Re: [Openstack] Can't launch VM with XCP plus OpenStack Essex release version

2012-06-11 Thread John Garbutt
It is not well documented (I promise to work on that soon), but you should use 
an EXT3 local SR, not LVM. It is certainly worth moving to EXT3, just to rule a 
few things out.

I see you are using the Ubuntu XCP packages. This is a version of XenAPI I have 
not yet started doing much QA on, but I did get a VM started OK when using 
EXT3. However, I used to get LVM working for the basics with earlier versions 
of xenapi, so I am shocked that is totally broken on the latest.

Also, what sort of VM are you trying to launch? Is it a three part image with 
external kernel?

If you want to get going with XenAPI and OpenStack quickly, I would recommend 
DevStack and XenServer 6.0.2 (free download) or XCP 1.5 Beta. That is the 
combination I am testing today. Hoping to test Ubuntu XCP and OpenStack in the 
next few weeks.

Thanks,
John

PS:
I think the problem you are hitting is because there was a recent change in 
XCP/XenAPI where mounting the VDI into dom0 no longer automatically exposes the 
partitions of that disk to the Dom0 Kernel, so you need to use kpartx (or 
similar) to mount to appropriate partition. I can see you are hit this because 
the device returned is /dev/sm/guid/guid rather than /dev/sda1 or 
something like that. I got normal VMs working OK despite this, but have not yet 
gone through all the XenAPI plugins to ensure all the issues are fixed.

 -Original Message-
 From: openstack-bounces+john.garbutt=eu.citrix@lists.launchpad.net
 [mailto:openstack-bounces+john.garbutt=eu.citrix@lists.launchpad.net]
 On Behalf Of Takaaki Suzuki
 Sent: 11 June 2012 04:07
 To: openstack@lists.launchpad.net
 Cc: dev; Ryu Ishimoto
 Subject: [Openstack] Can't launch VM with XCP plus OpenStack Essex release
 version

 Hi all.

 I need your support again.
 We had this parted error with XCP plus OpenStack Essex release version.
 Nova can't launch VM with XCP with OpenStack. I use SR with Local LVM.
 Did you get parted command error? If you have any idea.
 Please let me know.

 --
 nova-compute log:
 --
 2012-06-11 11:42:49 DEBUG nova.compute.manager
 [req-85e34ab4-885c-4006-ac22-47b0323b9e21
 4a7c390b523d40c49aab4ca4dfd92ced 6f9b5c555cc94910bf5b7b4a0b3af4b6]
 [instance: ae09a885-ab5c-4b10-9582-73423932e40c] Deallocati ng network for
 instance from (pid=3057) _deallocate_network
 /usr/lib/python2.7/dist-packages/nova/compute/manager.py:616
 2012-06-11 11:42:49 DEBUG nova.rpc.amqp
 [req-85e34ab4-885c-4006-ac22-47b0323b9e21
 4a7c390b523d40c49aab4ca4dfd92ced 6f9b5c555cc94910bf5b7b4a0b3af4b6]
 Making asynchronous cast on network... from (pid=3057) cast /usr/l
 ib/python2.7/dist-packages/nova/rpc/amqp.py:346
 2012-06-11 11:42:49 ERROR nova.rpc.amqp
 [req-85e34ab4-885c-4006-ac22-47b0323b9e21
 4a7c390b523d40c49aab4ca4dfd92ced 6f9b5c555cc94910bf5b7b4a0b3af4b6]
 Exception during message handling
 2012-06-11 11:42:49 TRACE nova.rpc.amqp Traceback (most recent call last):
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/dist-packages/nova/rpc/amqp.py, line 252, in
 _process_data
 2012-06-11 11:42:49 TRACE nova.rpc.amqp rval =
 node_func(context=ctxt, **node_args)
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/dist-packages/nova/exception.py, line 114, in wrapped
 2012-06-11 11:42:49 TRACE nova.rpc.amqp return f(*args, **kw)
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/dist-packages/nova/compute/manager.py, line 177, in
 decorated_function
 2012-06-11 11:42:49 TRACE nova.rpc.amqp sys.exc_info())
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/contextlib.py, line 24, in __exit__
 2012-06-11 11:42:49 TRACE nova.rpc.amqp self.gen.next()
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/dist-packages/nova/compute/manager.py, line 171, in
 decorated_function
 2012-06-11 11:42:49 TRACE nova.rpc.amqp return function(self,
 context, instance_uuid, *args, **kwargs)
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/dist-packages/nova/compute/manager.py, line 651, in
 run_instance
 2012-06-11 11:42:49 TRACE nova.rpc.amqp do_run_instance()
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/dist-packages/nova/utils.py, line 945, in inner
 2012-06-11 11:42:49 TRACE nova.rpc.amqp retval = f(*args, **kwargs)
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/dist-packages/nova/compute/manager.py, line 650, in
 do_run_instance
 2012-06-11 11:42:49 TRACE nova.rpc.amqp
 self._run_instance(context, instance_uuid, **kwargs)
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/dist-packages/nova/compute/manager.py, line 451, in
 _run_instance
 2012-06-11 11:42:49 TRACE nova.rpc.amqp
 self._set_instance_error_state(context, instance_uuid)
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/contextlib.py, line 24, in __exit__
 2012-06-11 11:42:49 TRACE nova.rpc.amqp 

Re: [Openstack] Can't launch VM with XCP plus OpenStack Essex release version

2012-06-11 Thread Takaaki Suzuki
Hi John.

Thank you for your support again :)

 Also, what sort of VM are you trying to launch? Is it a three part image with 
 external kernel?
Yes, I use this VM images.
ttylinux-uec-amd64-12.1_2.6.35-22_1
 - ttylinux-uec-amd64-12.1_2.6.35-22_1.img
 - ttylinux-uec-amd64-12.1_2.6.35-22_1-initrd
 - ttylinux-uec-amd64-12.1_2.6.35-22_1-loader

cirros-0.3.0-x86_64-uec
 - cirros-0.3.0-x86_64-blank.img
 - cirros-0.3.0-x86_64-initrd
 - cirros-0.3.0-x86_64-vmlinuz

I'll check use EXT3 local SR and kpartx command.

Thanks!
Suzuki

On Mon, Jun 11, 2012 at 7:03 PM, John Garbutt john.garb...@citrix.com wrote:
 It is not well documented (I promise to work on that soon), but you should 
 use an EXT3 local SR, not LVM. It is certainly worth moving to EXT3, just to 
 rule a few things out.

 I see you are using the Ubuntu XCP packages. This is a version of XenAPI I 
 have not yet started doing much QA on, but I did get a VM started OK when 
 using EXT3. However, I used to get LVM working for the basics with earlier 
 versions of xenapi, so I am shocked that is totally broken on the latest.

 Also, what sort of VM are you trying to launch? Is it a three part image with 
 external kernel?

 If you want to get going with XenAPI and OpenStack quickly, I would recommend 
 DevStack and XenServer 6.0.2 (free download) or XCP 1.5 Beta. That is the 
 combination I am testing today. Hoping to test Ubuntu XCP and OpenStack in 
 the next few weeks.

 Thanks,
 John

 PS:
 I think the problem you are hitting is because there was a recent change in 
 XCP/XenAPI where mounting the VDI into dom0 no longer automatically exposes 
 the partitions of that disk to the Dom0 Kernel, so you need to use kpartx (or 
 similar) to mount to appropriate partition. I can see you are hit this 
 because the device returned is /dev/sm/guid/guid rather than 
 /dev/sda1 or something like that. I got normal VMs working OK despite this, 
 but have not yet gone through all the XenAPI plugins to ensure all the issues 
 are fixed.

 -Original Message-
 From: openstack-bounces+john.garbutt=eu.citrix@lists.launchpad.net
 [mailto:openstack-bounces+john.garbutt=eu.citrix@lists.launchpad.net]
 On Behalf Of Takaaki Suzuki
 Sent: 11 June 2012 04:07
 To: openstack@lists.launchpad.net
 Cc: dev; Ryu Ishimoto
 Subject: [Openstack] Can't launch VM with XCP plus OpenStack Essex release
 version

 Hi all.

 I need your support again.
 We had this parted error with XCP plus OpenStack Essex release version.
 Nova can't launch VM with XCP with OpenStack. I use SR with Local LVM.
 Did you get parted command error? If you have any idea.
 Please let me know.

 --
 nova-compute log:
 --
 2012-06-11 11:42:49 DEBUG nova.compute.manager
 [req-85e34ab4-885c-4006-ac22-47b0323b9e21
 4a7c390b523d40c49aab4ca4dfd92ced 6f9b5c555cc94910bf5b7b4a0b3af4b6]
 [instance: ae09a885-ab5c-4b10-9582-73423932e40c] Deallocati ng network for
 instance from (pid=3057) _deallocate_network
 /usr/lib/python2.7/dist-packages/nova/compute/manager.py:616
 2012-06-11 11:42:49 DEBUG nova.rpc.amqp
 [req-85e34ab4-885c-4006-ac22-47b0323b9e21
 4a7c390b523d40c49aab4ca4dfd92ced 6f9b5c555cc94910bf5b7b4a0b3af4b6]
 Making asynchronous cast on network... from (pid=3057) cast /usr/l
 ib/python2.7/dist-packages/nova/rpc/amqp.py:346
 2012-06-11 11:42:49 ERROR nova.rpc.amqp
 [req-85e34ab4-885c-4006-ac22-47b0323b9e21
 4a7c390b523d40c49aab4ca4dfd92ced 6f9b5c555cc94910bf5b7b4a0b3af4b6]
 Exception during message handling
 2012-06-11 11:42:49 TRACE nova.rpc.amqp Traceback (most recent call last):
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/dist-packages/nova/rpc/amqp.py, line 252, in
 _process_data
 2012-06-11 11:42:49 TRACE nova.rpc.amqp     rval =
 node_func(context=ctxt, **node_args)
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/dist-packages/nova/exception.py, line 114, in wrapped
 2012-06-11 11:42:49 TRACE nova.rpc.amqp     return f(*args, **kw)
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/dist-packages/nova/compute/manager.py, line 177, in
 decorated_function
 2012-06-11 11:42:49 TRACE nova.rpc.amqp     sys.exc_info())
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/contextlib.py, line 24, in __exit__
 2012-06-11 11:42:49 TRACE nova.rpc.amqp     self.gen.next()
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/dist-packages/nova/compute/manager.py, line 171, in
 decorated_function
 2012-06-11 11:42:49 TRACE nova.rpc.amqp     return function(self,
 context, instance_uuid, *args, **kwargs)
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/dist-packages/nova/compute/manager.py, line 651, in
 run_instance
 2012-06-11 11:42:49 TRACE nova.rpc.amqp     do_run_instance()
 2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
 /usr/lib/python2.7/dist-packages/nova/utils.py, line 945, in inner
 2012-06-11 11:42:49 TRACE nova.rpc.amqp     retval = f(*args, 

Re: [Openstack] Can't launch VM with XCP plus OpenStack Essex release version

2012-06-11 Thread John Garbutt
Happy to help :-)

You shouldn't need to use kpartx yourself.
We may need to use it in the nova code where we try to mount VDIs in Dom0 (in 
XenAPI plugins)
Long term, I should probably look at ways around doing that.

Cheers,
John

 -Original Message-
 From: suz...@midokura.jp [mailto:suz...@midokura.jp] On Behalf Of
 Takaaki Suzuki
 Sent: 11 June 2012 11:37
 To: John Garbutt
 Cc: openstack@lists.launchpad.net; dev; Ryu Ishimoto; Mike McClurg
 Subject: Re: [Openstack] Can't launch VM with XCP plus OpenStack Essex
 release version

 Hi John.

 Thank you for your support again :)

  Also, what sort of VM are you trying to launch? Is it a three part image 
  with
 external kernel?
 Yes, I use this VM images.
 ttylinux-uec-amd64-12.1_2.6.35-22_1
  - ttylinux-uec-amd64-12.1_2.6.35-22_1.img
  - ttylinux-uec-amd64-12.1_2.6.35-22_1-initrd
  - ttylinux-uec-amd64-12.1_2.6.35-22_1-loader

 cirros-0.3.0-x86_64-uec
  - cirros-0.3.0-x86_64-blank.img
  - cirros-0.3.0-x86_64-initrd
  - cirros-0.3.0-x86_64-vmlinuz

 I'll check use EXT3 local SR and kpartx command.

 Thanks!
 Suzuki

 On Mon, Jun 11, 2012 at 7:03 PM, John Garbutt john.garb...@citrix.com
 wrote:
  It is not well documented (I promise to work on that soon), but you should
 use an EXT3 local SR, not LVM. It is certainly worth moving to EXT3, just to
 rule a few things out.
 
  I see you are using the Ubuntu XCP packages. This is a version of XenAPI I
 have not yet started doing much QA on, but I did get a VM started OK when
 using EXT3. However, I used to get LVM working for the basics with earlier
 versions of xenapi, so I am shocked that is totally broken on the latest.
 
  Also, what sort of VM are you trying to launch? Is it a three part image 
  with
 external kernel?
 
  If you want to get going with XenAPI and OpenStack quickly, I would
 recommend DevStack and XenServer 6.0.2 (free download) or XCP 1.5 Beta.
 That is the combination I am testing today. Hoping to test Ubuntu XCP and
 OpenStack in the next few weeks.
 
  Thanks,
  John
 
  PS:
  I think the problem you are hitting is because there was a recent change in
 XCP/XenAPI where mounting the VDI into dom0 no longer automatically
 exposes the partitions of that disk to the Dom0 Kernel, so you need to use
 kpartx (or similar) to mount to appropriate partition. I can see you are hit 
 this
 because the device returned is /dev/sm/guid/guid rather than
 /dev/sda1 or something like that. I got normal VMs working OK despite this,
 but have not yet gone through all the XenAPI plugins to ensure all the issues
 are fixed.
 
  -Original Message-
  From:
  openstack-bounces+john.garbutt=eu.citrix@lists.launchpad.net
  [mailto:openstack-bounces+john.garbutt=eu.citrix.com@lists.launchpad.
  net]
  On Behalf Of Takaaki Suzuki
  Sent: 11 June 2012 04:07
  To: openstack@lists.launchpad.net
  Cc: dev; Ryu Ishimoto
  Subject: [Openstack] Can't launch VM with XCP plus OpenStack Essex
  release version
 
  Hi all.
 
  I need your support again.
  We had this parted error with XCP plus OpenStack Essex release version.
  Nova can't launch VM with XCP with OpenStack. I use SR with Local LVM.
  Did you get parted command error? If you have any idea.
  Please let me know.
 
  --
  nova-compute log:
  --
  2012-06-11 11:42:49 DEBUG nova.compute.manager
  [req-85e34ab4-885c-4006-ac22-47b0323b9e21
  4a7c390b523d40c49aab4ca4dfd92ced 6f9b5c555cc94910bf5b7b4a0b3af4b6]
  [instance: ae09a885-ab5c-4b10-9582-73423932e40c] Deallocati ng
  network for instance from (pid=3057) _deallocate_network
  /usr/lib/python2.7/dist-packages/nova/compute/manager.py:616
  2012-06-11 11:42:49 DEBUG nova.rpc.amqp
  [req-85e34ab4-885c-4006-ac22-47b0323b9e21
  4a7c390b523d40c49aab4ca4dfd92ced 6f9b5c555cc94910bf5b7b4a0b3af4b6]
  Making asynchronous cast on network... from (pid=3057) cast /usr/l
  ib/python2.7/dist-packages/nova/rpc/amqp.py:346
  2012-06-11 11:42:49 ERROR nova.rpc.amqp
  [req-85e34ab4-885c-4006-ac22-47b0323b9e21
  4a7c390b523d40c49aab4ca4dfd92ced 6f9b5c555cc94910bf5b7b4a0b3af4b6]
  Exception during message handling
  2012-06-11 11:42:49 TRACE nova.rpc.amqp Traceback (most recent call
 last):
  2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
  /usr/lib/python2.7/dist-packages/nova/rpc/amqp.py, line 252, in
  _process_data
  2012-06-11 11:42:49 TRACE nova.rpc.amqp rval =
  node_func(context=ctxt, **node_args)
  2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
  /usr/lib/python2.7/dist-packages/nova/exception.py, line 114, in
  wrapped
  2012-06-11 11:42:49 TRACE nova.rpc.amqp return f(*args, **kw)
  2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
  /usr/lib/python2.7/dist-packages/nova/compute/manager.py, line 177,
  in decorated_function
  2012-06-11 11:42:49 TRACE nova.rpc.amqp sys.exc_info())
  2012-06-11 11:42:49 TRACE nova.rpc.amqp   File
  /usr/lib/python2.7/contextlib.py, line 24, in __exit__
  2012-06-11 11:42:49 TRACE nova.rpc.amqp self.gen.next