Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

Bug 1110380

Summary: Expose nova.conf virt_type parameter in staypuft
Product: Red Hat OpenStack Reporter: Lars Kellogg-Stedman <lars>
Component: rubygem-staypuftAssignee: Scott Seago <sseago>
Status: CLOSED ERRATA QA Contact: Toure Dunnon <tdunnon>
Severity: low Docs Contact:
Priority: high    
Version: Foreman (RHEL 6)CC: ajeain, hbrock, mburns, rhos-maint, sgordon
Target Milestone: ga   
Target Release: Installer   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2014-08-21 18:04:34 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description Lars Kellogg-Stedman 2014-06-17 14:14:34 UTC
For testing the staypuft installer in a virtual environment it would be tremendously useful to be able to specify virt_type=qemu in nova.conf.  Setting this manually requires disabling puppet, which will otherwise reset virt_type to the default settinge every time it runs.

Comment 2 Russell Bryant 2014-06-19 14:29:32 UTC
I checked with danpb on IRC, and confirmed that when nova asks libvirt for a VM with KVM, it will invoke qemu such that it will fail if KVM can not be used.  So, this is needed for nova to work in a test environment using VMs.

Comment 3 Stephen Gordon 2014-06-19 14:38:40 UTC
To be explicit, appropriate action for now is to expose the kvm v qemu option.

Comment 4 Russell Bryant 2014-06-19 14:41:00 UTC
(In reply to Stephen Gordon from comment #3)
> To be explicit, appropriate action for now is to expose the kvm v qemu
> option.

Right, there are other values that nova accepts for this option, but they should *not* be allowed.

Comment 5 Stephen Gordon 2014-06-19 14:46:26 UTC
In the future we may also add support for other drivers, so I am thinking a radio option like this:

    Driver:
        (    ) Libvirt/KVM
        (    ) Libvirt/QEMU
        (    ) vCenter

Even though Libvirt/KVM and Libvirt/QEMU are really the same driver with different virt_type setting the this should be handled behind the scenes.

Similarly when we add vCenter it has a host of other options to configure, different to the Libvirt ones, but we can consider that in a later phase.

Comment 6 Russell Bryant 2014-06-19 14:56:00 UTC
Also note that packstack does this detection and setting automatically using the following bits:

  packstack/puppet/modules/packstack/lib/facter/is_virtual_packstack.rb

  Facter.add("is_virtual_packstack") do
    setcode do
      Facter::Util::Resolution.exec('grep hypervisor /proc/cpuinfo > /dev/null && echo true || echo false')
    end
  end

Comment 7 Russell Bryant 2014-06-19 14:57:55 UTC
(In reply to Russell Bryant from comment #6)
> Also note that packstack does this detection and setting automatically using
> the following bits:
> 
>   packstack/puppet/modules/packstack/lib/facter/is_virtual_packstack.rb
> 
>   Facter.add("is_virtual_packstack") do
>     setcode do
>       Facter::Util::Resolution.exec('grep hypervisor /proc/cpuinfo >
> /dev/null && echo true || echo false')
>     end
>   end

While this seems like a handy puppet approach to doing the check, I don't think it's right.  I don't get this on my laptop for example.  I think the check needs to be for something else (vmx or svm, I think).

Comment 18 errata-xmlrpc 2014-08-21 18:04:34 UTC
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

http://rhn.redhat.com/errata/RHBA-2014-1090.html