Re: [Openstack] [xenserver][ocata] overcommit vcpu

[Bob Ball]

Hi Adhi,

Do you mean that you can’t run two VMs each with 8 vCPUs, or do you mean that 
you are trying to run one VM with more than 8 vCPUs?

I believe that the cpu_allocation_ratio means that you can re-use each physical 
cpu (thread) up to 16 times with different VMs, but each VM is still restricted 
to at must the number of physical cpu threads you have on the machine.

Bob
From: Adhi Priharmanto [mailto:adhi@gmail.com]
Sent: 29 November 2017 11:38
To: openstack 
Subject: [Openstack] [xenserver][ocata] overcommit vcpu

Hi,

I've been running ocata release with xenserver as hypervisor, my question, how 
to overcommit vCPU of xenserver ?

my xenserver have 4 physical core and 8 vcpu. I already add 
"cpu_allocation_ratio = 16.0" option into my nova.conf at compute nodes, but 
its still read 8 vCPUs by scheduler.


--
​​

Cheers,


Adhi Priharmanto
about.me/a_dhi






+62-812-82121584

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [xenserver-ocata] neutron-dhcp/metadata agent on XenServer

[Bob Ball]

Hi Adhi,

If the Neutron agents would normally run with the nova compute services, then 
they would need to run in the compute VM when deploying with XenServer, not in 
domain 0.

I would assume that the DHCP agents should also run in the compute VM.

Also added openst...@citrix.com to include others.

Bob

From: Adhi Priharmanto [mailto:adhi@gmail.com]
Sent: 22 August 2017 06:58
To: openstack 
Subject: [Openstack] [xenserver-ocata] neutron-dhcp/metadata agent on XenServer

Hi,
According to this guide:
https://docs.openstack.org/ocata/networking-guide/config-dhcp-ha.html

If I want to add HA on my openstack-xenserver environment , is it possible to 
include neutron-dhcp/metadata agent on Compute-node ?

or I have to add neutron-dhcp/metadata agent on xenserver dom0 ?
or I have to create more neutron-dhcp/metadata agent on separate node ?

--
Cheers,


Adhi Priharmanto
about.me/a_dhi






+62-812-82121584

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [Ocata-Xenserver] Nova-compute wouldn't start

[Bob Ball]

Hi Adhi,

Please see bug report at https://bugs.launchpad.net/nova/+bug/1693147 which I 
believe covers your issue.

Huan is working on a fix at https://review.openstack.org/#/c/467926

Thanks,

Bob

From: Adhi Priharmanto [mailto:adhi@gmail.com]
Sent: 18 May 2017 17:52
To: Bob Ball 
Cc: openstack ; #OpenStack External Email 

Subject: Re: [Openstack] [Ocata-Xenserver] Nova-compute wouldn't start

Hi bob,
Thanks for your suggest, because it's non production use, I decide to hardcoded 
clean up instance record from nova table, and finally nova-compute service was 
running again. :)

On Thu, May 18, 2017 at 10:53 PM, Bob Ball 
mailto:bob.b...@citrix.com>> wrote:
Hi Adhi,

The point is that there is a mismatch between what Nova believes exists, and 
what really exists on the hypervisor.  How were the instances deleted?

Check the hypervisors that each server Nova knows about - I suspect at least 
one is expected to exist on the XenServer for which the compute will not start.

Bob

From: Adhi Priharmanto [mailto:adhi@gmail.com<mailto:adhi@gmail.com>]
Sent: 18 May 2017 16:40
To: Bob Ball mailto:bob.b...@citrix.com>>
Cc: openstack 
mailto:openstack@lists.openstack.org>>; 
#OpenStack External Email mailto:openst...@citrix.com>>
Subject: Re: [Openstack] [Ocata-Xenserver] Nova-compute wouldn't start

hi Bob,
I don't have any vm/instance on that compute/xenserver , I'm deleted all vm 
since last week.

On Thu, May 18, 2017 at 6:03 PM, Bob Ball 
mailto:bob.b...@citrix.com>> wrote:
Hi Adhi,

Very interesting.  I suspect that  self._get_vif_ref, below, is returning None. 
 This appears to be when the VM does not exist on the host - perhaps the VM was 
renamed, deleted, or it has been migrated to a different host and Nova’s 
records didn’t get updated?

I believe the issue is most likely occurring at 
http://git.openstack.org/cgit/openstack/nova/tree/nova/virt/xenapi/vif.py?h=15.0.4#n246
 where we attempt to detect the VM based on the instance’s name in Nova’s 
database.
Do you have VMs on this host? They don’t have to be running (for example if the 
host rebooted then the VMs may exist but in a shutdown state)

Could you check that the VMs that are on the host still have the name 
associated with them according to Nova?
For example:
root@host~/openrc# openstack server show 3662f9b8-1303-496a-8c21-97bba312432c
…
| OS-EXT-SRV-ATTR:instance_name| instance-914e  

 |
…
| name | 
dsvm-devstack-citrix-mia-nodepool-683420

|

In this case, the server has the name dsvm-devstack-citrix-mia-nodepool-683420 
but on the hypervisor the VM itself has the name  instance-914e

Thanks,

Bob


From: Adhi Priharmanto [mailto:adhi@gmail.com<mailto:adhi@gmail.com>]
Sent: 18 May 2017 10:34
To: openstack 
mailto:openstack@lists.openstack.org>>
Subject: [Openstack] [Ocata-Xenserver] Nova-compute wouldn't start

Hi all,
I have openstack RDO ocata using xenserver as Hypervisor, setup manually. It 
works fine about 2 weeks, today my nova-compute wouldn't start and when I check 
the nova-compute log :
2017-05-18 16:29:24.769 10764 INFO nova.service [-] Starting compute node 
(version 15.0.0-1.el7)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
[req-f29142ce-c68a-4d0a-b6ff-ba5b0a67ec6a - - - - -] Error starting thread.
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service Traceback (most recent 
call last):
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/oslo_service/service.py", line 722, in 
run_service
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service service.start()
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/service.py", line 144, in start
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
self.manager.init_host()
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/compute/manager.py", line 1152, in 
init_host
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
self._init_instance(context, instance)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/compute/manager.py", line 949, in 
_init_instance
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
self.driver.plug_vifs(instance, net_info)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/virt/xenapi/driver.py", line 309, in 
plug_vifs
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   

Re: [Openstack] [Ocata-Xenserver] Nova-compute wouldn't start

[Bob Ball]

Hi Adhi,

The point is that there is a mismatch between what Nova believes exists, and 
what really exists on the hypervisor.  How were the instances deleted?

Check the hypervisors that each server Nova knows about - I suspect at least 
one is expected to exist on the XenServer for which the compute will not start.

Bob

From: Adhi Priharmanto [mailto:adhi@gmail.com]
Sent: 18 May 2017 16:40
To: Bob Ball 
Cc: openstack ; #OpenStack External Email 

Subject: Re: [Openstack] [Ocata-Xenserver] Nova-compute wouldn't start

hi Bob,
I don't have any vm/instance on that compute/xenserver , I'm deleted all vm 
since last week.

On Thu, May 18, 2017 at 6:03 PM, Bob Ball 
mailto:bob.b...@citrix.com>> wrote:
Hi Adhi,

Very interesting.  I suspect that  self._get_vif_ref, below, is returning None. 
 This appears to be when the VM does not exist on the host - perhaps the VM was 
renamed, deleted, or it has been migrated to a different host and Nova’s 
records didn’t get updated?

I believe the issue is most likely occurring at 
http://git.openstack.org/cgit/openstack/nova/tree/nova/virt/xenapi/vif.py?h=15.0.4#n246
 where we attempt to detect the VM based on the instance’s name in Nova’s 
database.
Do you have VMs on this host? They don’t have to be running (for example if the 
host rebooted then the VMs may exist but in a shutdown state)

Could you check that the VMs that are on the host still have the name 
associated with them according to Nova?
For example:
root@host~/openrc# openstack server show 3662f9b8-1303-496a-8c21-97bba312432c
…
| OS-EXT-SRV-ATTR:instance_name| instance-914e  

 |
…
| name | 
dsvm-devstack-citrix-mia-nodepool-683420

|

In this case, the server has the name dsvm-devstack-citrix-mia-nodepool-683420 
but on the hypervisor the VM itself has the name  instance-914e

Thanks,

Bob


From: Adhi Priharmanto [mailto:adhi@gmail.com<mailto:adhi@gmail.com>]
Sent: 18 May 2017 10:34
To: openstack 
mailto:openstack@lists.openstack.org>>
Subject: [Openstack] [Ocata-Xenserver] Nova-compute wouldn't start

Hi all,
I have openstack RDO ocata using xenserver as Hypervisor, setup manually. It 
works fine about 2 weeks, today my nova-compute wouldn't start and when I check 
the nova-compute log :
2017-05-18 16:29:24.769 10764 INFO nova.service [-] Starting compute node 
(version 15.0.0-1.el7)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
[req-f29142ce-c68a-4d0a-b6ff-ba5b0a67ec6a - - - - -] Error starting thread.
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service Traceback (most recent 
call last):
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/oslo_service/service.py", line 722, in 
run_service
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service service.start()
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/service.py", line 144, in start
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
self.manager.init_host()
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/compute/manager.py", line 1152, in 
init_host
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
self._init_instance(context, instance)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/compute/manager.py", line 949, in 
_init_instance
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
self.driver.plug_vifs(instance, net_info)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/virt/xenapi/driver.py", line 309, in 
plug_vifs
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
self._vmops.plug_vifs(instance, network_info)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/virt/xenapi/vmops.py", line 1959, in 
plug_vifs
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
self.vif_driver.plug(instance, vif, device=device)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/virt/xenapi/vif.py", line 250, in plug
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service vif_ref = 
self._get_vif_ref(vif, vm_ref)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/virt/xenapi/vif.py", line 42, in 
_get_vif_ref
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service vif_refs = 
self._session.c

Re: [Openstack] [Ocata-Xenserver] Nova-compute wouldn't start

[Bob Ball]

Hi Adhi,

Very interesting.  I suspect that  self._get_vif_ref, below, is returning None. 
 This appears to be when the VM does not exist on the host - perhaps the VM was 
renamed, deleted, or it has been migrated to a different host and Nova’s 
records didn’t get updated?

I believe the issue is most likely occurring at 
http://git.openstack.org/cgit/openstack/nova/tree/nova/virt/xenapi/vif.py?h=15.0.4#n246
 where we attempt to detect the VM based on the instance’s name in Nova’s 
database.
Do you have VMs on this host? They don’t have to be running (for example if the 
host rebooted then the VMs may exist but in a shutdown state)

Could you check that the VMs that are on the host still have the name 
associated with them according to Nova?
For example:
root@host~/openrc# openstack server show 3662f9b8-1303-496a-8c21-97bba312432c
…
| OS-EXT-SRV-ATTR:instance_name| instance-914e  

 |
…
| name | 
dsvm-devstack-citrix-mia-nodepool-683420

|

In this case, the server has the name dsvm-devstack-citrix-mia-nodepool-683420 
but on the hypervisor the VM itself has the name  instance-914e

Thanks,

Bob


From: Adhi Priharmanto [mailto:adhi@gmail.com]
Sent: 18 May 2017 10:34
To: openstack 
Subject: [Openstack] [Ocata-Xenserver] Nova-compute wouldn't start

Hi all,
I have openstack RDO ocata using xenserver as Hypervisor, setup manually. It 
works fine about 2 weeks, today my nova-compute wouldn't start and when I check 
the nova-compute log :
2017-05-18 16:29:24.769 10764 INFO nova.service [-] Starting compute node 
(version 15.0.0-1.el7)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
[req-f29142ce-c68a-4d0a-b6ff-ba5b0a67ec6a - - - - -] Error starting thread.
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service Traceback (most recent 
call last):
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/oslo_service/service.py", line 722, in 
run_service
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service service.start()
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/service.py", line 144, in start
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
self.manager.init_host()
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/compute/manager.py", line 1152, in 
init_host
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
self._init_instance(context, instance)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/compute/manager.py", line 949, in 
_init_instance
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
self.driver.plug_vifs(instance, net_info)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/virt/xenapi/driver.py", line 309, in 
plug_vifs
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
self._vmops.plug_vifs(instance, network_info)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/virt/xenapi/vmops.py", line 1959, in 
plug_vifs
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service 
self.vif_driver.plug(instance, vif, device=device)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/virt/xenapi/vif.py", line 250, in plug
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service vif_ref = 
self._get_vif_ref(vif, vm_ref)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/nova/virt/xenapi/vif.py", line 42, in 
_get_vif_ref
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service vif_refs = 
self._session.call_xenapi("VM.get_VIFs", vm_ref)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/os_xenapi/client/session.py", line 200, in 
call_xenapi
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service return 
session.xenapi_request(method, args)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib/python2.7/site-packages/os_xenapi/client/XenAPI.py", line 130, in 
xenapi_request
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service result = 
_parse_result(getattr(self, methodname)(*full_params))
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib64/python2.7/xmlrpclib.py", line 1233, in __call__
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service return 
self.__send(self.__name, args)
2017-05-18 16:29:24.847 10764 ERROR oslo_service.service   File 
"/usr/lib64/python2.7/xmlrpclib.py", line

Re: [Openstack] Security Groups Can't Apply in Kilo with Neutron & XenServer

[Bob Ball]

Hi Adhi,

Ccould you provide more details about how you’re installing Liberty, including 
your local.conf (if using devstack) and the nova / neutron configuration files?
Also details about how you’re booting the instance and what security groups 
you’re expecting to be applied.

Thanks,

Bob

From: Adhi Priharmanto [mailto:adhi@gmail.com]
Sent: 15 September 2016 08:48
To: Huan Xie 
Cc: openstack@lists.openstack.org
Subject: Re: [Openstack] Security Groups Can't Apply in Kilo with Neutron & 
XenServer

Hi, I still no luck for this problem, even I using liberty release, Security 
groups still not applied on network. can you help me again ?

On Thu, Mar 17, 2016 at 10:55 AM, Adhi Priharmanto 
mailto:adhi@gmail.com>> wrote:
Ok, 'll try to patched my neutron

On Tue, Mar 15, 2016 at 8:52 AM, Huan Xie 
mailto:huan@citrix.com>> wrote:
Hi,
For apply the patch, you need to download the changed file with this 
https://review.openstack.org/#/c/251271/ and its dependent changes, you can 
find its dependent changes in the right corner(Related Changes) in you open the 
link.
For files that you need edit, in the middle of the code review page, you can 
find a section called “Files”, this part shows you which files are changed.

Best Regards//Huan

From: Adhi Priharmanto [mailto:adhi@gmail.com]
Sent: Monday, March 14, 2016 6:21 PM
To: Huan Xie
Cc: openstack@lists.openstack.org
Subject: Re: [Openstack] Security Groups Can't Apply in Kilo with Neutron & 
XenServer

Hi Xie,

I also commented on your post at blog.citrix :) , for step 1 - 3 was clear for 
me. I still confused about patched code in 
https://review.openstack.org/#/c/251271/ for some file, could you more explain 
how to, which file that I should edit ?

Thanks before

On Mon, Mar 14, 2016 at 3:34 PM, Huan Xie 
mailto:huan@citrix.com>> wrote:
Hi Adhi,

Do you use devstack to deploy XenServer + Kilo or manually?
Current Kilo release does not support XenServer + Neutron security group, 
because security group is implemented via iptables on Linux bridge, however, 
there is no Linux bridge created when booting a new instance.
But we now have a new fix to support neutron security group, we have tested 
that it can work, this will be implemented as a blue print 
https://review.openstack.org/#/c/251271/
So, if you want to use neutron security group in Kilo, you should add some 
patch for your code and also please make the configurations as below:


1.   In nova.conf, two configurations should be set
[DEFAULT]
firewall_driver = nova.virt.firewall.NoopFirewallDriver

security_group_api=neutron


[xenserver]
ovs_integration_bridge =
vif_driver = nova.virt.xenapi.vif.XenAPIOpenVswitchDriver

If you don’t know how to configure ovs_integration_bridge, then 
you can refer this blog 
https://www.citrix.com/blogs/2015/11/30/integrating-xenserver-rdo-and-neutron/


2.   In neutron,  check configurations ml2_conf.ini in compute node which 
is used for neutron L2 agent

[agent]

minimize_polling = False

root_helper_daemon =

root_helper = /usr/local/bin/neutron-rootwrap-xen-dom0 
/etc/neutron/rootwrap.conf



[ovs]

integration_bridge =

bridge_mappings =

Also for ovs configuration items, if you don’t clear on how to 
configure them, refer the blog


3.   In neutron, check configurations /etc/neutron/rootwrap.conf in compute 
node

[xenapi]

# XenAPI configuration is only required by the L2 agent if it is to

# target a XenServer/XCP compute host's dom0.

xenapi_connection_url=

xenapi_connection_username=

xenapi_connection_password=


Best Regards//Huan


 Original Message 
Subject: [Openstack] Security Groups Can't Apply in Kilo with Neutron & 
XenServer
From: Adhi Priharmanto
To: openstack@lists.openstack.org
CC:
Hi all,

I had Openstack Kilo installed on my lab, for Compute Hypervisor I use 
XenServer 6.5, and networking Using Neutron OVS. For Controller, Network, and 
Compute node I'm using Ubuntu 14.04.

My problem was Security Groups rules doesn't applied to the instance that 
created. For example, there is no rule for SSH port 22 in security group i 
defined to the instance, but instance with floating IP able to login by ssh 
from external network.

I've already add this option on my nova.conf

firewall_driver=nova.virt.xenapi.firewall.Dom0IptablesFirewallDriver

and also defined firewall_driver on my ml2_conf.ini at Controller, Network, and 
Compute node

[ovs]
enable_security_group = True
enable_ipset = True
firewall_driver = 
neutron.agent.linux.iptables_firewall.OVSHybridIptablesFirewallDriver

can somebody help me with this problem ?


--
Cheers,


Adhi Priharmanto
about.me/a_dhi









___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : 
openstack@lists.openstack.org

Re: [Openstack] [OpenStack-Infra] [third-party][openstack][CI] Examples required for "single_node_ci_data.yaml" config file

[Bob Ball]

You need your own credentials where I put  - I just didn’t want to 
share my credentials with the whole OpenStack community ☺

Bob

From: Abhishek Shrivastava [mailto:abhis...@cloudbyte.com]
Sent: 10 May 2016 10:04
To: Bob Ball 
Cc: openstack-in...@lists.openstack.org; openstack@lists.openstack.org
Subject: Re: [OpenStack-Infra] [third-party][openstack][CI] Examples required 
for "single_node_ci_data.yaml" config file

Thanks Bob,

So just need to put the following in the section without any credentials?

On Tue, May 10, 2016 at 1:35 PM, Bob Ball 
mailto:bob.b...@citrix.com>> wrote:
Hi Abishek,

I use the following which works well:
oscc_file_contents: |
# Do Not Edit - Generated & Managed by Puppet
clouds:
rax:
profile: rackspace
regions:
- IAD
- ORD
- DWF
auth:
username: 
password: 
project-id: 
#

Then my nodepool config has “cloud: rax” and “region-name: ‘IAD’” instead of 
any credentials.

Thanks,

Bob

From: Abhishek Shrivastava 
[mailto:abhis...@cloudbyte.com<mailto:abhis...@cloudbyte.com>]
Sent: 10 May 2016 07:13
To: 
openstack-in...@lists.openstack.org<mailto:openstack-in...@lists.openstack.org>;
 openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: [OpenStack-Infra] [third-party][openstack][CI] Examples required for 
"single_node_ci_data.yaml" config file

Hi Folks,

I have been configuring the new OpenStack CI and was not able to figure out 
what exactly needed to be put in the section "oscc_file_contents:". So, if 
anyone has done the configuration and is running the new OpenStack CI please 
let me know what to put in the configuration.

Also, need to know how to add slaves to Jenkins using the Nodepool images.

--
[https://docs.google.com/uc?export=download&id=0Byq0j7ZjFlFKV3ZCWnlMRXBCcU0&revid=0Byq0j7ZjFlFKa2V5VjdBSjIwUGx6bUROS2IrenNwc0kzd2IwPQ]
Thanks & Regards,
Abhishek
Cloudbyte Inc.<http://www.cloudbyte.com>



--
[https://docs.google.com/uc?export=download&id=0Byq0j7ZjFlFKV3ZCWnlMRXBCcU0&revid=0Byq0j7ZjFlFKa2V5VjdBSjIwUGx6bUROS2IrenNwc0kzd2IwPQ]
Thanks & Regards,
Abhishek
Cloudbyte Inc.<http://www.cloudbyte.com>
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [OpenStack-Infra] [third-party][openstack][CI] Examples required for "single_node_ci_data.yaml" config file

[Bob Ball]

Hi Abishek,

I use the following which works well:
oscc_file_contents: |
# Do Not Edit - Generated & Managed by Puppet
clouds:
rax:
profile: rackspace
regions:
- IAD
- ORD
- DWF
auth:
username: 
password: 
project-id: 
#

Then my nodepool config has “cloud: rax” and “region-name: ‘IAD’” instead of 
any credentials.

Thanks,

Bob

From: Abhishek Shrivastava [mailto:abhis...@cloudbyte.com]
Sent: 10 May 2016 07:13
To: openstack-in...@lists.openstack.org; openstack@lists.openstack.org
Subject: [OpenStack-Infra] [third-party][openstack][CI] Examples required for 
"single_node_ci_data.yaml" config file

Hi Folks,

I have been configuring the new OpenStack CI and was not able to figure out 
what exactly needed to be put in the section "oscc_file_contents:". So, if 
anyone has done the configuration and is running the new OpenStack CI please 
let me know what to put in the configuration.

Also, need to know how to add slaves to Jenkins using the Nodepool images.

--
[https://docs.google.com/uc?export=download&id=0Byq0j7ZjFlFKV3ZCWnlMRXBCcU0&revid=0Byq0j7ZjFlFKa2V5VjdBSjIwUGx6bUROS2IrenNwc0kzd2IwPQ]
Thanks & Regards,
Abhishek
Cloudbyte Inc.
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [Fuel] How to configure Xen + Libvirt using Fuel 8.0

[Bob Ball]

Hi Martin,

I've never seen or heard of this before; do you have any more details or a bug 
on launchpad I can read?

XenServer is used under openstack for hundreds of thousands of VM instances, so 
some configurations are definitely stable and don't suffer from the issue you 
saw.

Were you - by chance - using devstack? If a devstack VM reboot occurs then it 
will intentionally destroy the VMs and may not delete all disks - but since 
this cannot be a production environment and devstack is intended only for 
development, we have not pursued a fix.

Thanks,

Bob

 Original Message 
Subject: Re: [Openstack] [Fuel] How to configure Xen + Libvirt using Fuel 8.0
From: Martinx - ジェームズ
To: Bob Ball
CC: Mike Scherbakov ,Ashish Yadav ,Fabrizio Soppelsa 
,openstack@lists.openstack.org

>From time to time, randomly, XenServer loses all virtual machine definitions, 
>and you'll end up with many orphan disk images that only God knows where those 
>images belongs to.

I gave up on XenServer because of this, it happened more than once, on very 
different places.

It is a big pain to restore it.

Open Source Xen looks more reliable and stable.

But theses, I prefer KVM on Ubuntu anyway...

On 1 April 2016 at 13:43, Bob Ball 
mailto:bob.b...@citrix.com>> wrote:
XenServer is fully open source - XCP is no longer. You can download XenServer 
for free from www.XenServer.org<http://www.XenServer.org> without any contact 
with Citrix.

For libvirt+Xen AFAIK that plugin doesn't exist and you would need to manually 
configure each compute node after installation.

The XenServer plugin will, of course, work with the version downloaded from 
above and does not require anything only available direct from Citrix, nor does 
it require a support contract.

Thanks,

Bob


 Original Message 
Subject: Re: [Openstack] [Fuel] How to configure Xen + Libvirt using Fuel 8.0
From: Mike Scherbakov
To: Ashish Yadav ,Fabrizio Soppelsa
CC: openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>

We used open source Xen Cloud Platform (XCP) around Cactus/Diablo release of 
OpenStack. At that time, there was no good support via libvirt but via xenapi 
it worked pretty well. So if you are just looking for an open source working 
solution, take a careful look - xenapi & XCP used to be fully open source, and 
no XenServer is needed.

Thanks,

On Tue, Mar 29, 2016 at 11:15 PM Ashish Yadav 
mailto:gwalash...@gmail.com>> wrote:
Hi Fabrizio,

On Tue, Mar 29, 2016 at 8:58 PM, Fabrizio Soppelsa 
mailto:fsoppe...@mirantis.com>> wrote:
Hi Ashish,

you may want to check this out: Citrix - XenServer Plugin
https://www.mirantis.com/products/openstack-drivers-and-plugins/fuel-plugins/

Thanks for this but I know about this plugin. I wanted to use Xen + Libvirt not 
xenserver,
because it was using XenAPI (XAPI) tool-stack not XL.

I search through for Xen + Liibvirt but no luck till now.

--Regards
Ashishkumar S. Yadav
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : 
openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
--
Mike Scherbakov
#mihgen

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : 
openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack


___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [Fuel] How to configure Xen + Libvirt using Fuel 8.0

[Bob Ball]

XenServer is fully open source - XCP is no longer. You can download XenServer 
for free from www.XenServer.org without any contact with Citrix.

For libvirt+Xen AFAIK that plugin doesn't exist and you would need to manually 
configure each compute node after installation.

The XenServer plugin will, of course, work with the version downloaded from 
above and does not require anything only available direct from Citrix, nor does 
it require a support contract.

Thanks,

Bob

 Original Message 
Subject: Re: [Openstack] [Fuel] How to configure Xen + Libvirt using Fuel 8.0
From: Mike Scherbakov
To: Ashish Yadav ,Fabrizio Soppelsa
CC: openstack@lists.openstack.org

We used open source Xen Cloud Platform (XCP) around Cactus/Diablo release of 
OpenStack. At that time, there was no good support via libvirt but via xenapi 
it worked pretty well. So if you are just looking for an open source working 
solution, take a careful look - xenapi & XCP used to be fully open source, and 
no XenServer is needed.

Thanks,

On Tue, Mar 29, 2016 at 11:15 PM Ashish Yadav 
mailto:gwalash...@gmail.com>> wrote:
Hi Fabrizio,

On Tue, Mar 29, 2016 at 8:58 PM, Fabrizio Soppelsa 
mailto:fsoppe...@mirantis.com>> wrote:
Hi Ashish,

you may want to check this out: Citrix - XenServer Plugin
https://www.mirantis.com/products/openstack-drivers-and-plugins/fuel-plugins/

Thanks for this but I know about this plugin. I wanted to use Xen + Libvirt not 
xenserver,
because it was using XenAPI (XAPI) tool-stack not XL.

I search through for Xen + Liibvirt but no luck till now.

--Regards
Ashishkumar S. Yadav
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : 
openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
--
Mike Scherbakov
#mihgen
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [Fuel] v7.0 error during deployment

[Bob Ball]

I’ve consistently had the same problem.

A quick fix for me was to change the NTP servers to “10.20.0.1” (or whatever 
the local IP address of your Fuel master is), then you only have to worry about 
the NTP settings for the fuel master.

Bob

From: Adam Lawson [mailto:alaw...@aqorn.com]
Sent: 22 October 2015 22:17
To: openstack
Subject: [Openstack] [Fuel] v7.0 error during deployment

Can someone explain this error? It comes up every time and fails the whole 
deployment.

Puppet logs:
ERROR: Unable to communicate with at least one of NTP server, checked the 
following host(s): ["0.fuel.pool.ntp.org", 
"1.fuel.pool.ntp.org", 
"2.fuel.pool.ntp.org"] on node 
node-6.x.com


priority: 1500
fail_on_error: true
type: puppet
uids:
- '6'
parameters:
  puppet_modules: "/etc/puppet/modules"
  puppet_manifest: "/etc/puppet/modules/osnailyfacter/modular/ntp/ntp-check.pp"
  timeout: 600
  cwd: "/"
.
Inspect Astute logs for the details
What is required for these nodes to talk to the internet and is it possible to 
disable external ntp?

Adam Lawson

AQORN, Inc.
427 North Tatnall Street
Ste. 58461
Wilmington, Delaware 19801-2230
Toll-free: (844) 4-AQORN-NOW ext. 101
International: +1 302-387-4660
Direct: +1 916-246-2072
[http://www.aqorn.com/images/logo.png]
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] XenServer 6.5 or KVM

[Bob Ball]

I would always recommend the latest version of XenServer and of OpenStack.  In 
particular there were a couple of fixes for race conditions that have gone into 
Kilo, so I would recommend using that version.

XenServer 6.5 with SP1 is the newest XenServer and has a number of performance 
and scalability enhancements that, even if you are happy with the feature set 
in earlier versions of XenServer, mean that I would recommend it.
As a minimum, XenServer 6.2 would also work.

Thanks,

Bob


From: Leandro Mendes [theflock...@gmail.com]
Sent: 24 July 2015 15:15
To: Mārtiņš Jakubovičs
Cc: openstack@lists.openstack.org
Subject: Re: [Openstack] XenServer 6.5 or KVM

Thank you guys.

One last question: This 3 node setup is just a proof of concept of a bigger 
architecture (with around 30 nodes). I personally would prefer use KVM, so we 
can use Kilo and Neutron, but for some political reasons, we are really pushed 
to use XenServer (no problem about XS itself, I think it is great), but we 
don't have enough time and "hands" to make it work.

Thinking about it,i would ask (maybe Bob will give me the best answer) what 
should be the best choise for use Openstack with XenServer regarding their 
versions?
Maybe an older XCP with Juno or Icehouse?

And what about VMWare as it is as XenServer in Group B?

Thank you again.

On Fri, Jul 24, 2015 at 4:20 AM, Mārtiņš Jakubovičs 
mailto:mart...@vertigs.lv>> wrote:
Hello Leandro,

>From my experience OpenStack + XenServer is really difficult to build, but it 
>work. Newest features like neutron, advanced cinder drivers, XenServer pool's, 
>etc is not working well. At least with XenServer 6.2 and  Icehouse release. If 
>you know both technologies and don't know which to choose then use KVM.
If you want to use XenServer then community member Bob Ball 
(bob.b...@citrix.com<mailto:bob.b...@citrix.com>) can help you more with this.

Best regards,
Martins


On 2015.07.24. 05:03, Leandro Mendes wrote:
Hi Guys,

I'm trying to setup a 3 node (controller, neutron and compute) Openstack setup, 
but i'm having a lot of trouble regarding the XenServer(6.5) setup.

Looking around the internet for docs about XenServer as compute node, i've got 
the impression that the main effort of OpenStack is use KVM instead of 
XenServer, because the docs (and lots of another links outhere) are outdated.

May i consider it true? Should i forget about XenServer and move to KVM?

What do you guys think about that?

Thank you.



___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : 
openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack



___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : 
openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack


___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] XenServer 6.5 or KVM

[Bob Ball]

Hi Leandro,

It is true that the majority of development work, and the majority of 
deployments, happen with the libvirt+KVM combination however there are some 
major deployments that use XenServer highly effectively.

Regarding Neutron support, yes there are a number of known areas where we need 
more development effort to fully support this with XenServer, and these are 
ongoing.  If you're set on deploying Neutron today then you may be best served 
by looking at other hypervisors.

XenServer works easily with nova-network and, as Rackspace has proven possible, 
can also work with Neutron (just not as easily!).  If there are specific docs 
that you've identified that are out of date please let me know and we can have 
a look.

Thanks,

Bob

On 24 Jul 2015 3:14 am, Leandro Mendes  wrote:
Hi Guys,

I'm trying to setup a 3 node (controller, neutron and compute) Openstack setup, 
but i'm having a lot of trouble regarding the XenServer(6.5) setup.

Looking around the internet for docs about XenServer as compute node, i've got 
the impression that the main effort of OpenStack is use KVM instead of 
XenServer, because the docs (and lots of another links outhere) are outdated.

May i consider it true? Should i forget about XenServer and move to KVM?

What do you guys think about that?

Thank you.
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] VM_MISSING_PV_DRIVERS

[Bob Ball]

Hi Leandro,

This is most likely that your compute VM has been created as an HVM guest, 
rather than a PV guest.  This will typically happen if you use the "Other 
Install Media" template in XenServer rather than the template for the specific 
Linux version.

The error reported is that we cannot dynamically plug a disk into a VM - and 
this is most likely to be the compute VM.  Installing the PV tools into the VM 
will get you passed this particular error however you are then highly likely to 
hit https://bugs.launchpad.net/nova/+bug/1318544

The best fix is to reinstall the compute node as a PV guest.
Another approach could be to convert the compute node from HVM to PV.  Note 
that if you're using recent guests that use grub2 as the bootloader you must be 
using XenServer 6.5 as earlier versions didn't parse the grub2 config correctly 
when submenus were involved.  
http://wiki.univention.com/index.php?title=Citrix_Xen_Server#Convert_the_UCS_system_to_paravirtualization

It's also worth noting that the image you upload probably needs to be in a 
slightly different format: 
http://blogs.citrix.com/2012/10/17/upload-custom-images-to-a-xenserver-powered-openstack-cloud/
 You probably just want to extract the XVA and create a tgz as with the 
instructions on the blog page.

Thanks,

Bob


From: Leandro Mendes [theflock...@gmail.com]
Sent: 20 July 2015 22:40
To: openstack@lists.openstack.org
Subject: [Openstack] VM_MISSING_PV_DRIVERS

Hi guys,

Trying to register a CentOS7 VHD through glance, but when deploying the VM, i 
always get the follow error:

Failure: ['VM_MISSING_PV_DRIVERS', 
'OpaqueRef:9391569a-f5e1-1a34-1f90-fb7fb0983277']

I've created a VHD using a XVA template exported through XenCenter.

The command used to register the image was:

 glance image-create --name "CentOS 7 (x86_64)" --file 
 --disk-format vhd --container-format bare 
--visibility public --progress

Do you guys have any clue about what's missing?

Thanks

Leandro
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] xenserver and cinder

[Bob Ball]

I don't understand the CloudStack case for data disks; but in OpenStack you 
should look at ephemeral storage - see the flavour definition at 
http://docs.openstack.org/openstack-ops/content/flavors.html

This will allow you to have a main disk and a data disk visible to the VM; 
however it is not possible to move the data disk between two VMs - the 
ephemeral disk exists only for the life of the instance.

If you want the data disk to be portable between instances (or exist beyond the 
life of the instance) then yes, I believe you need to use Cinder and have a VM 
that provides the storage backend for Cinder.

Bob

From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com]
Sent: 14 July 2015 16:13
To: Bob Ball; openstack@lists.openstack.org
Subject: RE: xenserver and cinder

Well, we are migrating from cloudstack now and it has this feature

So, do I understand It right. In current situation I cannot use data disk's at 
local storages at all?
The only solution is to make additional layer through cinder+lvm and share 
cinder vols through iscsi?

Do you know what's the best practice for a system like mine?
I mean XS with local storages

Regards,
IT engineer
Farheap, Russia
Ivan Derbenev

From: Bob Ball [mailto:bob.b...@citrix.com]
Sent: Tuesday, July 14, 2015 5:18 PM
To: Ivan Derbenev; 
openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: RE: xenserver and cinder

Hi Ivan,

In that case then no, I don't believe what you are trying to achieve is 
currently possible with OpenStack.  My understanding is that detachable block 
storage is provided only by Cinder and Cinder must be network-attached storage 
as there is no concept of tying the storage to particular VMs.

For example, you create the cinder volumes _before_ you attach it to a virtual 
machine - so there is no knowledge of which VM or which hypervisor will be 
attaching to the volume.

Bob

From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com]
Sent: 14 July 2015 15:10
To: Bob Ball; 
openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: RE: xenserver and cinder

Ok, well...
I don't have the reason to use precisely Cinder.
What I need is the ability to create, delete and attach data volumes to 
instances. As far as I know I can't do it without Cinder.
If you can show me a path how to do it, and use local storages only - that 
would be really great

Regards,
IT engineer
Farheap, Russia
Ivan Derbenev

From: Bob Ball [mailto:bob.b...@citrix.com]
Sent: Tuesday, July 14, 2015 4:58 PM
To: Ivan Derbenev; 
openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: RE: xenserver and cinder

As I understand it Cinder provides volumes that must be accessible to VMs on 
any of the hosts - i.e. must be accessed over the network.  Therefore if you 
want to use Cinder to provide data volumes (or BFV) then you need a VM/physical 
machine/storage array providing the actual storage which is accessible over the 
network.

I'm not aware of a way to use Cinder create volumes that are only accessible to 
VMs on a single host - which is what you would need to use storage that is 
entirely local to the hypervisor.

Just to understand, why are you trying to use Cinder backed by local storage 
only?  What's the use case here?

Bob

From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com]
Sent: 14 July 2015 14:50
To: Bob Ball; 
openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: RE: xenserver and cinder

So, do I understand it right - I can't get rid of additional layer for cinder 
volumes, if I want to use local storage?

Regards,
IT engineer
Farheap, Russia
Ivan Derbenev

From: Bob Ball [mailto:bob.b...@citrix.com]
Sent: Tuesday, July 14, 2015 12:34 PM
To: Ivan Derbenev; 
openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: RE: xenserver and cinder

Hi Ivan,

XenAPINFSDriver was primarily useful for pooled scenarios (which in turn relied 
on Nova aggregates) - however it's not the easiest way to consume Cinder 
volumes.  The XenServer Nova integration supports BFV and volume attach for 
Cinder volumes presented over iSCSI, so however those are managed by Cinder is 
independent of how XenServer can consume them.

Even when using the deprecated XenAPINFSDriver, the storage had to be remote 
storage (i.e. provided to the host through an NFS server) so that does not let 
you use the hypervisor-local storage as block storage provisionable by Cinder 
and you would have needed a separate VM if you actually need your Cinder 
volumes to be provisioned from local storage.

Bob

From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com]
Sent: 13 July 2015 18:53
To: openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: [Openstack] xenserver and cinder


Hello, guys!
We are currently making an installation of XS6.5+OS Juno (maybe we'll switch to 
kilo)
We have 2

Re: [Openstack] xenserver and cinder

[Bob Ball]

Hi Ivan,

In that case then no, I don't believe what you are trying to achieve is 
currently possible with OpenStack.  My understanding is that detachable block 
storage is provided only by Cinder and Cinder must be network-attached storage 
as there is no concept of tying the storage to particular VMs.

For example, you create the cinder volumes _before_ you attach it to a virtual 
machine - so there is no knowledge of which VM or which hypervisor will be 
attaching to the volume.

Bob

From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com]
Sent: 14 July 2015 15:10
To: Bob Ball; openstack@lists.openstack.org
Subject: RE: xenserver and cinder

Ok, well...
I don't have the reason to use precisely Cinder.
What I need is the ability to create, delete and attach data volumes to 
instances. As far as I know I can't do it without Cinder.
If you can show me a path how to do it, and use local storages only - that 
would be really great

Regards,
IT engineer
Farheap, Russia
Ivan Derbenev

From: Bob Ball [mailto:bob.b...@citrix.com]
Sent: Tuesday, July 14, 2015 4:58 PM
To: Ivan Derbenev; 
openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: RE: xenserver and cinder

As I understand it Cinder provides volumes that must be accessible to VMs on 
any of the hosts - i.e. must be accessed over the network.  Therefore if you 
want to use Cinder to provide data volumes (or BFV) then you need a VM/physical 
machine/storage array providing the actual storage which is accessible over the 
network.

I'm not aware of a way to use Cinder create volumes that are only accessible to 
VMs on a single host - which is what you would need to use storage that is 
entirely local to the hypervisor.

Just to understand, why are you trying to use Cinder backed by local storage 
only?  What's the use case here?

Bob

From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com]
Sent: 14 July 2015 14:50
To: Bob Ball; 
openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: RE: xenserver and cinder

So, do I understand it right - I can't get rid of additional layer for cinder 
volumes, if I want to use local storage?

Regards,
IT engineer
Farheap, Russia
Ivan Derbenev

From: Bob Ball [mailto:bob.b...@citrix.com]
Sent: Tuesday, July 14, 2015 12:34 PM
To: Ivan Derbenev; 
openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: RE: xenserver and cinder

Hi Ivan,

XenAPINFSDriver was primarily useful for pooled scenarios (which in turn relied 
on Nova aggregates) - however it's not the easiest way to consume Cinder 
volumes.  The XenServer Nova integration supports BFV and volume attach for 
Cinder volumes presented over iSCSI, so however those are managed by Cinder is 
independent of how XenServer can consume them.

Even when using the deprecated XenAPINFSDriver, the storage had to be remote 
storage (i.e. provided to the host through an NFS server) so that does not let 
you use the hypervisor-local storage as block storage provisionable by Cinder 
and you would have needed a separate VM if you actually need your Cinder 
volumes to be provisioned from local storage.

Bob

From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com]
Sent: 13 July 2015 18:53
To: openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: [Openstack] xenserver and cinder


Hello, guys!
We are currently making an installation of XS6.5+OS Juno (maybe we'll switch to 
kilo)
We have 2 servers, controller has glance, nova and keystone, and compute vms 
have only nova-compute installed.
We use local storage on each server, not shared one (this is important)

We want to implement cinder service to manage volumes easily, but i can't 
understand how can we do it.XenAPINFSDriver is deprecated and isn't supported 
any more.
The only solution i found so far is to create storage VM, give it some space 
and use this space for cinder volumes, and then mount it in VMs as ISCSI 
targets.
And in the same time nova uses xenAPI to create and manage volumes when it 
creates instances.

And what i want to make is to make cinder use Xenserver volumes both when nova 
creates VM and when i create volume with cinder.
Without second level of abstraction.

Can you tell me what are the best practices to use cinder for Xenservers with 
local storage?


Regards,
Ivan Derbenev

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] xenserver and cinder

[Bob Ball]

As I understand it Cinder provides volumes that must be accessible to VMs on 
any of the hosts - i.e. must be accessed over the network.  Therefore if you 
want to use Cinder to provide data volumes (or BFV) then you need a VM/physical 
machine/storage array providing the actual storage which is accessible over the 
network.

I'm not aware of a way to use Cinder create volumes that are only accessible to 
VMs on a single host - which is what you would need to use storage that is 
entirely local to the hypervisor.

Just to understand, why are you trying to use Cinder backed by local storage 
only?  What's the use case here?

Bob

From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com]
Sent: 14 July 2015 14:50
To: Bob Ball; openstack@lists.openstack.org
Subject: RE: xenserver and cinder

So, do I understand it right - I can't get rid of additional layer for cinder 
volumes, if I want to use local storage?

Regards,
IT engineer
Farheap, Russia
Ivan Derbenev

From: Bob Ball [mailto:bob.b...@citrix.com]
Sent: Tuesday, July 14, 2015 12:34 PM
To: Ivan Derbenev; 
openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: RE: xenserver and cinder

Hi Ivan,

XenAPINFSDriver was primarily useful for pooled scenarios (which in turn relied 
on Nova aggregates) - however it's not the easiest way to consume Cinder 
volumes.  The XenServer Nova integration supports BFV and volume attach for 
Cinder volumes presented over iSCSI, so however those are managed by Cinder is 
independent of how XenServer can consume them.

Even when using the deprecated XenAPINFSDriver, the storage had to be remote 
storage (i.e. provided to the host through an NFS server) so that does not let 
you use the hypervisor-local storage as block storage provisionable by Cinder 
and you would have needed a separate VM if you actually need your Cinder 
volumes to be provisioned from local storage.

Bob

From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com]
Sent: 13 July 2015 18:53
To: openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: [Openstack] xenserver and cinder


Hello, guys!
We are currently making an installation of XS6.5+OS Juno (maybe we'll switch to 
kilo)
We have 2 servers, controller has glance, nova and keystone, and compute vms 
have only nova-compute installed.
We use local storage on each server, not shared one (this is important)

We want to implement cinder service to manage volumes easily, but i can't 
understand how can we do it.XenAPINFSDriver is deprecated and isn't supported 
any more.
The only solution i found so far is to create storage VM, give it some space 
and use this space for cinder volumes, and then mount it in VMs as ISCSI 
targets.
And in the same time nova uses xenAPI to create and manage volumes when it 
creates instances.

And what i want to make is to make cinder use Xenserver volumes both when nova 
creates VM and when i create volume with cinder.
Without second level of abstraction.

Can you tell me what are the best practices to use cinder for Xenservers with 
local storage?


Regards,
Ivan Derbenev

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] xenserver and cinder

[Bob Ball]

Hi Ivan,

XenAPINFSDriver was primarily useful for pooled scenarios (which in turn relied 
on Nova aggregates) - however it's not the easiest way to consume Cinder 
volumes.  The XenServer Nova integration supports BFV and volume attach for 
Cinder volumes presented over iSCSI, so however those are managed by Cinder is 
independent of how XenServer can consume them.

Even when using the deprecated XenAPINFSDriver, the storage had to be remote 
storage (i.e. provided to the host through an NFS server) so that does not let 
you use the hypervisor-local storage as block storage provisionable by Cinder 
and you would have needed a separate VM if you actually need your Cinder 
volumes to be provisioned from local storage.

Bob

From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com]
Sent: 13 July 2015 18:53
To: openstack@lists.openstack.org
Subject: [Openstack] xenserver and cinder


Hello, guys!
We are currently making an installation of XS6.5+OS Juno (maybe we'll switch to 
kilo)
We have 2 servers, controller has glance, nova and keystone, and compute vms 
have only nova-compute installed.
We use local storage on each server, not shared one (this is important)

We want to implement cinder service to manage volumes easily, but i can't 
understand how can we do it.XenAPINFSDriver is deprecated and isn't supported 
any more.
The only solution i found so far is to create storage VM, give it some space 
and use this space for cinder volumes, and then mount it in VMs as ISCSI 
targets.
And in the same time nova uses xenAPI to create and manage volumes when it 
creates instances.

And what i want to make is to make cinder use Xenserver volumes both when nova 
creates VM and when i create volume with cinder.
Without second level of abstraction.

Can you tell me what are the best practices to use cinder for Xenservers with 
local storage?


Regards,
Ivan Derbenev

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Running Xen VMs on Openstack

[Bob Ball]

Hi Girija,

I can’t help with converting XenServer VMs to KVM but can I ask why running 
XenServer under your OpenStack cloud isn’t considered an option?

Clearly there are some differences with VM setup (e.g. the VM may be set up to 
use a specific disk name which won’t be present if you move to use the virtio 
interface) that might make it difficult to move an existing VM between any two 
hypervisors.  As such it may well be easier to use XenServer in your OpenStack 
infrastructure?

You could also have a mixed-hypervisor environment if that would meet your 
requirements.

Thanks,

Bob

From: Girija Sharan [mailto:girijasharansi...@gmail.com]
Sent: 21 May 2015 06:26
To: openstack@lists.openstack.org
Subject: [Openstack] Running Xen VMs on Openstack

HI All,
It's urgent and a interesting one too.
I have to run VMs from Xen Server on Openstack (with KVM).
In detail it is like this :
1. I have a Xen Server and couple of large VMs (with GBs of data) are running 
on that.
2. I have a Openstack multi-node setup with Nova configured to use KVM-Qemu as 
a hypervisor.
3. Now I have to run those Xen VMs (as it is, without loosing any data) on this 
Openstack setup.
4. I don't want to change Openstack configuration to let Nova use Xen 
hypervisor instead of KVM.
5. I tried launching VMs on Openstack using images of VMs running on Xen 
Server, but that is failing.
What is the way to run Xen server's VMs on KVM machine (basically on Openstack) 
?
What all I have to do to acheive this?
Any quick help or guidance or pointers will be really appreciated.
Thanks in advance. Waiting for early replies.

Thanks and Regards,
Girija Sharan Singh
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Does Xenserver support cinder volume?

[Bob Ball]

Hi,

The XenAPINFS driver has indeed been deprecated from Cinder and is not 
currently under development.  This driver was known broken (some required 
functionality was missing) and only useful in a very specific use case.

The XenAPI nova driver can, of course, attach Cinder drives exposed over iSCSI, 
so the XenAPINFS driver is not needed for generic integration.

Bob


From: 한승진 [yongi...@gmail.com]
Sent: 10 February 2015 09:07
To: openstack@lists.openstack.org
Subject: [Openstack] Does Xenserver support cinder volume?


I noticed that XenAPINFSdriver is not included in cinder trunk any longer.

Xen storage manager volume driver is introduced a couple year ago. However, 
there are no contribution codes though.

Does someone plan to develop this driver?
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Can't install domU on Xenserver

[Bob Ball]

Hi John,

That’s a very interesting error – one I’ve not seen before personally.

A quick way to get up and running is to use the nightly XVAs we build from 
devstack, see http://downloads.vmd.citrix.com/OpenStack/.  The blog explaining 
how to use these is at  
http://blogs.citrix.com/2013/10/16/devstack-virtual-appliance-for-xenserver/

In terms of the error listed, what’s the status of the XenServer?  A large 
output from the log would be useful to help me understand which steps to try 
next.

I guess the first question would be what is the output of the kpartx command if 
you run it manually?

Another instance of this error at 
https://lists.launchpad.net/openstack/msg16989.html suggests that the error is 
that the installation of Ubuntu failed in a way we weren’t expecting, which 
means that there are no partitions at all on the device.

The full log of install_os_domU.sh and your localrc would be a useful starting 
place to debug what’s going on here.  And, of course, this debugging can happen 
in parallel with using the XVAs described above to get you running sooner.

Thanks,

Bob

From: 한승진 [mailto:yongi...@gmail.com]
Sent: 05 January 2015 14:57
To: openstack@lists.openstack.org
Subject: [Openstack] Can't install domU on Xenserver

First of all I 'd like to wish to everyone of you happy new year!

I've been trying to install Openstack with XenServer.

Now, I'm using devstack for install Openstack on XenServer.

I followed this devstack xenserver guide in github.

https://github.com/openstack-dev/devstack/blob/master/tools/xen/README.md


However, I'm faced against this below error exception whenever I run 
install_os_domU.sh.

==
+++ xe vbd-list --minimal params=device 
uuid=19641d6d-cdb3-a8e3-5777-11b8baa8b248
++ 
dev=sm/backend/dee61403-815a-b970-9f28-cf9ba2efc910/ff2efba1-fdc9-4924-991e-e7b3ae62fd7a
++ [[ 
sm/backend/dee61403-815a-b970-9f28-cf9ba2efc910/ff2efba1-fdc9-4924-991e-e7b3ae62fd7a
 =~ sm/ ]]
++ DEBIAN_FRONTEND=noninteractive
++ apt-get --option Dpkg::Options::=--force-confold --assume-yes install kpartx
++ true
+++ kpartx -av 
/dev/sm/backend/dee61403-815a-b970-9f28-cf9ba2efc910/ff2efba1-fdc9-4924-991e-e7b3ae62fd7a
+++ sed -ne 's,^add map \([a-z0-9\-]*\).*$,\1,p'
+++ sed -ne 's,^\(.*1\)$,\1,p'
++ mapping=
++ '[' -z '' ']'
++ echo 'Failed to find mapping'
++ exit -1
+ mount_device='Failed to find mapping'
+ STAGING_DIR=
==


There is an old mailing list related to this issue which is saying that kpartx 
cannot mount disk of domU in the dom0 on Ubuntu OS.

http://www.gossamer-threads.com/lists/xen/api/276203

The source code of devstack is still same since that time.

Would anyone have good solution on this?

Please help..

Thanks.

--John Haan




___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Is it possible to use Xenserver with two virtual machines as Compute Node and Network node to set up openstack + neutron network?

[Bob Ball]

Hi Dhanesh,

Yes, this should be possible – there is no additional coupling between the 
compute node and the network node when running with XenServer and as such they 
can easily be running as two separate virtual machines.

In terms of the setup, you might want to look at 
http://blogs.citrix.com/2013/06/14/openstack-networking-quantum-on-xenserver-from-notworking-to-networking/
 which has a useful deployment diagram and should help explain what can be 
separated in which way.

Bob

From: dhanesh1212121212 [mailto:dhanesh1...@gmail.com]
Sent: 11 December 2014 08:07
To: openstack@lists.openstack.org
Subject: [Openstack] Is it possible to use Xenserver with two virtual machines 
as Compute Node and Network node to set up openstack + neutron network?

Hi All,


Is it possible to use Xenserver with  two virtual machines as Compute Node and 
Network node?
Will it effect the performance? or the set up in any ways?



Regards,
Dhanesh M.
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] openstack + libvirt + xl for xen

[Bob Ball]

Hi Xing,

It may be that the dom0 qemu helper process is not running.  It seems a very 
similar issue was tracked down to this on the thread at 
http://lists.xen.org/archives/html/xen-users/2014-05/msg00134.html

Ian C pointed out that this helper process is normally started by the 
xencommons script.

I hope this solves the issue.

Thanks,

Bob

From: Xing Lin [mailto:linxing...@gmail.com]
Sent: 05 November 2014 21:38
To: openstack@lists.openstack.org
Subject: [Openstack] openstack + libvirt + xl for xen

Hi,

I am aware that Xen via libvirt is in the group C support but since I am not 
able to install xenserver iso at compute machines I have, I have to consider to 
use xen with libvirt. I have three nodes, each one running ubuntu 14.04. I 
follow the instruction to install juno in ubuntu 14.04 and it works (I can 
create instances from horizon) when I use kvm as the hypervisor at the compute 
node . However, if I switch to use xen as the hypervisor, it will fail to 
create instances. Note that I am able to create a xen guest with virt-install 
from command line, with the following command. So, I believe I am quite close 
to get libvirt + xl work for openstack. Any help will be highly appreciated! 
Thanks,

"$ virt-install --connect=xen:/// --name u14.04.3 --ram 1024 --disk 
u14.04.3.img,size=4 --location 
http://ftp.ubuntu.com/ubuntu/dists/trusty/main/installer-amd64/ --network 
bridge=virbr0"

libxl log when creating an instance, launched from horizon (failed).
--
libxl: debug: libxl_create.c:1342:do_domain_create: ao 0x7f4e5c001890: create: 
how=(nil) callback=(nil) poller=0x7f4e5c000c50
libxl: debug: libxl_device.c:251:libxl__device_disk_set_backend: Disk vdev=sda 
spec.backend=qdisk
libxl: debug: libxl_create.c:797:initiate_domain_create: running bootloader
libxl: debug: libxl_device.c:251:libxl__device_disk_set_backend: Disk 
vdev=(null) spec.backend=qdisk
libxl: debug: libxl_device.c:251:libxl__device_disk_set_backend: Disk vdev=xvda 
spec.backend=qdisk
libxl: debug: libxl_event.c:618:libxl__ev_xswatch_deregister: watch 
w=0x7f4e5c001e20: deregister unregistered
libxl: debug: libxl.c:2712:local_device_attach_cb: locally attaching qdisk 
/dev/xvda
libxl: error: libxl_device.c:1224:libxl__wait_for_backend: Backend 
/local/domain/0/backend/qdisk/0/51712 not ready
libxl: error: libxl_bootloader.c:405:bootloader_disk_attached_cb: failed to 
attach local disk for bootloader execution
libxl: debug: libxl_event.c:618:libxl__ev_xswatch_deregister: watch 
w=0x7f4e5c001f48: deregister unregistered
libxl: error: libxl_bootloader.c:276:bootloader_local_detached_cb: unable to 
detach locally attached disk
libxl: error: libxl_create.c:1022:domcreate_rebuild_done: cannot (re-)build 
domain: -3
libxl: debug: libxl_event.c:1591:libxl__ao_complete: ao 0x7f4e5c001890: 
complete, rc=-3
libxl: debug: libxl_create.c:1356:do_domain_create: ao 0x7f4e5c001890: 
inprogress: poller=0x7f4e5c000c50, flags=ic
libxl: debug: libxl_event.c:1563:libxl__ao__destroy: ao 0x7f4e5c001890: destroy
xc: debug: hypercall buffer: total allocations:24 total releases:24
xc: debug: hypercall buffer: current allocations:0 maximum allocations:2
xc: debug: hypercall buffer: cache current size:2
xc: debug: hypercall buffer: cache hits:20 misses:2 toobig:2
libxl: debug: libxl_create.c:1342:do_domain_create: ao 0x7f4e540083e0: create: 
how=(nil) callback=(nil) poller=0x7f4e54007fb0




libxl log when creating an instance with virt-install from command line 
(succeed).

libxl: debug: libxl_create.c:1342:do_domain_create: ao 0x7f4e641b3ba0: create: 
how=(nil) callback=(nil) poller=0x7f4e640a94d0
libxl: debug: libxl_device.c:251:libxl__device_disk_set_backend: Disk vdev=xvda 
spec.backend=unknown
libxl: debug: libxl_device.c:197:disk_try_backend: Disk vdev=xvda, backend phy 
unsuitable as phys path not a block device
libxl: debug: libxl_device.c:286:libxl__device_disk_set_backend: Disk 
vdev=xvda, using backend qdisk
libxl: debug: libxl_create.c:797:initiate_domain_create: running bootloader
libxl: debug: libxl_bootloader.c:327:libxl__bootloader_run: no bootloader 
configured, using user supplied kernel
libxl: debug: libxl_event.c:618:libxl__ev_xswatch_deregister: watch 
w=0x7f4e641cd2f8: deregister unregistered
libxl: debug: libxl_numa.c:475:libxl__get_numa_candidate: New best NUMA 
placement candidate found: nr_nodes=1, nr_cpus=8, nr_vcpus=9, free_memkb=1293
libxl: detail: libxl_dom.c:195:numa_place_domain: NUMA placement candidate with 
1 nodes, 8 cpus and 1293 KB free selected
domainbuilder: detail: xc_dom_allocate: 
cmdline="method=http://ftp.ubuntu.com/ubuntu/dists/trusty/main/installer-amd64/";,
 features="(null)"
libxl: debug: libxl_dom.c:357:libxl__build_pv: pv kernel mapped 0 path 
/var/lib/xen/virtinst-vmlinuz.mBNvPH
domainbuilder: detail: xc_dom_kernel_file: 
filename="/var/lib/xen/virtinst-vmlinuz.mBNvPH"
domainbuilder: detail: xc_dom_malloc_filemap: 5643 kB
domainbuilder: detai

Re: [Openstack] Xen as a hypervisor

[Bob Ball]

Hi Mridhul,

See 
http://docs.openstack.org/trunk/config-reference/content/introduction-to-xen.html
 for the documentation.  While it has not been updated for some time, the 
current documentation is still valid and following it will give you a working 
environment on icehouse or indeed the current Juno versions.
Specifically note the recommendation to use EXT storage (several things only 
work if you pick EXT storage) but if you pick that, it works great.

Alternatively if you're looking for a quick and easy way to set up XenServer 
with OpenStack have a look at the script at 
https://github.com/citrix-openstack/qa/blob/master/install-devstack-xen.sh 
which will set up a devstack environment on your existing XenServer 
installation (more information on this setup at 
https://wiki.openstack.org/wiki/XenServer/DevStack)

Thanks!

Bob

From: Mridhul Pax [mailto:mrid...@live.com]
Sent: 14 August 2014 08:43
To: openstack@lists.openstack.org
Subject: [Openstack] Xen as a hypervisor

Stackers - Im planning to use Citrix Xenserver as a hypervisor in my 
environment (Icehouse) .

Any one implemented this successfully  ?

Any good good documentation available for the same ?
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [openstack-dev] Devstack (icehouse) - XCP - enable live migration

[Bob Ball]

Hi Afef,

There was a regression in Icehouse that broke XenAPI aggregates.  This has been 
fixed in Juno, however we would recommend you use live migrate with block 
migration (using XCP 1.6 or XenServer 6.2 – which is Free as well now, see 
http://lists.openstack.org/pipermail/openstack-dev/2013-September/014556.html).

Thanks,

Bob

From: Afef Mdhaffar [mailto:afef.mdhaf...@gmail.com]
Sent: 14 July 2014 22:58
To: openst...@lists.launchpad.net; openstack@lists.openstack.org; OpenStack 
Development Mailing List
Subject: [openstack-dev] Devstack (icehouse) - XCP - enable live migration

Hi all,

I have installed the latest release of openstack (icehouse) via devstack.
I use XCP and would like to activate the "live migration" functionality. 
Therefore, I tried to set up the pool with creating "host aggregate".
After adding the slave compute, nova-compute does not want to start any more 
and shows the following error. Could you please help me to fix this issue.
2014-07-14 21:45:13.933 CRITICAL nova [req-c7965812-76cb-4479-8947-edd70644cd3d 
None None] AttributeError: 'Aggregate' object has no attribute 'metadetails'

2014-07-14 21:45:13.933 TRACE nova Traceback (most recent call last):
2014-07-14 21:45:13.933 TRACE nova   File "/usr/local/bin/nova-compute", line 
10, in 
2014-07-14 21:45:13.933 TRACE nova sys.exit(main())
2014-07-14 21:45:13.933 TRACE nova   File 
"/opt/stack/nova/nova/cmd/compute.py", line 72, in main
2014-07-14 21:45:13.933 TRACE nova db_allowed=CONF.conductor.use_local)
2014-07-14 21:45:13.933 TRACE nova   File "/opt/stack/nova/nova/service.py", 
line 273, in create
2014-07-14 21:45:13.933 TRACE nova db_allowed=db_allowed)
2014-07-14 21:45:13.933 TRACE nova   File "/opt/stack/nova/nova/service.py", 
line 147, in __init__
2014-07-14 21:45:13.933 TRACE nova self.manager = 
manager_class(host=self.host, *args, **kwargs)
2014-07-14 21:45:13.933 TRACE nova   File 
"/opt/stack/nova/nova/compute/manager.py", line 597, in __init__
2014-07-14 21:45:13.933 TRACE nova self.driver = 
driver.load_compute_driver(self.virtapi, compute_driver)
2014-07-14 21:45:13.933 TRACE nova   File 
"/opt/stack/nova/nova/virt/driver.py", line 1299, in load_compute_driver
2014-07-14 21:45:13.933 TRACE nova virtapi)
2014-07-14 21:45:13.933 TRACE nova   File 
"/opt/stack/nova/nova/openstack/common/importutils.py", line 50, in 
import_object_ns
2014-07-14 21:45:13.933 TRACE nova return import_class(import_value)(*args, 
**kwargs)
2014-07-14 21:45:13.933 TRACE nova   File 
"/opt/stack/nova/nova/virt/xenapi/driver.py", line 156, in __init__
2014-07-14 21:45:13.933 TRACE nova self._session = 
session.XenAPISession(url, username, password)
2014-07-14 21:45:13.933 TRACE nova   File 
"/opt/stack/nova/nova/virt/xenapi/client/session.py", line 87, in __init__
2014-07-14 21:45:13.933 TRACE nova self.host_uuid = self._get_host_uuid()
2014-07-14 21:45:13.933 TRACE nova   File 
"/opt/stack/nova/nova/virt/xenapi/client/session.py", line 140, in 
_get_host_uuid
2014-07-14 21:45:13.933 TRACE nova return aggr.metadetails[CONF.host]
2014-07-14 21:45:13.933 TRACE nova AttributeError: 'Aggregate' object has no 
attribute 'metadetails'
2014-07-14 21:45:13.933 TRACE nova

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [xs-devel] Delete the imported VM image to release space -- Sisu

[Bob Ball]

Hi Sisu,

Sorry – I understand now.  When you upload an image to glance it will get 
stored in the DevStackOSDomU by default, as that is where Swift is storing it’s 
data.
The data is likely in /opt/stack/data/swift but is likely still owned by the 
glance service account.  Unfortunately by rebooting the machine you caused 
devstack to restart.  It’s very possible that the glance image repository is 
not maintained between devstack runs – if so, the image could be ‘lost’ within 
swift.

We typically only use very small virtual machines for testing (which is the 
primary purpose of the devstack-setup) which is why we have not seen such space 
used in the DomU.

Bob

From: Sisu Xi [mailto:xis...@gmail.com]
Sent: 16 June 2014 16:36
To: xs-de...@lists.xenserver.org
Cc: openstack@lists.openstack.org
Subject: Re: [Openstack] [xs-devel] Delete the imported VM image to release 
space -- Sisu

Hi, Bob:

But when I port image into nova, I need to get it into DevStackOSDomU, right?

I just don't know why the disk is full. Here is the process:
1. port self-build image using glance image-create;
2. shutdown machine
3. boot machine up, the disk space is gone, and I don't know how to free it.

Thanks.

Sisu




On Mon, Jun 16, 2014 at 4:43 AM, Bob Ball 
mailto:bob.b...@citrix.com>> wrote:
Nothing should be downloading files to DevStackOSDomU; they get downloaded 
straight into dom0 when you create an image.

Why do you think the missing disk space is related to mysql, or to /var?

Bob

From: 
xs-devel-requ...@lists.xenserver.org<mailto:xs-devel-requ...@lists.xenserver.org>
 
[mailto:xs-devel-requ...@lists.xenserver.org<mailto:xs-devel-requ...@lists.xenserver.org>]
 On Behalf Of Sisu Xi
Sent: 14 June 2014 23:19
To: xs-de...@lists.xenserver.org<mailto:xs-de...@lists.xenserver.org>; 
openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: [xs-devel] Delete the imported VM image to release space -- Sisu

Hi, all:

I am using OpenStack with XenServer. I am using a pre-build image found in:
http://downloads.vmd.citrix.com/OpenStack/
The image file is devstack-havana.xva.

This image has a disk size of 40G, I tried to import a self-build 15G image, it 
worked just fine.

When I reboot the machine, the previously imported 15G image is gone (which is 
expected).
stack@DevStackOSDomU:~$ glance image-list
+--+-+-+--+--++
| ID   | Name| Disk 
Format | Container Format | Size | Status |
+--+-+-+--+--++
| 22bcc4ed-c412-495d-995c-bdfc571e0991 | cirros-0.3.0-x86_64-disk| vhd  
   | ovf  | 9220018  | active |
| 93aefd5f-3e75-4969-a013-e49e9c85163d | cirros-0.3.1-x86_64-uec | ami  
   | ami  | 25165824 | active |
| 4d63a0d5-91e0-4fcb-944c-9fa25e7549e8 | cirros-0.3.1-x86_64-uec-kernel  | aki  
   | aki  | 4955792  | active |
| d8c961b7-5fe1-4520-8fa3-8bba97729eeb | cirros-0.3.1-x86_64-uec-ramdisk | ari  
   | ari  | 3714968  | active |
+--+-+-+--+--++


However, the disk-space is not released. And I cannot find the previously 
imported image.
-result of df-ah-
root@DevStackOSDomU:/var/lib/mysql# df -ah
Filesystem  Size  Used Avail Use% Mounted on
/dev/xvda1   38G   33G  3.1G  92% /
proc   0 0 0- /proc
sysfs  0 0 0- /sys
none   0 0 0- /sys/fs/fuse/connections
none   0 0 0- /sys/kernel/debug
none   0 0 0- /sys/kernel/security
udev1.5G  4.0K  1.5G   1% /dev
devpts 0 0 0- /dev/pts
tmpfs   300M  288K  300M   1% /run
none5.0M 0  5.0M   0% /run/lock
none1.5G 0  1.5G   0% /run/shm
none   0 0 0- /proc/xen

I checked the /var/lib/mysql, and also tried to use "du -a /var | sort -n -r | 
head -n 10" to find the top 10 files, but still cannot find any file larger 
than 1G.

Can anyone help me with this?

Thanks very much!

Sisu


--
[https://dl.dropboxusercontent.com/u/268069/Sisu_Contact.jpeg]

Sisu Xi, PhD Candidate

http://www.cse.wustl.edu/~xis/
Department of Computer Science and Engineering
Campus Box 1045
Washington University in St. Louis
One Brookings Drive
St. Louis, MO 63130



--
[https://dl.dropboxusercontent.com/u/268069/Sisu_Contact.jpeg]

Sisu Xi, PhD Candidate

http://www.cse.wustl.edu/~xis/
Department of Computer Science and Engineering
Campus Box 1045
Washington University in St. Louis
One Brookings Drive
St. Louis, MO 63130

Re: [Openstack] [xs-devel] Delete the imported VM image to release space -- Sisu

[Bob Ball]

Nothing should be downloading files to DevStackOSDomU; they get downloaded 
straight into dom0 when you create an image.

Why do you think the missing disk space is related to mysql, or to /var?

Bob

From: xs-devel-requ...@lists.xenserver.org 
[mailto:xs-devel-requ...@lists.xenserver.org] On Behalf Of Sisu Xi
Sent: 14 June 2014 23:19
To: xs-de...@lists.xenserver.org; openstack@lists.openstack.org
Subject: [xs-devel] Delete the imported VM image to release space -- Sisu

Hi, all:

I am using OpenStack with XenServer. I am using a pre-build image found in:
http://downloads.vmd.citrix.com/OpenStack/
The image file is devstack-havana.xva.

This image has a disk size of 40G, I tried to import a self-build 15G image, it 
worked just fine.

When I reboot the machine, the previously imported 15G image is gone (which is 
expected).
stack@DevStackOSDomU:~$ glance image-list
+--+-+-+--+--++
| ID   | Name| Disk 
Format | Container Format | Size | Status |
+--+-+-+--+--++
| 22bcc4ed-c412-495d-995c-bdfc571e0991 | cirros-0.3.0-x86_64-disk| vhd  
   | ovf  | 9220018  | active |
| 93aefd5f-3e75-4969-a013-e49e9c85163d | cirros-0.3.1-x86_64-uec | ami  
   | ami  | 25165824 | active |
| 4d63a0d5-91e0-4fcb-944c-9fa25e7549e8 | cirros-0.3.1-x86_64-uec-kernel  | aki  
   | aki  | 4955792  | active |
| d8c961b7-5fe1-4520-8fa3-8bba97729eeb | cirros-0.3.1-x86_64-uec-ramdisk | ari  
   | ari  | 3714968  | active |
+--+-+-+--+--++


However, the disk-space is not released. And I cannot find the previously 
imported image.
-result of df-ah-
root@DevStackOSDomU:/var/lib/mysql# df -ah
Filesystem  Size  Used Avail Use% Mounted on
/dev/xvda1   38G   33G  3.1G  92% /
proc   0 0 0- /proc
sysfs  0 0 0- /sys
none   0 0 0- /sys/fs/fuse/connections
none   0 0 0- /sys/kernel/debug
none   0 0 0- /sys/kernel/security
udev1.5G  4.0K  1.5G   1% /dev
devpts 0 0 0- /dev/pts
tmpfs   300M  288K  300M   1% /run
none5.0M 0  5.0M   0% /run/lock
none1.5G 0  1.5G   0% /run/shm
none   0 0 0- /proc/xen

I checked the /var/lib/mysql, and also tried to use "du -a /var | sort -n -r | 
head -n 10" to find the top 10 files, but still cannot find any file larger 
than 1G.

Can anyone help me with this?

Thanks very much!

Sisu


--
[https://dl.dropboxusercontent.com/u/268069/Sisu_Contact.jpeg]

Sisu Xi, PhD Candidate

http://www.cse.wustl.edu/~xis/
Department of Computer Science and Engineering
Campus Box 1045
Washington University in St. Louis
One Brookings Drive
St. Louis, MO 63130
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] XenServer domU packages

[Bob Ball]

It seems that this was historically provided by "nova-compute-xcp" but this 
does not appear on my installation.

Given this is lacking in the Ubuntu packaging, you may need to install Nova 
from source rather than relying on the packaging.

Bob

> -Original Message-
> From: Mārtiņš Jakubovičs [mailto:mart...@vertigs.lv]
> Sent: 09 June 2014 16:02
> To: openstack@lists.openstack.org
> Subject: [Openstack] XenServer domU packages
> 
> Hello,
> 
> Which package install necesery XenServer domU compute services? If try
> to install nova-compute (ubuntu 12.04) it install nova-compute-kvm and
> libvirt packages also. I want to try with clean install as possible.
> 
> Thanks.
> 
> ___
> Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe : http://lists.openstack.org/cgi-
> bin/mailman/listinfo/openstack

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] [xs-devel] XenSever with OpenStack, Live Migration Error

[Bob Ball]

Hi Sisu,

2014-06-05 16:33:27.509 TRACE nova.api.openstack.compute.contrib.admin_actions  
 File "/opt/stack/nova/nova/virt/xenapi/vmops.py", line 1840, in 
check_can_live_migrate_source
2014-06-05 16:33:27.509 TRACE nova.api.openstack.compute.contrib.admin_actions  
   raise exception.MigrationPreCheckError(reason=msg)
2014-06-05 16:33:27.509 TRACE nova.api.openstack.compute.contrib.admin_actions
2014-06-05 16:33:27.509 TRACE nova.api.openstack.compute.contrib.admin_actions 
MigrationPreCheckError: Migration pre-check error: VM.assert_can_migrate failed

The strange thing is sometimes it will migrate, but sometimes it cannot...

Can anyone help me with this? Why would the VM.assert_can_migrate fail?

The most typical failure in assert_can_migrate is around the compatability of 
the two host CPUs.  If they are different CPUs then they most likely support 
different feature sets.  When you have a VM it can only be migrated to a host 
if that new host has _at least_ all of the features that are available on the 
current host.

Check out http://support.citrix.com/article/CTX127059 - while OpenStack does 
not use pools, live migration depends on the same things, so setting 
cupid_feature_mask may allow the migration to succeed.

If this is not the issue, check /var/log/xensource.log for details on why XAPI 
determined that the VM cannot be live migrated.

Bob
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] xenserver-core not working with openstack

[Bob Ball]

Hi Qin,

This may be related to missing plugins in dom0 - check that you have the commit 
from https://review.openstack.org/#/c/81849/

Otherwise, please upload the relevant portions of the logs to 
paste.openstack.org

Thanks,

Bob



From: Qin Jia [jiaqin1...@gmail.com]
Sent: 07 April 2014 20:57
To: openstack@lists.openstack.org
Cc: Bob Ball; Euan Harris
Subject: Re: xenserver-core not working with openstack

In addtion, there is an error message when running run.sh in domU.

Error: Service n-cpu is not running

Thanks,
Qin


On Mon, Apr 7, 2014 at 3:53 PM, Qin Jia 
mailto:jiaqin1...@gmail.com>> wrote:
Hi,

I installed xenserver-core on centos. It could start and stop instances using 
xe commands. I tried to install openstack in domU using devstack scripts. But 
it fails at
[ERROR] /opt/stack/devstack/functions-common:1164 More details about the above 
errors can be found with screen, with ./rejoin-stack.sh
with the log
 Failure: Unable to log in to XenAPI (is the Dom0 disk full?)

I tried to wget the http://XENAPI_IP, it also says 404 not found.

In addition to that, in the dom0, rotate_xen_guest_logs.sh can't work because 
there is no list_domains after I install the xenserver-core.

Is there anything I missed?

Thank you in advance!
Qin



--
Best Regards,
Qin Jia
Computer Science
Cornell University
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Turn a XenServer AllInOne Node into a Compute Node?

[Bob Ball]

Hi Sisu,

You can easily convert a VM running all services in devstack to one running 
just the services needed for compute (nova-compute and nova-network in the case 
of the stock method for setting up devstack with XenServer).

There is a script at 
https://github.com/citrix-openstack/qa/blob/master/connect-devstack-local.sh - 
download this to the devstack instance you want to just run compute services, 
and run it passing the IP address of the devstack that hosts the rest of the 
services.
If you’ve got a split of where the other services are, check out the details in 
the script – it’s actually quite simple to change hosts in the localrcs.

Note that this will not convert an “AllInOne” devstack to a standalone compute 
node – it will still be using devstack, but just configuring it to run only the 
compute services.

Thanks,

Bob

From: Sisu Xi [mailto:xis...@gmail.com]
Sent: 17 January 2014 17:31
To: openstack@lists.openstack.org
Subject: [Openstack] Turn a XenServer AllInOne Node into a Compute Node?

Hi,

I was having trouble configuring a XenServer Node into an computer Node, the 
tutorial/devstack are all for using XenServer as a single allinone node.

For example, I found this really nice VM image from citrix at:
http://downloads.vmd.citrix.com/OpenStack/
which includes pre-build XenServer patch and images. I can just import them 
into XenServer and everything works.

Is there any easy way to turn this AllInOne XenServer Node into just a compute 
node? Of course I have another machine works as the controller node, installed 
following the OpenStack installation guide.

Thanks very much!

Sisu


--
Sisu Xi, PhD Candidate

http://www.cse.wustl.edu/~xis/
Department of Computer Science and Engineering
Campus Box 1045
Washington University in St. Louis
One Brookings Drive
St. Louis, MO 63130
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] XenServer support

[Bob Ball]

> -Original Message-
> From: Mārtiņš Jakubovičs [mailto:mart...@hostnet.lv]
> Sent: 18 December 2013 11:06
> To: openstack@lists.openstack.org
> Subject: [Openstack] XenServer support
> 
> I want to create small cloud but in testing I faced with many
> issues and it looks like OpenStack + XenServer is not usable for production.

Just to point out that OpenStack+XenServer certainly is useable - and being 
aggressively used - in production.  The production cases that I'm aware of do 
not use the feature that is broken in this instance, host aggregates, which is 
also one of the least well tested areas of the integration.  

> - Can't aggregate hosts. (After patch it is going to work, but in
> XenServer second host's PV domU nova-compute won't start anyway
> https://ask.openstack.org/en/question/8392/aggregate-hosts/)

I'm in the process of trying to create a back-port for Havana that will resolve 
this (https://review.openstack.org/#/c/61712/).  Note that trunk does not 
suffer from either the original issue or the new one that cropped up.  It's 
very disappointing that we didn't have the right testing in place to ensure 
that this functionality wasn't broken for the Havana release, but I'm also 
working on that.

> - Not normal shared storage support. Isn't possible to resize instances
> with shared storage.

"resize" current expects to be moving the images from one SR to another using 
rsync.  I agree that this is something that should be addressed - could you 
raise a bug report on that and give it the "xenserver" tag?

> - Shared storage is important, without shared storage XenServer can't
> create aggregate zone (Resource pool) and use live migration.

Storage in XenServer is based on a UUID.  There is a mode in XenServer 6.2 
which can be enabled by setting "relaxed_xsm_check=true" in /etc/xapi.conf 
which will test whether the SRs are the same based on the UUID and only copy 
the VDI if it does not exist in the target SR.

If you enable this option, then you can use shared storage on your XenServer 
hosts with storage live migration, rather than needing to use the pool live 
migration which does depend on host aggregates.  This may enable you to set up 
a pool independently of host aggregates and still benefit from the increased 
speed of pool-based live migration.

Thanks,

Bob

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Problem with aggregate second host with XenServer

[Bob Ball]

Sorry – you’re right – I went too far in the history and then mis-read the 
changeset I referenced!  This was actually fixed by the same fix with the 
aggregates objects – that doesn’t cleanly backport, which is one reason I 
didn’t suggest it in the first place.

I’m not sure what the right way to fix this is for good in Havana, since I’m 
nervous about trying to backport the big changes to use objects (which was the 
correct fix in trunk).

Can you remove the change you cherry-picked  as a result of my advice on 
ask.openstack.org and cherry-pick my new commit:

git checkout HEAD^1
git fetch https://review.openstack.org/openstack/nova refs/changes/12/61712/2 
&& git cherry-pick FETCH_HEAD



From: Mārtiņš Jakubovičs [mailto:mart...@hostnet.lv]
Sent: 12 December 2013 11:27
To: Bob Ball; openstack@lists.openstack.org
Subject: Re: [Openstack] Problem with aggregate second host with XenServer

This changes is included in code, at least in my.

On 2013.12.12. 13:14, Bob Ball wrote:
This issue seems to have been fixed by https://review.openstack.org/#/c/15910 - 
so to get it fixed in your environment, try cherry-picking the change using:
git fetch https://review.openstack.org/openstack/nova refs/changes/10/15910/1 
&& git cherry-pick FETCH_HEAD

I’ll look at trying to get these two fixes backported to Havana.

Bob

From: Mārtiņš Jakubovičs [mailto:mart...@hostnet.lv]
Sent: 12 December 2013 08:08
To: openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: [Openstack] Problem with aggregate second host with XenServer

Hello!

First of all I want to say, that aggregate function isn't working at all, but 
it is solved by this post: 
https://ask.openstack.org/en/question/8392/aggregate-hosts/
After fixing sources aggregate is working now, I can add master and second host 
to pool (second host's node-compute successfully reboots and host adds to pool).
But after adding second host I got problem with second hosts nova-compute, it 
won't start:

2013-12-12 09:52:08.114 13238 INFO nova.openstack.common.periodic_task [-] 
Skipping periodic task _periodic_update_dns because its interval is negative
2013-12-12 09:52:08.166 13238 INFO nova.virt.driver [-] Loading compute driver 
'xenapi.XenAPIDriver'
2013-12-12 09:52:08.535 13238 CRITICAL nova [-] 'XenAPISession' object has no 
attribute '_virtapi'
2013-12-12 09:52:08.535 13238 TRACE nova Traceback (most recent call last):
2013-12-12 09:52:08.535 13238 TRACE nova   File "/usr/bin/nova-compute", line 
10, in 
2013-12-12 09:52:08.535 13238 TRACE nova sys.exit(main())
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/cmd/compute.py", line 68, in main
2013-12-12 09:52:08.535 13238 TRACE nova db_allowed=False)
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/service.py", line 257, in create
2013-12-12 09:52:08.535 13238 TRACE nova db_allowed=db_allowed)
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/service.py", line 139, in __init__
2013-12-12 09:52:08.535 13238 TRACE nova self.manager = 
manager_class(host=self.host, *args, **kwargs)
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/compute/manager.py", line 450, in 
__init__
2013-12-12 09:52:08.535 13238 TRACE nova self.driver = 
driver.load_compute_driver(self.virtapi, compute_driver)
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/virt/driver.py", line 1108, in 
load_compute_driver
2013-12-12 09:52:08.535 13238 TRACE nova virtapi)
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/openstack/common/importutils.py", line 
52, in import_object_ns
2013-12-12 09:52:08.535 13238 TRACE nova return 
import_class(import_value)(*args, **kwargs)
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/driver.py", line 142, in 
__init__
2013-12-12 09:52:08.535 13238 TRACE nova self._session = XenAPISession(url, 
username, password, self.virtapi)
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/driver.py", line 666, in 
__init__
2013-12-12 09:52:08.535 13238 TRACE nova self.host_uuid = 
self._get_host_uuid()
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/driver.py", line 698, in 
_get_host_uuid
2013-12-12 09:52:08.535 13238 TRACE nova aggr = 
self._virtapi.aggregate_get_by_host(
2013-12-12 09:52:08.535 13238 TRACE nova AttributeError: 'XenAPISession' object 
has no attribute '_virtapi'
2013-12-12 09:52:08.535 13238 TRACE nova

I discover, that I can't set up nova-compute on "second host" in XenServe

Re: [Openstack] Problem with aggregate second host with XenServer

[Bob Ball]

This issue seems to have been fixed by https://review.openstack.org/#/c/15910 - 
so to get it fixed in your environment, try cherry-picking the change using:
git fetch https://review.openstack.org/openstack/nova refs/changes/10/15910/1 
&& git cherry-pick FETCH_HEAD

I’ll look at trying to get these two fixes backported to Havana.

Bob

From: Mārtiņš Jakubovičs [mailto:mart...@hostnet.lv]
Sent: 12 December 2013 08:08
To: openstack@lists.openstack.org
Subject: [Openstack] Problem with aggregate second host with XenServer

Hello!

First of all I want to say, that aggregate function isn't working at all, but 
it is solved by this post: 
https://ask.openstack.org/en/question/8392/aggregate-hosts/
After fixing sources aggregate is working now, I can add master and second host 
to pool (second host's node-compute successfully reboots and host adds to pool).
But after adding second host I got problem with second hosts nova-compute, it 
won't start:

2013-12-12 09:52:08.114 13238 INFO nova.openstack.common.periodic_task [-] 
Skipping periodic task _periodic_update_dns because its interval is negative
2013-12-12 09:52:08.166 13238 INFO nova.virt.driver [-] Loading compute driver 
'xenapi.XenAPIDriver'
2013-12-12 09:52:08.535 13238 CRITICAL nova [-] 'XenAPISession' object has no 
attribute '_virtapi'
2013-12-12 09:52:08.535 13238 TRACE nova Traceback (most recent call last):
2013-12-12 09:52:08.535 13238 TRACE nova   File "/usr/bin/nova-compute", line 
10, in 
2013-12-12 09:52:08.535 13238 TRACE nova sys.exit(main())
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/cmd/compute.py", line 68, in main
2013-12-12 09:52:08.535 13238 TRACE nova db_allowed=False)
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/service.py", line 257, in create
2013-12-12 09:52:08.535 13238 TRACE nova db_allowed=db_allowed)
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/service.py", line 139, in __init__
2013-12-12 09:52:08.535 13238 TRACE nova self.manager = 
manager_class(host=self.host, *args, **kwargs)
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/compute/manager.py", line 450, in 
__init__
2013-12-12 09:52:08.535 13238 TRACE nova self.driver = 
driver.load_compute_driver(self.virtapi, compute_driver)
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/virt/driver.py", line 1108, in 
load_compute_driver
2013-12-12 09:52:08.535 13238 TRACE nova virtapi)
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/openstack/common/importutils.py", line 
52, in import_object_ns
2013-12-12 09:52:08.535 13238 TRACE nova return 
import_class(import_value)(*args, **kwargs)
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/driver.py", line 142, in 
__init__
2013-12-12 09:52:08.535 13238 TRACE nova self._session = XenAPISession(url, 
username, password, self.virtapi)
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/driver.py", line 666, in 
__init__
2013-12-12 09:52:08.535 13238 TRACE nova self.host_uuid = 
self._get_host_uuid()
2013-12-12 09:52:08.535 13238 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/driver.py", line 698, in 
_get_host_uuid
2013-12-12 09:52:08.535 13238 TRACE nova aggr = 
self._virtapi.aggregate_get_by_host(
2013-12-12 09:52:08.535 13238 TRACE nova AttributeError: 'XenAPISession' object 
has no attribute '_virtapi'
2013-12-12 09:52:08.535 13238 TRACE nova

I discover, that I can't set up nova-compute on "second host" in XenServer pool 
which is created by XenServer (without OpenStack aggregate), error is same. To 
master host nova-compute connects successfully. I tested this in two different 
XenServer 6.2 pools.
Is there some solution how can start to work "second host's" nova-computes?

Thanks.

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] devstack log files

[Bob Ball]

Yes.  You should set the SCREEN_LOGDIR and LOGFILE parameters - the former is 
for all of the services that are started and the latter is for stack.sh itself.

Bob

> -Original Message-
> From: John Smith [mailto:lbalba...@gmail.com]
> Sent: 26 November 2013 12:05
> To: openstack@lists.openstack.org
> Subject: [Openstack] devstack log files
> 
> Hi,
> 
> 
> Im running devstack, but cant seem to find the log files. I cannot
> find them in /opt/stack, nor in /var/log.
> 
> Do I need to do some configuring before devstack will create log
> files, or am i simply looking in the wrong location ?
> 
> Thanks.
> 
> 
> 
> Regards,
> 
> 
> John Smith.
> 
> ___
> Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe : http://lists.openstack.org/cgi-
> bin/mailman/listinfo/openstack

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Cannot log into openstack domu, xenserver+devstack.

[Bob Ball]

Hi Alex,

DomU's username is root or stack.  Password is the same for both and configured 
by the localrc - but if you didn't set it then it might have defaulted to 
'citrix'.

The PW you entered might just have been for the XAPI connection?

Bob

Alexander Sibetheros  wrote:



Hello,
I recently installed Xenserver6.2 on a disk, ran post install steps and then 
installed devstack(install_os_domU.sh).
I logged into the newly created domU(openstack domu, ubuntu 12.04), but do not 
know the username.
I image the password is the pw I entered in the localrc file before running the 
install script.
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Openstack and xen issues.

[Bob Ball]

Hi Alvin,

I hopethis is a configuration typo - the other-config value in XAPI is 
"ceph_sr" but the sr_matching_filter is set up to look for "ceph-sr".

I hope if you match the two this will resolve the issue.

Bob


From: Alvin Starr [al...@netvel.net]
Sent: 22 November 2013 14:31
To: Bob Ball; openstack@lists.openstack.org
Subject: Re: [Openstack] Openstack and xen issues.

I have put Openstack on a separate machine to try and separate and isolate the 
various components I need to work with in the interests of making my debugging 
easier.
This in retrospect may not have been the best idea.

I have had a very long history with xen and that may be more of an impediment 
because I think I know things about it that are no longer true.

I am using the default devstack scripts as of a few weeks ago so it should be 
grabbing the latest version of Openstack or at least that is my belief.

Here is my sr-param-list.

uuid ( RO): 7d56f548-174b-d42b-12f2-e0849588e503
  name-label ( RW): Ceph Storage
name-description ( RW):
host ( RO): localhost
  allowed-operations (SRO): unplug; plug; PBD.create; PBD.destroy; 
VDI.clone; scan; VDI.create; VDI.destroy
  current-operations (SRO):
VDIs (SRO):
PBDs (SRO): 40dd29a3-154a-e841-ce52-4547c817d856
  virtual-allocation ( RO): 348064577384
physical-utilisation ( RO): 342363992064
   physical-size ( RO): 18986006446080
type ( RO): libvirt
content-type ( RO):
  shared ( RW): true
   introduced-by ( RO): 
other-config (MRW): ceph_sr: true
   sm-config (MRO):
   blobs ( RO):
 local-cache-enabled ( RO): false
tags (SRW):


I started tracing the xenapi transactions over the network and could see the 
pool.get_all and pool.get_default when the sr_filter was not set but once I set 
it  nova would crash complaining about no repository.
I checked the TCP transactions and did not see any SR.get_all while some 
debugging prints assured me that the code was being exercised.



On 11/22/2013 04:40 AM, Bob Ball wrote:
Hi Alvin,

Yes, we typically do expect Nova to be running in a DomU.  It’s worth checking 
out 
http://docs.openstack.org/trunk/openstack-compute/install/yum/content/introduction-to-xen.html
 just to make sure you’ve got everything covered there.

I say typically because in some configurations (notably using xenserver-core) 
it may be possible to run Nova in dom0 by setting the connection URL to 
“unix://local”.  This is an experimental configuration and was added near the 
end of Havana – see https://blueprints.launchpad.net/nova/+spec/xenserver-core.

In terms of sr_matching_filter, check that you’re setting it in the right 
group.  If you’re using the latest builds of Icehouse then it should be in the 
xenserver group.  I’m also assuming that the other-config for the SR does 
indeed contain ceph-sr=true?

Is the SR that is used for VMs still the default-SR?

Thanks,

Bob

From: Alvin Starr [mailto:al...@netvel.net]
Sent: 22 November 2013 01:32
To: openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: [Openstack] Openstack and xen issues.


I am trying to use xen with Ceph and openstack using the devstack package.
I am slowly wacking my way through things and have noticed a few issues.

  1.   openstack expects to be running in a domU and generates error messages 
even if xenapi_check_host is false. I am not sure if this causes other side 
effects. The tests for the local dom0 should be completley bypassed if the 
check is disabled.
  2.  Open stack tries to read the xen SRs and checks the default one which 
ends up being the xen local storage and not any other SR. If I set the 
sr_matching_filter = other-config:ceph-sr=true there should be a xapi 
SR.get_all request generated but it looks like it is not generated at all. I 
have tracked the http traffic and no out put is generated even though the 
approprate code is being called.



--

Alvin Starr   ||   voice: (905)513-7688

Netvel Inc.   ||   Cell:  (416)806-0133

al...@netvel.net<mailto:al...@netvel.net>  ||



--
Alvin Starr   ||   voice: (905)513-7688
Netvel Inc.   ||   Cell:  (416)806-0133
al...@netvel.net<mailto:al...@netvel.net>  ||

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Openstack and xen issues.

[Bob Ball]

Hi Alvin,

Yes, we typically do expect Nova to be running in a DomU.  It's worth checking 
out 
http://docs.openstack.org/trunk/openstack-compute/install/yum/content/introduction-to-xen.html
 just to make sure you've got everything covered there.

I say typically because in some configurations (notably using xenserver-core) 
it may be possible to run Nova in dom0 by setting the connection URL to 
"unix://local".  This is an experimental configuration and was added near the 
end of Havana - see https://blueprints.launchpad.net/nova/+spec/xenserver-core.

In terms of sr_matching_filter, check that you're setting it in the right 
group.  If you're using the latest builds of Icehouse then it should be in the 
xenserver group.  I'm also assuming that the other-config for the SR does 
indeed contain ceph-sr=true?

Is the SR that is used for VMs still the default-SR?

Thanks,

Bob

From: Alvin Starr [mailto:al...@netvel.net]
Sent: 22 November 2013 01:32
To: openstack@lists.openstack.org
Subject: [Openstack] Openstack and xen issues.


I am trying to use xen with Ceph and openstack using the devstack package.
I am slowly wacking my way through things and have noticed a few issues.

  1.   openstack expects to be running in a domU and generates error messages 
even if xenapi_check_host is false. I am not sure if this causes other side 
effects. The tests for the local dom0 should be completley bypassed if the 
check is disabled.
  2.  Open stack tries to read the xen SRs and checks the default one which 
ends up being the xen local storage and not any other SR. If I set the 
sr_matching_filter = other-config:ceph-sr=true there should be a xapi 
SR.get_all request generated but it looks like it is not generated at all. I 
have tracked the http traffic and no out put is generated even though the 
approprate code is being called.



--

Alvin Starr   ||   voice: (905)513-7688

Netvel Inc.   ||   Cell:  (416)806-0133

al...@netvel.net  ||
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Ceilometer-collector does not receive anything

[Bob Ball]

Hi,

Ceilometer is currently KVM specific.

I hope this will get fixed during the icehouse cycle, however ceilometer and 
XenServer/XCP do not work together quite yet.

Bob

Martinx - ジェームズ  wrote:



Hi!

I followed this document: 
http://docs.openstack.org/havana/install-guide/install/apt/content/ch_ceilometer.html

My Ceilometer does not receives any data... Still "0 bytes" showing at the 
Admin Panel (/admin/metering)...

I installed Ceilometer endpoint / mongodb / packages in a dedicated Virtual 
Machine, not at the crowd controller, keystone users / info created okay but, 
it doesn't seems to be working as expected.

Also, I noted that nothing is changed at the "nova.conf" of the controller 
itself, only nova.conf from compute node gets updated for Ceilometer AND, the 
Network Node also haven't changed.

Let us know if it is working for you...

Best,
Thiago


On 1 November 2013 13:01, Afef MDHAFFAR 
mailto:afef.mdhaf...@gmail.com>> wrote:
Hi all,

We have launched devstack to install openstack, while including ceilometer.
We use XCP as hypervisor, Ubuntu 12.04. The used openstack version is 2014.1.

When running ceilometer-collector, consumers get registered on every topic and 
exchange, but never actually consume anything. We tested this with different 
amqp libraries, while sending messages to those topics and exchanges. Not even 
ceilometer.openstack.common.rpc.impl_kombu.BasicConsumer._callback gets called.

As we are interested in notifications from our own source, has anyone 
experienced this behaviour or a working system with notifications?

Thank you

Best regards,
Afef

Here are our configuration files:

---
ceilometer-collector.conf:
[DEFAULT]
verbose=True
auth_strategy=keystone
allow_resize_to_same_host=True
api_paste_config=/etc/nova/api-paste.ini
rootwrap_config=/etc/nova/rootwrap.conf
compute_scheduler_driver=nova.scheduler.filter_scheduler.FilterScheduler
dhcpbridge_flagfile=/etc/nova/nova.conf
force_dhcp_release=True
fixed_range=10.0.0.0/24
s3_host=ip
s3_port=
osapi_compute_extension=nova.api.openstack.compute.contrib.standard_extensions
my_ip=ip
sql_connection=mysql://user:passwd@localhost/nova?charset=utf8
libvirt_type=qemu
libvirt_cpu_mode=none
instance_name_template=instance-%08x
enabled_apis=ec2,osapi_compute,metadata
volume_api_class=nova.volume.cinder.API
state_path=/opt/stack/data/nova
lock_path=/opt/stack/data/nova
instances_path=/opt/stack/data/nova/instances
multi_host=True
send_arp_for_ha=True
logging_debug_format_suffix=^[[00;33mfrom (pid=%(process)d) %(funcName)s 
%(pathname)s:%(lineno)d^[[00m
logging_exception_prefix=%(color)s%(asctime)s TRACE %(name)s 
^[[01;35m%(instance)s^[[00m
instance_usage_audit=True
instance_usage_audit_period=hour
notification_driver=nova.openstack.common.notifier.rpc_notifier
notification_driver=ceilometer.compute.nova_notifier
notify_on_state_change=vm_and_task_state
notify_api_faults=true
network_manager=nova.network.manager.FlatDHCPManager
public_interface=eth3
vlan_interface=eth1
flat_network_bridge=xapi0
flat_interface=eth1
novncproxy_base_url=http://ip:6080/vnc_auto.html
xvpvncproxy_base_url=http://ip:6081/console
vncserver_listen=127.0.0.1
vncserver_proxyclient_address=169.254.0.1
ec2_dmz_host=ip
rabbit_host=localhost
rabbit_password=passwd
glance_api_servers=ip:9292
compute_driver=xenapi.XenAPIDriver
xenapi_connection_url=http://ip1
xenapi_connection_username=user
xenapi_connection_password=passwd
flat_injected=False
firewall_driver=nova.virt.firewall.IptablesFirewallDriver
sr_matching_filter=default-sr:true

---
ceilometer.conf
[DEFAULT]
policy_file = /etc/ceilometer/policy.json

mysql_engine=InnoDB
sqlite_db=ceilometer.sqlite
#connection=sqlite:ceilometer/openstack/common/db/$sqlite_db

backend=sqlalchemy

connection=sqlite:opt/stack/ceilometer/ceilometer/openstack/common/db/ceilometer.sqlite

http_control_exchanges=nova
http_control_exchanges=glance
#http_control_exchanges=neutron
#http_control_exchanges=cinder

pipeline_cfg_file=pipeline.yaml

sample_source=openstack

nova_control_exchange=nova
glance_control_exchange=glance

verbose=true

publish_errors=true

notification_driver=ceilometer.openstack.common.notifier.rpc_notifier2
default_notification_level=INFO

policy_file=policy.json
policy_default_rule=default

rabbit_password = passwd
rabbit_host = localhost
verbose = True
notification_topics = notifications,glance_notifications
rpc_backend = ceilometer.openstack.common.rpc.impl_kombu
auth_strategy=keystone
control_exchange=openstack

[publisher_rpc]
metering_topic=metering

[keystone_authtoken]
admin_tenant_name = demo
admin_password = passwd
admin_user = admin
auth_protocol = http

rabbit_ha_queues=false

[service_credentials]
# Username to use for openstack service access (string value)
os_username=admin

# Password to use for openstack service access (string value)
os_password=passwd

# Tenant ID to use for openstack service access (string value)
os_tenant_id=

# Te

Re: [Openstack] [openstack-dev] Unable to logging to guest console on XCP/xenserver

[Bob Ball]

The trace below seems to be wanting to access the console log, rather than the 
VNC console.

I believe there is a blog post about how to enable this, but basically you need 
a cronjob to run in dom0 to rotate the logs.  This is because the Xen console 
logging is not ring-based, and therefore without a cronjob it will fill the 
logging up.  Therefore it is not enabled by default.

To enable the console log, check out 
https://github.com/openstack/nova/blob/master/tools/xenserver/rotate_xen_guest_logs.sh
 and add crontab entries for that script as described at the top of the file.

The first invocation of the script will enable the console logs for your VMs 
and you then won't see the error log.

Thanks,

Bob


From: Rajshree Thorat [rajshree.tho...@gslab.com]
Sent: 29 October 2013 09:25
To: openstack Users; OpenStack Development Mailing List
Subject: [openstack-dev] Unable to logging to guest console on XCP/xenserver

Hi,

I am trying to use Openstack Havana to control XCP hypervisor with neutron OVS 
plugin. I can launch instances normally but unable to logging to guest console. 
Please see the below log from nova compute to get a clear idea.

Log:

2013-10-29 14:09:10.203 954 AUDIT nova.compute.manager 
[req-157fb348-69fb-4b7f-b50c-44ef85e85f11 42ffb12172244726a1a15d044167de86 
2c4acdf64eed40f7a9efcf3b7dd13259] [instance: 
cf5678ec-5284-48cc-a8d6-005eac42118e] Get console output
2013-10-29 14:09:10.296 954 ERROR nova.virt.xenapi.vmops 
[req-157fb348-69fb-4b7f-b50c-44ef85e85f11 42ffb12172244726a1a15d044167de86 
2c4acdf64eed40f7a9efcf3b7dd13259] ['XENAPI_PLUGIN_FAILURE', 'get_console_log', 
'IOError', "[Errno 2] No such file or directory: 
'/var/log/xen/guest/console.7'"]
2013-10-29 14:09:10.296 954 TRACE nova.virt.xenapi.vmops Traceback (most recent 
call last):
2013-10-29 14:09:10.296 954 TRACE nova.virt.xenapi.vmops   File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/vmops.py", line 1446, in 
get_console_output
2013-10-29 14:09:10.296 954 TRACE nova.virt.xenapi.vmops 'get_console_log', 
{'dom_id': dom_id})
2013-10-29 14:09:10.296 954 TRACE nova.virt.xenapi.vmops   File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/driver.py", line 796, in 
call_plugin
2013-10-29 14:09:10.296 954 TRACE nova.virt.xenapi.vmops host, plugin, fn, 
args)
2013-10-29 14:09:10.296 954 TRACE nova.virt.xenapi.vmops   File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/driver.py", line 851, in 
_unwrap_plugin_exceptions
2013-10-29 14:09:10.296 954 TRACE nova.virt.xenapi.vmops return func(*args, 
**kwargs)
2013-10-29 14:09:10.296 954 TRACE nova.virt.xenapi.vmops   File 
"/usr/local/lib/python2.7/dist-packages/XenAPI.py", line 229, in __call__
2013-10-29 14:09:10.296 954 TRACE nova.virt.xenapi.vmops return 
self.__send(self.__name, args)
2013-10-29 14:09:10.296 954 TRACE nova.virt.xenapi.vmops   File 
"/usr/local/lib/python2.7/dist-packages/XenAPI.py", line 133, in xenapi_request
2013-10-29 14:09:10.296 954 TRACE nova.virt.xenapi.vmops result = 
_parse_result(getattr(self, methodname)(*full_params))
2013-10-29 14:09:10.296 954 TRACE nova.virt.xenapi.vmops   File 
"/usr/local/lib/python2.7/dist-packages/XenAPI.py", line 203, in _parse_result
2013-10-29 14:09:10.296 954 TRACE nova.virt.xenapi.vmops raise 
Failure(result['ErrorDescription'])
2013-10-29 14:09:10.296 954 TRACE nova.virt.xenapi.vmops Failure: 
['XENAPI_PLUGIN_FAILURE', 'get_console_log', 'IOError', "[Errno 2] No such file 
or directory: '/var/log/xen/guest/console.7'"]
2013-10-29 14:09:10.296 954 TRACE nova.virt.xenapi.vmops
2013-10-29 14:09:10.358 954 ERROR nova.openstack.common.rpc.amqp 
[req-157fb348-69fb-4b7f-b50c-44ef85e85f11 42ffb12172244726a1a15d044167de86 
2c4acdf64eed40f7a9efcf3b7dd13259] Exception during message handling
2013-10-29 14:09:10.358 954 TRACE nova.openstack.common.rpc.amqp Traceback 
(most recent call last):
2013-10-29 14:09:10.358 954 TRACE nova.openstack.common.rpc.amqp   File 
"/usr/lib/python2.7/dist-packages/nova/openstack/common/rpc/amqp.py", line 461, 
in _process_data
2013-10-29 14:09:10.358 954 TRACE nova.openstack.common.rpc.amqp **args)
2013-10-29 14:09:10.358 954 TRACE nova.openstack.common.rpc.amqp   File 
"/usr/lib/python2.7/dist-packages/nova/openstack/common/rpc/dispatcher.py", 
line 172, in dispatch
2013-10-29 14:09:10.358 954 TRACE nova.openstack.common.rpc.amqp result = 
getattr(proxyobj, method)(ctxt, **kwargs)
2013-10-29 14:09:10.358 954 TRACE nova.openstack.common.rpc.amqp   File 
"/usr/lib/python2.7/dist-packages/nova/exception.py", line 90, in wrapped
2013-10-29 14:09:10.358 954 TRACE nova.openstack.common.rpc.amqp payload)
2013-10-29 14:09:10.358 954 TRACE nova.openstack.common.rpc.amqp   File 
"/usr/lib/python2.7/dist-packages/nova/exception.py", line 73, in wrapped
2013-10-29 14:09:10.358 954 TRACE nova.openstack.common.rpc.amqp return 
f(self, context, *args, **kw)
2013-10-29 14:09:10.358 954 TRACE nova.openstack.common.rpc.amqp   Fil

Re: [Openstack] What Exactly Flavor Work?

[Bob Ball]

Yes, just to confirm the XenAPI driver will automatically resize disks where it 
can without needing an agent / cloud-init installed in the VM.

It does this by mounting the disk in the domain running nova and resizing the 
filesystem/device.

Bob

From: Joe Topjian [mailto:joe.topj...@cybera.ca]
Sent: 09 October 2013 14:52
To: Tom Fifield
Cc: openstack@lists.openstack.org
Subject: Re: [Openstack] What Exactly Flavor Work?


My understanding is that automatic resize will occur if the following are all 
true:
* auto_disk_config=True is set as a property on the image in the Image Registry.
* The disk on the image has only one partition.
* The file system on the one partition is ext3 or ext4.

I think Xen provides some extra help in this area by doing the resize itself (I 
*think*). KVM defers the job to the user.

There was a long discussion on the openstack-operators list about building 
CentOS images:

http://www.gossamer-threads.com/lists/openstack/operators/25790

The td;dr version would be this post where I link to some scripts that are able 
to build a complete CentOS image that resizes as you would expect:

http://www.gossamer-threads.com/lists/openstack/operators/28733#28733

Hope that helps,
Joe


--
Joe Topjian
Systems Architect
Cybera Inc.

www.cybera.ca

Cybera is a not-for-profit organization that works to spur and support 
innovation, for the economic benefit of Alberta, through the use of 
cyberinfrastructure.
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Openstack XCP with OVS Quantum

[Bob Ball]

While I'm a big supporter of xenserver-core (which is the name of the packages 
Endre mentions) - the packages and their OpenStack integration is currently 
very new.  We're working through a number of issues in the packaging at the 
moment and will hopefully have a more stable point in a few weeks' time.

For the time being I would personally recommend using XenServer 6.2 - which is 
also free, unlicensed and fully open source.

On the other hand, if you chose to use xenserver-core and find issues, please 
report them on github (https://github.com/xapi-project/xenserver-core/) and 
we'll get them fixed as soon as we can - or submit a pull request if you find 
the fix yourself!

Bob

From: Endre Karlson [mailto:endre.karl...@gmail.com]
Sent: 04 October 2013 14:57
To: Rajshree Thorat
Cc: openst...@ask.openstack.org; openstack Users
Subject: Re: [Openstack] Openstack XCP with OVS Quantum

I'll suggest you to move away from XCP if you can and check out the new work 
done by Citrix to make XenServer packages / functionality available to be 
installed on stuff like CentOS 6 and Ubuntu + others.

Endre.

2013/10/4 Rajshree Thorat 
mailto:rajshree.tho...@gslab.com>>
Hi,

I am trying to use Openstack Grizzly to control XCP hypervisor with quantum ovs 
plugin. I tried with the steps which are mentioned in official document 
(http://docs.openstack.org/grizzly/openstack-compute/install/apt/content/introduction-to-xen.html#xen-config-reference)
 but it doesn't work for me. I am not able to start nova-compute service. It is 
giving me following error:

2013-10-04 18:23:44.861 32548 INFO nova.manager [-] Skipping periodic task 
_periodic_update_dns because its interval is negative
2013-10-04 18:23:44.902 32548 INFO nova.virt.driver [-] Loading compute driver 
'xenapi.XenAPIDriver'
2013-10-04 18:23:44.922 32548 CRITICAL nova [-] [Errno 111] ECONNREFUSED
2013-10-04 18:23:44.922 32548 TRACE nova Traceback (most recent call last):
2013-10-04 18:23:44.922 32548 TRACE nova   File "/usr/bin/nova-compute", line 
83, in 
2013-10-04 18:23:44.922 32548 TRACE nova db_allowed=False)
2013-10-04 18:23:44.922 32548 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/service.py", line 534, in create
2013-10-04 18:23:44.922 32548 TRACE nova db_allowed=db_allowed)
2013-10-04 18:23:44.922 32548 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/service.py", line 413, in __init__
2013-10-04 18:23:44.922 32548 TRACE nova self.manager = 
manager_class(host=self.host, *args, **kwargs)
2013-10-04 18:23:44.922 32548 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/compute/manager.py", line 353, in 
__init__
2013-10-04 18:23:44.922 32548 TRACE nova self.driver = 
driver.load_compute_driver(self.virtapi, compute_driver)
2013-10-04 18:23:44.922 32548 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/virt/driver.py", line 931, in 
load_compute_driver
2013-10-04 18:23:44.922 32548 TRACE nova virtapi)
2013-10-04 18:23:44.922 32548 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/openstack/common/importutils.py", line 
51, in import_object_ns
2013-10-04 18:23:44.922 32548 TRACE nova return 
import_class(import_value)(*args, **kwargs)
2013-10-04 18:23:44.922 32548 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/driver.py", line 139, in 
__init__
2013-10-04 18:23:44.922 32548 TRACE nova self._session = XenAPISession(url, 
username, password, self.virtapi)
2013-10-04 18:23:44.922 32548 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/driver.py", line 630, in 
__init__
2013-10-04 18:23:44.922 32548 TRACE nova url = 
self._create_first_session(url, user, pw, exception)
2013-10-04 18:23:44.922 32548 TRACE nova   File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/driver.py", line 641, in 
_create_first_session
2013-10-04 18:23:44.922 32548 TRACE nova session.login_with_password(user, pw)
2013-10-04 18:23:44.922 32548 TRACE nova   File 
"/usr/local/lib/python2.7/dist-packages/XenAPI.py", line 182, in 
2013-10-04 18:23:44.922 32548 TRACE nova return lambda *params: 
self._login(name, params)
2013-10-04 18:23:44.922 32548 TRACE nova   File 
"/usr/local/lib/python2.7/dist-packages/XenAPI.py", line 148, in _login
2013-10-04 18:23:44.922 32548 TRACE nova result = 
_parse_result(getattr(self, 'session.%s' % method)(*params))
2013-10-04 18:23:44.922 32548 TRACE nova   File 
"/usr/lib/python2.7/xmlrpclib.py", line 1224, in __call__
2013-10-04 18:23:44.922 32548 TRACE nova return self.__send(self.__name, 
args)
2013-10-04 18:23:44.922 32548 TRACE nova   File 
"/usr/lib/python2.7/xmlrpclib.py", line 1578, in __request
2013-10-04 18:23:44.922 32548 TRACE nova verbose=self.__verbose
2013-10-04 18:23:44.922 32548 TRACE nova   File 
"/usr/lib/python2.7/xmlrpclib.py", line 1264, in request
2013-10-04 18:23:44.922 32548 TRACE nova return self.single_request(host, 
handler, request_body, verbose)
2013-10-04 18:23:44.922 32548 TRACE nova   

Re: [Openstack] Openstack XCP with OVS Quantum

[Bob Ball]

The Quantum support with XenServer/XCP was only added in Havana (See 
http://blogs.citrix.com/2013/06/14/openstack-networking-quantum-on-xenserver-from-notworking-to-networking/).
  If you want to use Grizzly then I'd recommend using nova-network.

Having said that, the error below indicates that you cannot connect from your 
VM with nova-compute through to XAPI.
Check your XENAPI_CONNECTION_URL setting in nova and your host networking 
configuration.
You should be able to wget a web page from XENAPI_CONNECTION_URL from the DomU 
that you're running your nova-compute in.

Bob

> -Original Message-
> From: Rajshree Thorat [mailto:rajshree.tho...@gslab.com]
> Sent: 04 October 2013 14:44
> To: openstack Users; openst...@ask.openstack.org
> Subject: [Openstack] Openstack XCP with OVS Quantum
> 
> Hi,
> 
> I am trying to use Openstack Grizzly to control XCP hypervisor with
> quantum ovs plugin. I tried with the steps which are mentioned in
> official document
> (http://docs.openstack.org/grizzly/openstack-
> compute/install/apt/content/introduction-to-xen.html#xen-config-
> reference)
> but it doesn't work for me. I am not able to start nova-compute service.
> It is giving me following error:
> 
> 2013-10-04 18:23:44.861 32548 INFO nova.manager [-] Skipping periodic
> task _periodic_update_dns because its interval is negative
> 2013-10-04 18:23:44.902 32548 INFO nova.virt.driver [-] Loading compute
> driver 'xenapi.XenAPIDriver'
> 2013-10-04 18:23:44.922 32548 CRITICAL nova [-] [Errno 111] ECONNREFUSED
> 2013-10-04 18:23:44.922 32548 TRACE nova Traceback (most recent call last):
> 2013-10-04 18:23:44.922 32548 TRACE nova   File "/usr/bin/nova-compute",
> line 83, in 
> 2013-10-04 18:23:44.922 32548 TRACE nova db_allowed=False)
> 2013-10-04 18:23:44.922 32548 TRACE nova   File
> "/usr/lib/python2.7/dist-packages/nova/service.py", line 534, in create
> 2013-10-04 18:23:44.922 32548 TRACE nova db_allowed=db_allowed)
> 2013-10-04 18:23:44.922 32548 TRACE nova   File
> "/usr/lib/python2.7/dist-packages/nova/service.py", line 413, in __init__
> 2013-10-04 18:23:44.922 32548 TRACE nova self.manager =
> manager_class(host=self.host, *args, **kwargs)
> 2013-10-04 18:23:44.922 32548 TRACE nova   File
> "/usr/lib/python2.7/dist-packages/nova/compute/manager.py", line 353, in
> __init__
> 2013-10-04 18:23:44.922 32548 TRACE nova self.driver =
> driver.load_compute_driver(self.virtapi, compute_driver)
> 2013-10-04 18:23:44.922 32548 TRACE nova   File
> "/usr/lib/python2.7/dist-packages/nova/virt/driver.py", line 931, in
> load_compute_driver
> 2013-10-04 18:23:44.922 32548 TRACE nova virtapi)
> 2013-10-04 18:23:44.922 32548 TRACE nova   File
> "/usr/lib/python2.7/dist-
> packages/nova/openstack/common/importutils.py",
> line 51, in import_object_ns
> 2013-10-04 18:23:44.922 32548 TRACE nova return
> import_class(import_value)(*args, **kwargs)
> 2013-10-04 18:23:44.922 32548 TRACE nova   File
> "/usr/lib/python2.7/dist-packages/nova/virt/xenapi/driver.py", line 139,
> in __init__
> 2013-10-04 18:23:44.922 32548 TRACE nova self._session =
> XenAPISession(url, username, password, self.virtapi)
> 2013-10-04 18:23:44.922 32548 TRACE nova   File
> "/usr/lib/python2.7/dist-packages/nova/virt/xenapi/driver.py", line 630,
> in __init__
> 2013-10-04 18:23:44.922 32548 TRACE nova url =
> self._create_first_session(url, user, pw, exception)
> 2013-10-04 18:23:44.922 32548 TRACE nova   File
> "/usr/lib/python2.7/dist-packages/nova/virt/xenapi/driver.py", line 641,
> in _create_first_session
> 2013-10-04 18:23:44.922 32548 TRACE nova
> session.login_with_password(user, pw)
> 2013-10-04 18:23:44.922 32548 TRACE nova   File
> "/usr/local/lib/python2.7/dist-packages/XenAPI.py", line 182, in 
> 2013-10-04 18:23:44.922 32548 TRACE nova return lambda *params:
> self._login(name, params)
> 2013-10-04 18:23:44.922 32548 TRACE nova   File
> "/usr/local/lib/python2.7/dist-packages/XenAPI.py", line 148, in _login
> 2013-10-04 18:23:44.922 32548 TRACE nova result =
> _parse_result(getattr(self, 'session.%s' % method)(*params))
> 2013-10-04 18:23:44.922 32548 TRACE nova   File
> "/usr/lib/python2.7/xmlrpclib.py", line 1224, in __call__
> 2013-10-04 18:23:44.922 32548 TRACE nova return
> self.__send(self.__name, args)
> 2013-10-04 18:23:44.922 32548 TRACE nova   File
> "/usr/lib/python2.7/xmlrpclib.py", line 1578, in __request
> 2013-10-04 18:23:44.922 32548 TRACE nova verbose=self.__verbose
> 2013-10-04 18:23:44.922 32548 TRACE nova   File
> "/usr/lib/python2.7/xmlrpclib.py", line 1264, in request
> 2013-10-04 18:23:44.922 32548 TRACE nova return
> self.single_request(host, handler, request_body, verbose)
> 2013-10-04 18:23:44.922 32548 TRACE nova   File
> "/usr/lib/python2.7/xmlrpclib.py", line 1292, in single_request
> 2013-10-04 18:23:44.922 32548 TRACE nova self.send_content(h,
> request_body)
> 2013-10-04 18:23:44.922 32548 TRACE nova   File
> "/usr/lib/python2.7/xmlrpclib.py", line

Re: [Openstack] Openstack IN a Virtual Machine?

[Bob Ball]

I can't speak for KVM, but this is certainly possible using XenServer.  We've 
run nested virtualisation and nested OpenStack instances (primarily devstack 
instances for testing) many times.
There is a degradation in performance, as you would expect from nested 
virtualisation, but it's definitely functional.

Bob

From: Marco Fornaro [mailto:marco.forn...@huawei.com]
Sent: 02 October 2013 08:58
To: openstack@lists.openstack.org
Subject: [Openstack] Openstack IN a Virtual Machine?

Hi All,

I have a question on Openstack that is quite important for my cloud 
architecture:

--> Can I have a complete Openstack instance running ON a VIRTUAL MACHINE?

Or, generalizing: can I have "nested" openstack instances?

...and If this is possible...are there any constraints?

NOTES:
->I'm interested to openstack instances WITH all network features available, 
meaning Quantum server with (preferably) openwswitch plugin, so I would be 
interested in VMs with a nested openstack image that do provide or access 
quatum server features...BUT even a flat DHCP network would be better than 
nothing!
->I normally use KVM as Hypervisor


Thanks so much for any help


Best Regards

Marco Fornaro

P.S.: if this is a "newbie" question I apology, but I cannot find a clear 
answer in doc (see for instance 
https://wiki.openstack.org/wiki/HypervisorSupportMatrix) AND perhaps some 
little words could save a lot of time when trying to realize such a nested 
achitecture
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Loading Kernel Module during reboot

[Bob Ball]

It depends what type of image you are booting.

If you are using a 3-part image, where the kernel and the initrd were 
originally outside of the root disk, then I've never tested how to make the 
necessary changes so the next boot pulls the kernel/initrd out from the image.

If the original image was a single-part and the kernel/initrd were extracted 
automatically at boot time then yes, simply updating grub.conf to reference the 
new kernel+initrd should be sufficient.

Bob

From: Qin Jia [mailto:jiaqin1...@gmail.com]
Sent: 23 September 2013 05:57
To: openstack@lists.openstack.org
Subject: [Openstack] Loading Kernel Module during reboot

Hello,

I have openstack run on top of Xenserver. I have created images (initrd and 
kernel images)  inside the /boot directory of the instance. I was wondering if 
it is possible to load the image during rebooting of the instance by changing 
the grub.conf?

Thanks,
Qin
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Instance failed to spawn

[Bob Ball]

That error is down to having the networking setup incorrectly.

Check flat_interface, flat_network_bridge, public_interface and vlan_interface 
in /etc/nova/nova.conf.

These should be set to an appropriate physical device (for _interface settings) 
and XAPI network (for _bridge) settings.

In your case I'm assuming that flat_network_bridge is defined as br100, which 
does not exist as a XAPI network.

Bob

From: Rajesh Upadhayay [mailto:rupadha...@xavient.com]
Sent: 19 September 2013 10:31
To: Bob Ball; 'openstack@lists.openstack.org'
Subject: RE: Instance failed to spawn

Hi Bob,
Thanks for replying back..  I have resolved this issue by removing physical 
CD-ROM from XEN -Dom0;
But now I have a new  error

2013-09-19 14:37:52.286 ERROR nova.scheduler.filter_scheduler 
[req-0c4412e1-77dc-4c59-a869-ce337b0e1c1b 93bcfd34b6174a638d0563af20a1db4d 
e60e26b10d0348629bb0dadce475ed1
f] [instance: 06664ac2-9d2b-409e-b1f0-dc537968f0f9] Error from last host: 
cloudbox.xavient.com (node xenserver1.xavient.com): [u'Traceback (most recent 
call last):\n',
u'  File "/usr/lib/python2.7/dist-packages/nova/compute/manager.py", line 848, 
in _run_instance\nset_access_ip=set_access_ip)\n', u'  File 
"/usr/lib/python2.7/dist-
packages/nova/compute/manager.py", line 1107, in _spawn\n
LOG.exception(_(\'Instance failed to spawn\'), instance=instance)\n', u'  File 
"/usr/lib/python2.7/contextl
ib.py", line 24, in __exit__\nself.gen.next()\n', u'  File 
"/usr/lib/python2.7/dist-packages/nova/compute/manager.py", line 1103, in 
_spawn\nblock_device_info)\
n', u'  File "/usr/lib/python2.7/dist-packages/nova/virt/xenapi/driver.py", 
line 177, in spawn\nadmin_password, network_info, block_device_info)\n', u' 
 File "/usr/
lib/python2.7/dist-packages/nova/virt/xenapi/vmops.py", line 480, in spawn\n
undo_mgr.rollback_and_reraise(msg=msg, instance=instance)\n', u'  File 
"/usr/lib/python2
.7/dist-packages/nova/utils.py", line 1250, in rollback_and_reraise\n
self._rollback()\n', u'  File "/usr/lib/python2.7/contextlib.py", line 24, in 
__exit__\nsel
f.gen.next()\n', u'  File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/vmops.py", line 468, in 
spawn\nsetup_network_step(undo_mgr, vm_ref, vdis)\n', u'  File
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/vmops.py", line 142, in 
inner\nrv = f(*args, **kwargs)\n', u'  File 
"/usr/lib/python2.7/dist-packages/nova/virt/x
enapi/vmops.py", line 416, in setup_network_step\nrescue)\n', u'  File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/vmops.py", line 507, in 
_setup_vm_networki
ng\nself._create_vifs(vm_ref, instance, network_info)\n', u'  File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/vmops.py", line 1479, in 
_create_vifs\nvm_
ref=vm_ref, device=device)\n', u'  File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/vif.py", line 56, in plug\n  
  self._session, vif[\'network\'][\'bridge\'])\n
', u'  File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/network_utils.py", line 49, 
in find_network_with_bridge\nraise Exception(_(\'Found no network for bri
dge %s\') % bridge)\n', u'Exception: Found no network for bridge br100\n']



Thanks & Regards,
Rajesh Upadhayay

From: Bob Ball [mailto:bob.b...@citrix.com]
Sent: Thursday, September 19, 2013 2:59 PM
To: Rajesh Upadhayay; 'openstack@lists.openstack.org'
Subject: RE: Instance failed to spawn

Hi Rajesh,

This isn't something I've seen before - typically the time we wait for a block 
device to exist (10 seconds) is more than enough.

There might be some more information in /var/log/xensource.log, dmesg or 
further up in the compute log - where it's talking about plugging the VBDs.

Is there anything special about this device? E.g. is it a cinder device, or is 
it remote storage?

Bob

From: Rajesh Upadhayay [mailto:rupadha...@xavient.com]
Sent: 19 September 2013 08:17
To: 'openstack@lists.openstack.org'
Subject: Re: [Openstack] Instance failed to spawn

Still struggling to resolve this issue. Please help me if someone face same 
issue.

Thanks & Regards,
Rajesh Upadhayay

From: Rajesh Upadhayay
Sent: Tuesday, September 10, 2013 1:31 PM
To: openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: Instance failed to spawn

Hi,

Can someone please help me as  I have successfully setup Openstack grizzly on 
DomU but getting below error  and not able to launch instance due to Spawning 
issue.


2013-09-10 13:19:43.536 WARNING nova.scheduler.driver 
[req-8cef211b-5ea4-4657-b823-1ca967efd232 33156467441344e8817046e95f0c381c 
2

Re: [Openstack] Instance failed to spawn

[Bob Ball]

Hi Rajesh,

This isn't something I've seen before - typically the time we wait for a block 
device to exist (10 seconds) is more than enough.

There might be some more information in /var/log/xensource.log, dmesg or 
further up in the compute log - where it's talking about plugging the VBDs.

Is there anything special about this device? E.g. is it a cinder device, or is 
it remote storage?

Bob

From: Rajesh Upadhayay [mailto:rupadha...@xavient.com]
Sent: 19 September 2013 08:17
To: 'openstack@lists.openstack.org'
Subject: Re: [Openstack] Instance failed to spawn

Still struggling to resolve this issue. Please help me if someone face same 
issue.

Thanks & Regards,
Rajesh Upadhayay

From: Rajesh Upadhayay
Sent: Tuesday, September 10, 2013 1:31 PM
To: openstack@lists.openstack.org
Subject: Instance failed to spawn

Hi,

Can someone please help me as  I have successfully setup Openstack grizzly on 
DomU but getting below error  and not able to launch instance due to Spawning 
issue.


2013-09-10 13:19:43.536 WARNING nova.scheduler.driver 
[req-8cef211b-5ea4-4657-b823-1ca967efd232 33156467441344e8817046e95f0c381c 
2687d9c88b514bea9564eb682c5644a5] [inst
ance: ac0f94fc-7c6d-43d6-a49d-2958e0694e6c] Setting instance to ERROR state.
2013-09-10 13:22:40.843 ERROR nova.scheduler.filter_scheduler 
[req-973d4cd8-c722-4bb5-92b8-ced31c6ab4f0 33156467441344e8817046e95f0c381c 
2687d9c88b514bea9564eb682c5644a
5] [instance: fe706720--4498-9b57-f1a6daaeea45] Error from last host: 
xplinno1ctlr1 (node xenserver): [u'Traceback (most recent call last):\n', u'  
File "/usr/lib/p
ython2.7/dist-packages/nova/compute/manager.py", line 848, in _run_instance\n   
 set_access_ip=set_access_ip)\n', u'  File 
"/usr/lib/python2.7/dist-packages/nova/comput
e/manager.py", line 1107, in _spawn\nLOG.exception(_(\'Instance failed to 
spawn\'), instance=instance)\n', u'  File "/usr/lib/python2.7/contextlib.py", 
line 24, in
__exit__\nself.gen.next()\n', u'  File 
"/usr/lib/python2.7/dist-packages/nova/compute/manager.py", line 1103, in 
_spawn\nblock_device_info)\n', u'  File "/usr/l
ib/python2.7/dist-packages/nova/virt/xenapi/driver.py", line 177, in spawn\n
admin_password, network_info, block_device_info)\n', u'  File 
"/usr/lib/python2.7/dist-p
ackages/nova/virt/xenapi/vmops.py", line 480, in spawn\n
undo_mgr.rollback_and_reraise(msg=msg, instance=instance)\n', u'  File 
"/usr/lib/python2.7/dist-packages/nov
a/utils.py", line 1250, in rollback_and_reraise\nself._rollback()\n', u'  
File "/usr/lib/python2.7/contextlib.py", line 24, in __exit__\n
self.gen.next()\n', u'
File "/usr/lib/python2.7/dist-packages/nova/virt/xenapi/vmops.py", line 463, in 
spawn\nvdis = create_disks_step(undo_mgr, disk_image_type, image_meta)\n', 
u'  File
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/vmops.py", line 142, in 
inner\nrv = f(*args, **kwargs)\n', u'  File 
"/usr/lib/python2.7/dist-packages/nova/virt/
xenapi/vmops.py", line 346, in create_disks_step\n
block_device_info=block_device_info)\n', u'  File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/vmops.py", li
ne 317, in _create_disks\nblock_device_info=block_device_info)\n', u'  File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/vm_utils.py", line 518, in 
get_vdis_f
or_instance\ncontext, session, instance, name_label, image, image_type)\n', 
u'  File "/usr/lib/python2.7/dist-packages/nova/virt/xenapi/vm_utils.py", line 
974, in _
create_image\nimage_id, image_type)\n', u'  File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/vm_utils.py", line 919, in 
_create_cached_image\nimage_id, i
mage_type)\n', u'  File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/vm_utils.py", line 997, in 
_fetch_image\nimage_id, image_type)\n', u'  File "/usr/lib/pyt
hon2.7/dist-packages/nova/virt/xenapi/vm_utils.py", line 1209, in 
_fetch_disk_image\nwith vdi_attached_here(session, vdi_ref, 
read_only=False) as dev:\n', u'  File
"/usr/lib/python2.7/contextlib.py", line 17, in __enter__\nreturn 
self.gen.next()\n', u'  File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/vm_utils.py", line
1907, in vdi_attached_here\n_wait_for_device(dev)\n', u'  File 
"/usr/lib/python2.7/dist-packages/nova/virt/xenapi/vm_utils.py", line 1864, in 
_wait_for_device\n
_(\'Timeout waiting for device %s to be created\') % dev)\n', u'StorageError: 
Timeout waiting for device hdc to be created\n']
2013-09-10 13:22:40.878 WARNING nova.scheduler.driver 
[req-973d4cd8-c722-4bb5-92b8-ced31c6ab4f0 33156467441344e8817046e95f0c381c 
2687d9c88b514bea9564eb682c5644a5] [inst
ance: fe706720--4498-9b57-f1a6daaeea45] Setting instance to ERROR state.

root@xplinno1ctlr1:~# nova hypervisor-list
++-+
| ID | Hypervisor hostname |
++-+
| 1  | xenserver   |
++-+
root@xplinno1ctlr1:~#



Thanks & Regards,
Rajesh Upadhayay
_

Re: [Openstack] Image Resizing

[Bob Ball]

Hi,

The resizing code depends on things like a single partition, and the type of 
partition - see 
https://github.com/openstack/nova/blob/master/nova/virt/xenapi/vm_utils.py#L2042

I assume that the CentOS / Fedora images you are talking about uses a different 
base filesystem.

The other requirement is that auto_disk_config needs to be set on the image 
(see 
https://github.com/openstack/nova/blob/master/nova/virt/xenapi/vm_utils.py#L803)

Thanks,

Bob


From: Qin Jia [jiaqin1...@gmail.com]
Sent: 12 September 2013 20:22
To: openstack@lists.openstack.org
Subject: [Openstack] Image Resizing

Hello,

I have deployed an openstack on top of xenserver. I downloaded the centos / 
fedora images from Rackspace Cloud Builders Images 
(https://github.com/rackerjoe/oz-image-build). Converted it to vhd format and 
uploaded to the openstack with "glance image-create". However, I find that the 
VMs started from the image will have the same size although I assigned 
different flavors.

I tried other official images (ubuntu and fedora) and resizing works well. I 
would like to ask how could I get a resizable centos image?

Thanks,
Qin Jia
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] can instances be resized?

[Bob Ball]

https://wiki.openstack.org/wiki/HypervisorSupportMatrix  suggest that this 
feature only works with XenServer.

Bob

> -Original Message-
> From: Mark Chaney [mailto:m...@lists.macscr.com]
> Sent: 02 September 2013 14:56
> To: Bob Ball
> Cc: openstack@lists.openstack.org
> Subject: RE: [Openstack] can instances be resized?
> 
> Ok, there seems to be a lot of conflicting information out there. Some
> people are saying its possible, some people are saying the kvm gets
> destroyed first and pretty much created from scratch, and others are
> saying its possible with xenserver. FYI, I will be using KVM.
> 
> Without any crazy workarounds or custom templates, can Openstack resize
> kvm linux guests and keep data in tack? This is something I have to do
> quite often with my xen (xen.org) cluster that is managed with cloudmin,
> so its a very important feature to me that I can continue to be able to
> do it when switching to kvm/openstack.
> 
> Thanks everyone for your time!
> 
> -Mark
> 
> On 2013-09-02 03:31, Bob Ball wrote:
> > Hi Mark,
> >
> > Just to add to this - if you use XenServer as the underlying
> > hypervisor then yes, it can perform automatic resize of the partitions
> > when you resize an instance between two flavors.
> >
> > I believe you need to set auto_disk_config=True for the instance - at
> > which point the expectation is that you have  a single partition and
> > are using EXT3 or EXT4.  See
> >
> https://github.com/openstack/nova/blob/master/nova/virt/xenapi/vm_util
> s.py#L798 for the technical details.
> >
> > This occurs when you issue the nova command to resize an instance
> > between the flavors and requires you to be running XenServer with the
> > EXT3 SR type - see
> > http://docs.openstack.org/trunk/openstack-
> compute/admin/content/introduction-to-xen.html for more details on
> setting up XenServer.
> >
> > Thanks,
> >
> > Bob
> >
> >> -Original Message-
> >> From: Mark Chaney [mailto:m...@lists.macscr.com]
> >> Sent: 30 August 2013 07:15
> >> To: openstack@lists.openstack.org
> >> Subject: [Openstack] can instances be resized?
> >>
> >> I am going to be hosting mainly Ubuntu, CentOS, and some Debian
> >> instances with my openstack "cloud". Can instances be resized when it
> >> comes to their storage? Obviously not only does its disk need to be
> >> resized, but its FS as well. I am hoping openstack can take care of
> >> all
> >> of that.
> >>
> >> Thanks,
> >> Mark
> >>
> >> ___
> >> Mailing list:
> >> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> >> Post to : openstack@lists.openstack.org
> >> Unsubscribe : http://lists.openstack.org/cgi-
> >> bin/mailman/listinfo/openstack
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] can instances be resized?

[Bob Ball]

Hi Mark,

Just to add to this - if you use XenServer as the underlying hypervisor then 
yes, it can perform automatic resize of the partitions when you resize an 
instance between two flavors.

I believe you need to set auto_disk_config=True for the instance - at which 
point the expectation is that you have  a single partition and are using EXT3 
or EXT4.  See 
https://github.com/openstack/nova/blob/master/nova/virt/xenapi/vm_utils.py#L798 
for the technical details.

This occurs when you issue the nova command to resize an instance between the 
flavors and requires you to be running XenServer with the EXT3 SR type - see 
http://docs.openstack.org/trunk/openstack-compute/admin/content/introduction-to-xen.html
 for more details on setting up XenServer.

Thanks,

Bob

> -Original Message-
> From: Mark Chaney [mailto:m...@lists.macscr.com]
> Sent: 30 August 2013 07:15
> To: openstack@lists.openstack.org
> Subject: [Openstack] can instances be resized?
> 
> I am going to be hosting mainly Ubuntu, CentOS, and some Debian
> instances with my openstack "cloud". Can instances be resized when it
> comes to their storage? Obviously not only does its disk need to be
> resized, but its FS as well. I am hoping openstack can take care of all
> of that.
> 
> Thanks,
> Mark
> 
> ___
> Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe : http://lists.openstack.org/cgi-
> bin/mailman/listinfo/openstack

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Unable to see xen in controller

[Bob Ball]

I'd check what's happening from the nova side; the logs should tell you what 
it's reporting back in terms of the CPU, disk and memory.

Bob

From: Jitendra Bhaskar [mailto:jeetuind...@gmail.com]
Sent: 30 August 2013 12:07
To: openstack@lists.openstack.org
Subject: Re: [Openstack] Unable to see xen in controller

Can anybody let me know hot to proceed ?

On Fri, Aug 23, 2013 at 5:17 PM, Jitendra Bhaskar 
mailto:jeetuind...@gmail.com>> wrote:
Hi All,
I have openstack(Grizzly) setup with two compute , one network(quantum) and one 
controller. I configured one more compute over XCP(1.6). Able to see in control 
with commant : nova hypervisor-list, but I am not getting correct vcpus, 
diskspace and memory info.
Attached file contains : select * from compute_nodes; from nova db .

Nova.conf on xen-compute node

[DEFAULT]
dhcpbridge_flagfile=/etc/nova/nova.conf
dhcpbridge=/usr/bin/nova-dhcpbridge
logdir=/var/log/nova
state_path=/var/lib/nova
lock_path=/var/lock/nova
force_dhcp_release=True
iscsi_helper=tgtadm
libvirt_use_virtio_for_bridges=True
#connection_type=libvirt
root_helper=sudo nova-rootwrap /etc/nova/rootwrap.conf
verbose=True
ec2_private_dns_show_ip=True
api_paste_config=/etc/nova/api-paste.ini
volumes_path=/var/lib/nova/volumes
enabled_apis=ec2,osapi_compute,metadata
libvirt_type=xen

# General
verbose=True
my_ip=10.10.10.13
rabbit_host=10.10.10.10
rabbit_password=password

auth_strategy=keystone
ec2_host=10.10.10.10
ec2_url=http://10.10.10.10:8773/services/Cloud

# Networking
#libvirt_use_virtio_for_bridges=True
network_api_class=nova.network.quantumv2.api.API
quantum_url=http://10.10.10.10:9696
quantum_auth_strategy=keystone
quantum_admin_tenant_name=service
quantum_admin_username=quantum
quantum_admin_password=password
quantum_admin_auth_url=http://10.10.10.10:35357/v2.0
xenapi_ovs_integration_bridge=br-int

# Security Groups
firewall_driver=nova.virt.firewall.NoopFirewallDriver
security_group_api=quantum

# Compute
compute_driver=xenapi.XenAPIDriver
xenapi_connection_url=http://192.168.1.21
xenapi_connection_username=root
xenapi_connection_password=pramati123
connection_type=xapi

# Cinder
volume_api_class=nova.volume.cinder.API

# Glance
glance_api_servers=10.10.10.10:9292
image_service=nova.image.glance.GlanceImageService

# novnc
vnc_enabled=true
vncserver_proxyclient_address=10.10.10.13
novncproxy_base_url=http://192.168.2.200:6080/vnc_auto.html
vncserver_listen=0.0.0.0


--
 Thanks & Regards
Jitendra




--
 Thanks & Regards
Jitendra Kumar Bhaskar
Cell:- +91 7306311531
  :- +91 9989743042

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Openstack in centos and Xen sever- Single node setup

[Bob Ball]

Just for completeness, a little explanation.

The XenServer installation for devstack (assuming you used 
https://raw.github.com/citrix-openstack/qa/master/install-devstack-xen.sh) does 
use a combination of Nova master and fixes needed to run master on XenServer 
(all of which changes - if there are any - are pending review).

This is through the settings:
NOVA_REPO=https://github.com/citrix-openstack/nova.git
NOVA_BRANCH=build

If you want to install Nova from master then either edit the script to remove 
those lines, or run the script without any parameters and it will generate a 
localrc.template which can be edited to remove the lines.

We've recently added the unit tests to the tests that we run in our CI (rather 
than relying on the unit tests run through Jenkins) - so hopefully the issue 
you encountered won't occur again.

On the other hand, there have been periods where nova master does not work 
correctly with XenServer without these fix-ups, so either approach has a risk 
since it's not the stable release.

Bob

From: Christopher Lefelhocz [mailto:christopher.lefel...@rackspace.com]
Sent: 28 August 2013 04:54
To: Bob Ball; SriHarish Ramesh; openstack@lists.openstack.org
Subject: Re: [Openstack] Openstack in centos and Xen sever- Single node setup

To Bob's point, I was able to follow these instructions in the past week and 
successfully get XenServer 6.2 up and running.  The only caution I would give 
is that buried in there is a link to "XenServer and DevStack".  While it does 
build a working running devstack, the branch of nova is not the master.  When I 
ran the unit testing on the branch many unit tests failed.  It's a relatively 
easy fix to move to the master, which to this point has not had any issues for 
me (though my testing has been limited).

Christopher

From: Bob Ball mailto:bob.b...@citrix.com>>
Date: Tuesday, August 27, 2013 1:33 AM
To: SriHarish Ramesh mailto:eie.har...@gmail.com>>, 
"openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>" 
mailto:openstack@lists.openstack.org>>
Subject: Re: [Openstack] Openstack in centos and Xen sever- Single node setup

The guide for setting up OpenStack with XenServer is at 
http://docs.openstack.org/trunk/openstack-compute/admin/content/introduction-to-xen.html.

I would personally recommend using XenServer 6.2 as that is fully open source 
and includes features such as the live storage migration in the free version.  
XenServer 5.6 is now very dated.

The "No host is available" is simply a message that says the controller node 
has not been set up properly - although it's not clear from your nova.conf 
snippet why that is.  You'll need to check the output produced by Nova to see 
what the error in connecting was.  It might be to do with network setup, or the 
username/password details.

Unfortunately I don't have experience of RDO but I would hope that the nova 
output is redirected to a log file, possibly in /var/log.

Thanks,

Bob

From: SriHarish Ramesh [mailto:eie.har...@gmail.com]
Sent: 26 August 2013 13:29
To: openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: [Openstack] Openstack in centos and Xen sever- Single node setup

Hi

I have installed XEN 5.6 in a physical machine and Openstack(Grizzly)in CentOS 
VM separately following RDO guide.
I have changed the answer file and updated the xen machine details. xen machine 
and centos vm are in the same network.




compute_driver=xenapi.XenAPIDriver

xenapi_connection_url=http://10.0.1.23

xenapi_connection_username=root

xenapi_connection_password=password
I tried restarting the centos machine and via dashboard when trying to create a 
VM, "No host is available" message is thrown.
I managed to deploy openstack in centos, but not  able to add host to the 
openstack.
Is there a step by step guide to configure the host to openstack?

--
Thanks and Regards
SRIHARISH
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Dealing with Out of Date wiki pages

[Bob Ball]

> -Original Message-
> From: Daniel P. Berrange [mailto:berra...@redhat.com]
> Sent: 28 August 2013 10:27
> 
> On Wed, Aug 28, 2013 at 01:06:01PM +1000, Tom Fifield wrote:
> > TL;DR - Obsolete page on the wiki? Consider moving it to the
> > Obsolete namespace.
> > Alternative proposals and thoughts very welcome!!
> 
> So I think a better option is to define an "Obsolete Content"
> template which would add a prominent banner across the top of the
> page which warns users that the content is out of date and asks
> for contributions to update it. 

I agree that using a template in this way would be better than moving pages to 
the Obsolete namespace.

I do think that we should have some automatic process to mark pages as possibly 
obsolete as well.

Perhaps we could make better use of the categories?  My thoughts are having a 
category for pages that will automatically be marked as obsolete, a category 
for potentially obsolete pages, one for obsolete and another for pages which 
will never be obsolete, or pages kept in that state for the archive (e.g. 
blueprints).

Each page could also have a designated owner who can be contacted when a page 
is being considered for moving to the obsolete category - even if it's just a 
comment at the bottom of the page, or another category - to give the human 
review element.

I'm not very familiar with mediawiki - but perhaps templates would be better 
used for all of the above.

Bob

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Openstack in centos and Xen sever- Single node setup

[Bob Ball]

The guide for setting up OpenStack with XenServer is at 
http://docs.openstack.org/trunk/openstack-compute/admin/content/introduction-to-xen.html.

I would personally recommend using XenServer 6.2 as that is fully open source 
and includes features such as the live storage migration in the free version.  
XenServer 5.6 is now very dated.

The "No host is available" is simply a message that says the controller node 
has not been set up properly - although it's not clear from your nova.conf 
snippet why that is.  You'll need to check the output produced by Nova to see 
what the error in connecting was.  It might be to do with network setup, or the 
username/password details.

Unfortunately I don't have experience of RDO but I would hope that the nova 
output is redirected to a log file, possibly in /var/log.

Thanks,

Bob

From: SriHarish Ramesh [mailto:eie.har...@gmail.com]
Sent: 26 August 2013 13:29
To: openstack@lists.openstack.org
Subject: [Openstack] Openstack in centos and Xen sever- Single node setup

Hi

I have installed XEN 5.6 in a physical machine and Openstack(Grizzly)in CentOS 
VM separately following RDO guide.
I have changed the answer file and updated the xen machine details. xen machine 
and centos vm are in the same network.




compute_driver=xenapi.XenAPIDriver

xenapi_connection_url=http://10.0.1.23

xenapi_connection_username=root

xenapi_connection_password=password
I tried restarting the centos machine and via dashboard when trying to create a 
VM, "No host is available" message is thrown.
I managed to deploy openstack in centos, but not  able to add host to the 
openstack.
Is there a step by step guide to configure the host to openstack?

--
Thanks and Regards
SRIHARISH
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Adding Ubuntu Images to Xenserver Powered Devstack

[Bob Ball]

Hi Qin,

(Moved openstack@lists.openstack.org<mailto:openstack@lists.openstack.org> to 
BCC while we track down the problem - I hope to post the resolution when it is 
identified)

I see - it shouldn't get shut off, that's strange - sorry for not spotting that 
in your original email.

If you could upload the nova log file for the VM booting, and possibly a 
bugtool from the XenServer host (generated using xen-bugtool -y) somewhere so 
we can have a look at what happened?

Specifically we're interested in whether there were any errors reported in the 
nova process (n-cpu if you're using devstack) and what was reported from the 
xensource.log files at the time this occurred.

Thanks,

Bob


From: Qin Jia [mailto:jiaqin1...@gmail.com]
Sent: 06 August 2013 15:28
To: Bob Ball
Cc: openstack@lists.openstack.org
Subject: Re: [Openstack] Adding Ubuntu Images to Xenserver Powered Devstack

Thank you Bob!

However, it doesn't help to wait for 5 minutes or inject credentials to the 
instance. It still got automatically shut-off at first, and then after I 
restarted it at the xenserver command or xencenter, it got stuck at that that 
message and never changed states for hours.

Do you mean the three-image the machine image/kernel/initrd?  I also downloaded 
the machine image and kernel in 
http://cloud-images.ubuntu.com/precise/current/. There is no initrd image in 
precise. I added the ami image with the kernel and without aki image. But I 
encountered the same problem as I described before.

Thanks again for your prompt response!

Qin

On Tue, Aug 6, 2013 at 4:32 AM, Bob Ball 
mailto:bob.b...@citrix.com>> wrote:
The qcow2 images provided by Ubuntu aren't set up correctly for the Xen console 
- this means that the output during the boot process doesn't go to the screen.

If you leave the image it should eventually (after perhaps 5 minutes on first 
boot) reach the login prompt.

If you have injected credentials then you should be able to login using them, 
however I believe that if you have left it to Ubuntu to generate a random 
password this is displayed on the screen at a point which is not visible 
through the VNC console.

The three-part images do not suffer from this, and are set up correctly for the 
Xen console.

Thanks,

Bob

From: Qin Jia [mailto:jiaqin1...@gmail.com<mailto:jiaqin1...@gmail.com>]
Sent: 06 August 2013 04:24
To: openstack@lists.openstack.org<mailto:openstack@lists.openstack.org>
Subject: [Openstack] Adding Ubuntu Images to Xenserver Powered Devstack

Hi,

I installed devstack on top of xenserver. Now I want to add a ubuntu image to 
it. I downloaded qcow2 image from here ( 
http://cloud-images.ubuntu.com/precise/current/) and used the guideline here 
(http://blogs.citrix.com/2012/10/17/upload-custom-images-to-a-xenserver-powered-openstack-cloud/)
 to convert the image to vhd format and add to openstack using glance.

But when I start the instance, it got shut-off automatically. When I try to 
restart the instance in Xensenver, it got stuck at the messages:

[ 1.612617] EXT4-fs (xvda1): mounted filesystem with ordered data mode. Opts: 
(null)

[ 2.245259] EXT4-fs (xvda1): re-mounted. Opts: (null)

One thing to mention is that I successfully added a cirros image to xenserver 
using the steps above.

So what's the problem here? Could any one give any suggestions?

Thanks in advance!

--
Best Regards,
Qin Jia
Computer Science
Cornell University



--
Best Regards,
Qin Jia
Computer Science
Cornell University
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Cinder basic questions

[Bob Ball]

The xentools (which can be more restrictively referred to as the guest agent) 
are usually independent of the PV drivers.  Most modern linux kernels include 
PV drivers - see http://wiki.xen.org/wiki/Xen_Linux_PV_on_HVM_drivers, so if 
you boot them in HVM mode they will negotiate up to use the PV drivers.  This 
isn't true for Windows and therefore you have to install the drivers on Windows 
- which are provided packaged with the guest agent too. 

There are many ways to confirm that the guest is using PV drivers - such as 
running "dmesg | egrep -i 'xen|front'.  This should then show lines such as 
"Initialising Xen virtual ethernet driver".

If you have any questions, let us know!

Thanks,

Bob

> -Original Message-
> From: Bryan Solie [mailto:bso...@gmail.com] On Behalf Of Bryan Solie
> Sent: 06 August 2013 14:30
> To: Bob Ball
> Cc: openstack@lists.openstack.org; #OpenStack External Email
> Subject: Re: [Openstack] Cinder basic questions
> 
> Thanks Bob.
> 
> I will monkey with the iscsi a little more to understand how the xenserver
> relationship works.  Thanks for the leads.
> 
> But one followup question: So all I need is xentools installed on the guest
> for it to be considered paravirtualized?
> 
> Bryan
> 
> On Aug 6, 2013, at 8:47 AM, Bob Ball  wrote:
> 
> > Hi Bryan,
> >
> > The volumes should appear in the guest under /dev/xvd when they
> are added.
> >
> > As long as the guest has PV drivers it should work - so HVM guests (such as
> Windows or if you have installed Linux as HVM) with PV drivers will also see
> these volumes added (although clearly with Windows they don't show up
> under /dev!)
> >
> > If using iSCSI then the target should appear on the xenserver host as a PBD
> - so "xe pbd-list" should show one up.  The code that creates this is
> https://github.com/openstack/nova/blob/master/nova/virt/xenapi/volume
> _utils.py#L51 - although that's not too obvious because we actually create a
> XenServer SR and allow that to create the iscsi connection.
> >
> > You're right that xensm isn't supported in the current code and the docs
> need to be updated for Havana, but you don't have to use XenAPINFS - that
> SR is primarily for setups using host aggregates where you want to do in-
> pool migration between hosts.
> >
> > Bob
> >
> >> -Original Message-
> >> From: Bryan Solie [mailto:bso...@well.com]
> >> Sent: 06 August 2013 13:17
> >> To: openstack@lists.openstack.org
> >> Subject: [Openstack] Cinder basic questions
> >>
> >> I am running OpenStack grizzly on Ubuntu 12.04 with XenServer.
> >>
> >> I have tried configuring Cinder with a few different backends, but am not
> >> able to get the end-to-end attachment process working.
> >>
> >> I have tried LVM/iscsi driver: The volumes can be created.  I attach them,
> >> and it returns successfully.  There is an iscsi target on the Cinder host.
> There
> >> is no sign that iscsi target is logged in from the instance I have 
> >> attached it
> to,
> >> and there's no indication from the CLI that it is actually attached to the
> >> instance.
> >>
> >> I have tried it with xensm storage backend, which I found in the Grizzly
> docs.
> >> But this backend no longer seems supported by the code.  Not sure if this
> is
> >> out of sync somehow; I don't know the history.
> >>
> >> I have tried it with XenapiNFS, and again, the volumes seem to be
> created,
> >> and I attach them without error, but they are not obviously visible from
> the
> >> instance.  I can't find an iscsi target when I create the volume in this 
> >> case;
> I
> >> don't know whether there is supposed to be one with this driver or not.
> >> Code suggests no, but I am a Python newbie.
> >>
> >> The one that seems most successful for me is the LVM/iscsi driver, but
> how
> >> do I get the volume to automatically login to the created iscsi target?
> >>
> >> Is there a requirement to use XenapiNFS if I am on xenserver?  What are
> the
> >> advantages?  And again, how do I get created volumes to be visible from
> >> inside the instance?
> >>
> >> The doc says with xenserver I can attach volumes only to paravirtualized
> >> guests.  I have looked at the PV-bootloader for the instances I have
> created,
> >> and it is blank.  As I understand it, they need to be on pygrub to be
> >> paravirtualized?  Again, I am a little new on this, and I have not been 
> >> able
> to
> >> get much out of google.
> >>
> >> Any help would be appreciated.
> >>
> >> Bryan
> >> ___
> >> Mailing list: http://lists.openstack.org/cgi-
> bin/mailman/listinfo/openstack
> >> Post to : openstack@lists.openstack.org
> >> Unsubscribe : http://lists.openstack.org/cgi-
> >> bin/mailman/listinfo/openstack


___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Cinder basic questions

[Bob Ball]

Hi Bryan,

The volumes should appear in the guest under /dev/xvd when they are added.

As long as the guest has PV drivers it should work - so HVM guests (such as 
Windows or if you have installed Linux as HVM) with PV drivers will also see 
these volumes added (although clearly with Windows they don't show up under 
/dev!)

If using iSCSI then the target should appear on the xenserver host as a PBD - 
so "xe pbd-list" should show one up.  The code that creates this is 
https://github.com/openstack/nova/blob/master/nova/virt/xenapi/volume_utils.py#L51
 - although that's not too obvious because we actually create a XenServer SR 
and allow that to create the iscsi connection.

You're right that xensm isn't supported in the current code and the docs need 
to be updated for Havana, but you don't have to use XenAPINFS - that SR is 
primarily for setups using host aggregates where you want to do in-pool 
migration between hosts.

Bob

> -Original Message-
> From: Bryan Solie [mailto:bso...@well.com]
> Sent: 06 August 2013 13:17
> To: openstack@lists.openstack.org
> Subject: [Openstack] Cinder basic questions
> 
> I am running OpenStack grizzly on Ubuntu 12.04 with XenServer.
> 
> I have tried configuring Cinder with a few different backends, but am not
> able to get the end-to-end attachment process working.
> 
> I have tried LVM/iscsi driver: The volumes can be created.  I attach them,
> and it returns successfully.  There is an iscsi target on the Cinder host. 
> There
> is no sign that iscsi target is logged in from the instance I have attached 
> it to,
> and there's no indication from the CLI that it is actually attached to the
> instance.
> 
> I have tried it with xensm storage backend, which I found in the Grizzly docs.
> But this backend no longer seems supported by the code.  Not sure if this is
> out of sync somehow; I don't know the history.
> 
> I have tried it with XenapiNFS, and again, the volumes seem to be created,
> and I attach them without error, but they are not obviously visible from the
> instance.  I can't find an iscsi target when I create the volume in this 
> case; I
> don't know whether there is supposed to be one with this driver or not.
> Code suggests no, but I am a Python newbie.
> 
> The one that seems most successful for me is the LVM/iscsi driver, but how
> do I get the volume to automatically login to the created iscsi target?
> 
> Is there a requirement to use XenapiNFS if I am on xenserver?  What are the
> advantages?  And again, how do I get created volumes to be visible from
> inside the instance?
> 
> The doc says with xenserver I can attach volumes only to paravirtualized
> guests.  I have looked at the PV-bootloader for the instances I have created,
> and it is blank.  As I understand it, they need to be on pygrub to be
> paravirtualized?  Again, I am a little new on this, and I have not been able 
> to
> get much out of google.
> 
> Any help would be appreciated.
> 
> Bryan
> ___
> Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
> Post to : openstack@lists.openstack.org
> Unsubscribe : http://lists.openstack.org/cgi-
> bin/mailman/listinfo/openstack

___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] Adding Ubuntu Images to Xenserver Powered Devstack

[Bob Ball]

The qcow2 images provided by Ubuntu aren't set up correctly for the Xen console 
- this means that the output during the boot process doesn't go to the screen.

If you leave the image it should eventually (after perhaps 5 minutes on first 
boot) reach the login prompt.

If you have injected credentials then you should be able to login using them, 
however I believe that if you have left it to Ubuntu to generate a random 
password this is displayed on the screen at a point which is not visible 
through the VNC console.

The three-part images do not suffer from this, and are set up correctly for the 
Xen console.

Thanks,

Bob

From: Qin Jia [mailto:jiaqin1...@gmail.com]
Sent: 06 August 2013 04:24
To: openstack@lists.openstack.org
Subject: [Openstack] Adding Ubuntu Images to Xenserver Powered Devstack

Hi,

I installed devstack on top of xenserver. Now I want to add a ubuntu image to 
it. I downloaded qcow2 image from here ( 
http://cloud-images.ubuntu.com/precise/current/) and used the guideline here 
(http://blogs.citrix.com/2012/10/17/upload-custom-images-to-a-xenserver-powered-openstack-cloud/)
 to convert the image to vhd format and add to openstack using glance.

But when I start the instance, it got shut-off automatically. When I try to 
restart the instance in Xensenver, it got stuck at the messages:

[ 1.612617] EXT4-fs (xvda1): mounted filesystem with ordered data mode. Opts: 
(null)

[ 2.245259] EXT4-fs (xvda1): re-mounted. Opts: (null)

One thing to mention is that I successfully added a cirros image to xenserver 
using the steps above.

So what's the problem here? Could any one give any suggestions?

Thanks in advance!

--
Best Regards,
Qin Jia
Computer Science
Cornell University
___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack

Re: [Openstack] openstack with xcp

[Bob Ball]

I would recommend using XenServr 6.2.  As Logan mentioned, XenServer is now 
fully open source and there is no paid for edition (although of course, Citrix 
will continue to offer support contracts).

XenServer 6.2 can be installed in "Pure OSS" or "XCP mode" which does not 
install the few components which are not owned by Citrix and cannot be open 
sourced - e.g. some drivers.

We will be updating this information soon to make it clear where to look for 
answers around what has happened with XCP - this is xenserver.org.

If you need any help with installing OpenStack on XenServer 6.2 (rather than 
the old XCP) - just give this list a shout (making sure you put Xen in the 
subject so that it is picked up by filters), or use the 
xs-de...@lists.xenserver.org mailing list as an alternative.

Thanks,

Bob


From: Logan McNaughton [lo...@bacoosta.com]
Sent: 02 August 2013 22:00
To: Jitendra Bhaskar
Cc: openstack@lists.openstack.org
Subject: Re: [Openstack] openstack with xcp

The relevant documentation is found here:
http://docs.openstack.org/grizzly/openstack-compute/admin/content/manual-install-on-xenserver.html
http://docs.openstack.org/grizzly/openstack-compute/admin/content/introduction-to-xen.html#xenapi-install

Essentially what you have to do is boot a VM on the XCP hypervisor (Ubuntu or 
CentOS or whatever), and install nova-compute on that VM, which will 
communicate with and manage the hypervisor.

As a side note... much of this information is out of date now since Citrix 
open-sourced XenServer, and discontinued XCP


On Fri, Aug 2, 2013 at 12:14 AM, Jitendra Bhaskar 
mailto:jeetuind...@gmail.com>> wrote:
Hi All,

I installed openstack with three node. Controller, network and compute. Now I 
installed xcp on another system. So I dont know how to proceed to integrate 
XCP. Can anybody share some clue ?




--
 Thanks & Regards
Jitendra Kumar Bhaskar
Cell:- +91 7306311531
  :- +91 9989743042


___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : 
openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack


___
Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Post to : openstack@lists.openstack.org
Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack