and you then may want to also
> > add a bridge on top of a VLAN, eg
If specifying a custom test driver, virtual networks were not 'activated'
on driver init. This differs from the behavior of domains and storage pools,
so fix it. Also improve a couple error messages in that area.
Signed-off-by: Cole Robinson
---
src/test.c |7 +--
1 files changed, 5 inse
If using a custom test driver, storage pool file parsing was broken, and
storage volume parsing was never implemented. Fix these issues, and add
some examples in docs/
Signed-off-by: Cole Robinson
---
docs/testnode.xml |3 ++
docs/testpool.xml | 15 +
docs/testvol.xml |6 +++
The storage driver arranges its parsing routines in a way that make them
difficult to use in the test driver for non-default file parsing. This
refactoring moves things to be consistent with the way domain_conf and
network_conf do things.
Signed-off-by: Cole Robinson
---
src/libvirt_private.syms
virEnumFromString doesn't check for a NULL string, and will segfault if
passed one. Lots of calling code protects against this, but at least
/pool/@type parsing does not.
Signed-off-by: Cole Robinson
---
src/util.c |3 +++
1 files changed, 3 insertions(+), 0 deletions(-)
diff --git a/src/ut
On Fri, Jun 19, 2009 at 06:13:37PM +0200, Daniel Veillard wrote:
> On Thu, Jun 18, 2009 at 10:50:10PM +, David Lutterkort wrote:
> > On Thu, 2009-06-18 at 20:48 +0100, Daniel P. Berrange wrote:
> > > On Thu, Jun 18, 2009 at 07:05:29PM +0100, Daniel P. Berrange wrote:
> > > >
> > > > I think th
On Fri, Jun 19, 2009 at 12:01:52PM -0400, David Allan wrote:
> * src/node_device.c src/node_device_hal_linux.c: check in both
> /sys/class/fc_host/hostN and /sys/class/scsi_host/hostN for the vport
> operations files
> ---
> src/node_device.c | 23 ++-
> src/node_
On Thu, Jun 18, 2009 at 10:50:10PM +, David Lutterkort wrote:
> On Thu, 2009-06-18 at 20:48 +0100, Daniel P. Berrange wrote:
> > On Thu, Jun 18, 2009 at 07:05:29PM +0100, Daniel P. Berrange wrote:
> > >
> > > I think this is a really unpleasant format to deal with. IMHO there should
> > > not
* src/node_device.c src/node_device_hal_linux.c: check in both
/sys/class/fc_host/hostN and /sys/class/scsi_host/hostN for the vport
operations files
---
src/node_device.c | 23 ++-
src/node_device_hal_linux.c | 39 +++
2 file
Ok, I redid the patch into a minimal change that fixes NPIV on older
kernels. I'll resubmit the rest of it when I decide how I want to
deal with the problem of things moving around in /sys.
Dave
--
Libvir-list mailing list
Libvir-list@redhat.com
https://www.redhat.com/mailman/listinfo/libvir-li
2009/6/19 Daniel P. Berrange :
> On Fri, Jun 19, 2009 at 10:01:20AM +0200, Matthias Bolte wrote:
>> Hi,
>>
>> As part of the VMware ESX driver I need to parse VMware VMX config
>> files. This files have basically the same syntax as Xen config files,
>> but the key part of the key=value format is al
On Fri, Jun 19, 2009 at 05:24:20PM +0200, Daniel Veillard wrote:
> On Fri, Jun 19, 2009 at 10:57:59AM -0400, Laine Stump wrote:
> >
> > This is a resubmission of the virsh commands to expose virInterface*
> > functionality. I've made the following changes from the original:
> >
> > 1) command nam
On Fri, Jun 19, 2009 at 10:57:59AM -0400, Laine Stump wrote:
>
> This is a resubmission of the virsh commands to expose virInterface*
> functionality. I've made the following changes from the original:
>
> 1) command names changed from if-* to iface-*
>
> 2) "if-create" is now "iface-start"
>
>
This is a resubmission of the virsh commands to expose virInterface*
functionality. I've made the following changes from the original:
1) command names changed from if-* to iface-*
2) "if-create" is now "iface-start"
3) iface-edit implementation is manually included directly in virsh.c
rathe
On Thu, Jun 18, 2009 at 01:25:46PM -0400, Amy Griffis wrote:
> Taking a second look at the log level code and documentation,
> there seems to be some confusion around the value of '0'. It's
> not one of the defined log priorities, but it is mentioned in the
> documentation on the website both as me
Daniel P. Berrange wrote:
On Fri, Jun 19, 2009 at 10:07:36AM +0200, Daniel Veillard wrote:
On Thu, Jun 18, 2009 at 09:40:21PM -0400, David Allan wrote:
* src/node_device_hal_linux.c, src/node_device.c: Older kernels had
vport_create and delete in /sys/class/scsi_host not /sys/class/fc_host.
Thi
On Thu, Jun 18, 2009 at 12:20:16PM +0100, Daniel P. Berrange wrote:
> On Thu, Jun 18, 2009 at 06:08:36PM +0900, Tatsuro Enokura wrote:
> > Hi all
> >
> > I try virsh dominfo in upstream libvirt on xen machine,
> > the commands returns -1 as follows:
>
> Opps, thanks for reporting this problem - s
On 06/19/2009 06:52 AM, Daniel P. Berrange wrote:
> On Thu, Jun 18, 2009 at 11:27:50AM +0200, Paul Reeves wrote:
>> A quick question -
>>
>> What is the correct root for dynamically generated KVM mac addresses? The
>> docs (man virt-install) say 54:52:00 and I've seen Ubuntu docs on-line that
>>
On Fri, Jun 19, 2009 at 10:01:20AM +0200, Matthias Bolte wrote:
> Hi,
>
> As part of the VMware ESX driver I need to parse VMware VMX config
> files. This files have basically the same syntax as Xen config files,
> but the key part of the key=value format is allowed to contain ':' and
> '.'.
>
>
On Thu, Jun 18, 2009 at 09:40:08AM +0100, Andreas Sommer wrote:
> I executed the commands "export LIBVIRT_DEBUG=1" and "virsh dominfo 2"
> (ID 2 was a running domU), this is the output:
>
> -
> DEBUG: libvirt.
On Fri, Jun 19, 2009 at 10:58:18AM +0900, Tatsuro Enokura wrote:
> Hi Dan,
>
> Daniel P. Berrange wrote:
> >>The explanation of virNodeGetSecurityModel() and
> >>virNodeGetSecurityModel() in libvirt.c is return -2
> >>when hypervisor drivers don't support these operations.
> >>But these functions
On Thu, Jun 18, 2009 at 11:27:50AM +0200, Paul Reeves wrote:
>
> A quick question -
>
> What is the correct root for dynamically generated KVM mac addresses? The
> docs (man virt-install) say 54:52:00 and I've seen Ubuntu docs on-line that
> say 52:54:00.
>
> Has someone got these mixed up?
On Thu, Jun 18, 2009 at 02:44:30PM -0400, Laine Stump wrote:
> ---
> qemud/remote.c | 51
> qemud/remote_dispatch_args.h |1 +
> qemud/remote_dispatch_prototypes.h | 14 +++
> qemud/remote_dispatch_ret.h|2 +
> qemud/remote
On Thu, Jun 18, 2009 at 02:44:29PM -0400, Laine Stump wrote:
> ---
> src/libvirt.c | 89
> +++--
> 1 files changed, 86 insertions(+), 3 deletions(-)
>
> diff --git a/src/libvirt.c b/src/libvirt.c
> index bf49018..c5c868d 100644
> --- a/src/li
On Thu, Jun 18, 2009 at 02:44:28PM -0400, Laine Stump wrote:
> ---
> src/driver.h |8
> 1 files changed, 8 insertions(+), 0 deletions(-)
>
> diff --git a/src/driver.h b/src/driver.h
> index ca759ff..2502c63 100644
> --- a/src/driver.h
> +++ b/src/driver.h
> @@ -508,6 +508,12 @@ typed
On Thu, Jun 18, 2009 at 02:44:27PM -0400, Laine Stump wrote:
> This adds virConnectListDefinedInterfaces() and
> virConnectNumOfDefinedInterfaces().
> ---
> include/libvirt/libvirt.h|5 +
> include/libvirt/libvirt.h.in |5 +
> src/libvirt_public.syms |5 +
> 3 file
On Fri, Jun 19, 2009 at 10:01:20AM +0200, Matthias Bolte wrote:
> Hi,
>
> As part of the VMware ESX driver I need to parse VMware VMX config
> files. This files have basically the same syntax as Xen config files,
> but the key part of the key=value format is allowed to contain ':' and
> '.'.
>
>
On Fri, Jun 19, 2009 at 10:07:36AM +0200, Daniel Veillard wrote:
> On Thu, Jun 18, 2009 at 09:40:21PM -0400, David Allan wrote:
> > * src/node_device_hal_linux.c, src/node_device.c: Older kernels had
> > vport_create and delete in /sys/class/scsi_host not /sys/class/fc_host.
> > This patch causes l
On Thu, Jun 18, 2009 at 03:13:06PM -0400, Cole Robinson wrote:
> Cole Robinson wrote:
> > CreateXMLFrom changes accidentally caused all raw volume creation to be
> > fully allocated (as though allocation == capacity). Fix this.
> >
> > Also force CreateXMLFrom to maintain the previous behavior: sp
On Thu, Jun 18, 2009 at 12:35:40PM -0400, Cole Robinson wrote:
> The ID of the existing VM was being unconditionally set to -1, which was
> upsetting virt-manager.
>
> Signed-off-by: Cole Robinson
ACK.
Daniel
--
|: Red Hat, Engineering, London -o- http://people.redhat.com/berrange/ :|
|: h
On Thu, Jun 18, 2009 at 12:35:39PM -0400, Cole Robinson wrote:
> Currently, 'info' will always report that mem = max mem. Make sure we
> actually return the correct mem value.
>
> Signed-off-by: Cole Robinson
Ahh, very subtle ! ACK
Daniel
--
|: Red Hat, Engineering, London -o- http://peopl
On Thu, Jun 18, 2009 at 10:56:44PM +, David Lutterkort wrote:
> On Thu, 2009-06-18 at 18:10 +0100, Daniel P. Berrange wrote:
> > > I don't see that that buys us anything that we wouldn't have with
> > >
> > >
> > >
> > >
> > >
> > >
> >
> > If you
Hi,
As part of the VMware ESX driver I need to parse VMware VMX config
files. This files have basically the same syntax as Xen config files,
but the key part of the key=value format is allowed to contain ':' and
'.'.
The attached patch adds a flag to the virConfParser to enabled a VMX
mode for th
On Thu, Jun 18, 2009 at 09:40:21PM -0400, David Allan wrote:
> * src/node_device_hal_linux.c, src/node_device.c: Older kernels had
> vport_create and delete in /sys/class/scsi_host not /sys/class/fc_host.
> This patch causes libvirt to look in both places.
Important patch as it allows NPIV suppor
On Thu, Jun 18, 2009 at 03:13:06PM -0400, Cole Robinson wrote:
> Cole Robinson wrote:
> > CreateXMLFrom changes accidentally caused all raw volume creation to be
> > fully allocated (as though allocation == capacity). Fix this.
> >
> > Also force CreateXMLFrom to maintain the previous behavior: sp
On Thu, Jun 18, 2009 at 12:35:40PM -0400, Cole Robinson wrote:
> The ID of the existing VM was being unconditionally set to -1, which was
> upsetting virt-manager.
Makes sense too, ACK :-)
Daniel
--
Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/
dan...@veillard.com
On Thu, Jun 18, 2009 at 12:35:39PM -0400, Cole Robinson wrote:
> Currently, 'info' will always report that mem = max mem. Make sure we
> actually return the correct mem value.
Yes that makes sense to me, ACK !
thanks !
Daniel
--
Daniel Veillard | libxml Gnome XML XSLT toolkit http:
On Thu, Jun 18, 2009 at 10:56:44PM +, David Lutterkort wrote:
> The one argument for tags is that it makes it cleaner to
> bundle addressing info like and routing info, to make sure that the
> user doesn't specify ipv6 routes for an interface without ipv6
> addresses.
Yes that grouping is
58 matches
Mail list logo