openstack
/
tripleo-ci
Code Issues Proposed changes

Default libvirt_cpu_mode to none when using qemu 

An upcoming update to qemu-kvm (2.6.0) with CentOS/RHEL 7.3 can break
existing implementations where qemu is the configured hypervisor and
'host-model' is used.
As such, ensure libvirt_cpu_mode is set to none when qemu is in use.

Worth noting that devstack has been setting 'none' [1] as well.

[1]: 1c442eebc8/lib/nova_plugins/hypervisor-libvirt (L42)

Change-Id: I4a7c3ced59c46fb2582122260922813b09a3d042
This commit is contained in:
David Moreau-Simard 2016-12-13 13:53:41 -05:00 committed by Emilien Macchi
parent a04150eb99
commit d4f72b0302
3 changed files with 8 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
test-environments/multinode.yaml
View File

@ -42,3 +42,5 @@ parameter_defaults:
ControllerExtraConfig:
nova::compute::libvirt::services::libvirt_virt_type: qemu
nova::compute::libvirt::libvirt_virt_type: qemu
# Required for Centos 7.3 and Qemu 2.6.0
nova::compute::libvirt::libvirt_cpu_mode: 'none'

3
test-environments/worker-config-mitaka-and-below.yaml
View File

@ -14,3 +14,6 @@ parameter_defaults:
NeutronWorkers: 1
NovaWorkers: 1
SwiftWorkers: 1
NovaComputeExtraConfig:
# Required for Centos 7.3 and Qemu 2.6.0
nova::compute::libvirt::libvirt_cpu_mode: 'none'

3
test-environments/worker-config.yaml
View File

@ -5,3 +5,6 @@ parameter_defaults:
heat::api_cloudwatch::enabled: false
heat::api_cfn::enabled: false
heat::rpc_response_timeout: 600
NovaComputeExtraConfig:
# Required for Centos 7.3 and Qemu 2.6.0
nova::compute::libvirt::libvirt_cpu_mode: 'none'